Five things I’d do this month.

1. Feature Banner is becoming the Header

Remember the Feature Banner ? That hero asset that only big apps were able to get? Well, now it’ll be available for all very soon as Header, and editable directly in App Store Connect. It can be either an image or a video, completely separate from your screenshots. Very versatile, it can be used for either branding, as seasonal asset, or a launch.

One approved asset now shows up across 3 surfaces:

• Your main Default page.

• Custom Product Pages: a different Header per page, plus upgraded deep linking that drops users onto a specific in-app screen after install.

• Product Page Optimization: a testable element , that can also be previewed before testing.

All of it flows from one Asset Library, and you submit assets for review without updating your app version or even declaring where you'll use them. Approved once, it lives on organic and paid whenever you want.

2. New Apple Ads creative asset

In Apple Ads you also get a new asset for Today tab and Search Results ads.

This means your ad creative can carry real promotional copy, from seasonal hooks, sale messaging to a flat "book now and save".

Do this now: create assets for the new format too so you’re ready for the launch in Fall. Think Black Friday offers, Christmas package, New Year promotions… First movers get a cleaner, less crowded placement while everyone else is still producing v1 at launch.

3. Maximize Conversions: the AI bidding with target CPA

If you're still building around CPA caps, stop. Apple is deprecating the CPA cap in favor of Target CPA via Maximize Conversions, with a campaign average, not a per-ad-group ceiling, so bids flex up on valuable queries and down elsewhere while your daily budget caps spend.

Two things worth knowing:

• It's a keyword-discovery machine that settles with data. For example, a campaign at 50 installs/day learns in two weeks what a 5/day one takes ten. So switch on a controlled test now, and it'll be dialed before Q4 instead of calibrating on expensive November traffic.

• If you automate anything, note Apple shipped Campaign Management API 5 with breaking changes. Older integrations need a rebuild (ask me how I know 🙃).

From what I heard and what I’ve been testing, Maximize Conversion has been working quite well. You can also check out latest Apple Ads Success Stories:

• Toca Boca World got a clear read on what was working within a week, and up to 88% better acquisition efficiency in new markets.

• Taskrabbit held positive ROAS for several consecutive months and shifted budget into seasonal categories, without touching their default storefront.

Quick setup: budget at 10× your target CPA/day (it needs 5 conversions/day just to learn), leave the Automated Ad Group untouched, add campaign-level negatives so you’re not bidding against your own Exact terms, and don’t touch anything for two weeks.

🎯 You Might Also Like:

4. Map your Q4 > Q5 in one sitting

Block an hour.

Q4 is the loud stretch, from Halloween to Black Friday & Christmas.

But don’t stop at Dec 25, before Q5 is coming right after, from late December into late January, where you get the lowest CPIs of the year on ASA, with freshly-activated devices, and most categories at peak intent.

Here’s what you should do:

• Pull last year’s numbers: Which moments actually drove installs and revenue? You might want to keep some budget for the end of January, and let others spend everything at the start.

• Draw the calendar: Halloween → Black Friday → Christmas → Q5 resolution surge. Mark the dates that matter for your category, not the generic retail ones.

• Split the budget: plan for Q4 AND Q5, and prepare before you take time off, because holidays might be the biggest opportunity to scale.

• Lock creative themes per moment, then reverse-engineer your asset deadlines. Header + ad creative both go through review, so back-date from the launch.

• Decide which campaigns carry which moment: brand defense over the holidays, Today tab to grasp the high traffic, discovery / non brand into the Q5 window.

• Confirm Maximize Conversions is live and settled before December, so it enters Q5 already trained.

• The new Apple Ads Platform API is live (v1.0.0), and it’s one API for both App Store and Maps campaigns. It’s what replaces Campaign Management API 5, which Apple is sunsetting Jan 26, 2027. If you automate anything, build on the Platform API now; don’t start anything new on v5.

5. Build one AI process for all of this

The gap between needing a tool and having one has closed.

Start with the highest-leverage, most tedious job: keyword categorization. The brilliant Kevser Imirogullari open-sourced her 3D Keyword Framework as a Claude Code skill. Feed it an export from AppTweak, Sensor Tower, MobileAction or Astro and it sorts your pool by Type, Relevancy and Segment into five groups, each with its own metadata slot, ASA treatment and bid strategy. Several hours of work down to 30 minutes, and it’s only Act 1 of 3, with campaign creation and the optimization loop coming.

From there the same pattern scales: pull metrics via the Apple Ads API and let an agent flag your under-performers, or wire Astro’s tracking into the loop. Which is exactly the guide I’ve been working on for few weeks now (and promised). But since Apple Ads Platform API v1.0.0 landed, I’ve had to update my current setup and it’s taking longer than expected.

Apple Ads × Claude Code is coming, hopefully next issue.

🎯 You Might Also Like:

See you soon?

After half a year travelling abroad, I’m back for industry events.

• Mobidictum in Istanbul, Oct 7–8

• Business of Apps Berlin, Nov 12

Open to meeting anyone building in this space, so feel free to reach out!

📣 Good ideas are better shared.

Send this to your team, or anyone building an app right now, they might find it useful!

1. Start with the keyword strategy

Before touching anything, work out what the feature actually is in search terms.

A few rules I don’t break:

• The title stays untouched, at least until the launch is over with, then you can test again. It’s very risky to try something new at this point, so keep your sure assets as they are.

• The subtitle carries more weight than the keyword field, so I move it carefully and keep the riskier tests for the keyword field.

• New keywords replace the old ones. Pull the terms you’re not ranking on anyway, put the feature terms instead.

US first, then check other English geos before doing some research for the rest of the World. English-speaking markets can usually share the same subtitle, but the keyword field is where you localize (yes even for English speaking countries): different competitors, different search terms popularity and difficulty, different audience entirely.

2. Prepare the screenshots, and localize them properly

Your screenshots have to show the feature. If it’s the headline, it goes in the first one to three, the ones people see in the search results before they even tap.

Expect your conversion rate to move, most likely down at first. That’s normal, changing proven creative almost always costs you something short-term. Don’t revert in a panic. Look at it over time, and treat it as the start of a testing cycle rather than a final answer.

Then localize them, and properly. Duplicate the set and translate the copy for every language you’re targeting, then swap the in-app screens for the right localized ones and adjust the type, the sizing, the fit, so the translation doesn’t break the layout. The copy is the bulk of the work; the screens are the finishing.

The same base works for your Custom Product Page variants, which is exactly why it’s worth setting it up cleanly the first time.

3. Don’t sleep on the in-app event

The in-app event isn’t just an announcement, as the metadata on the event card starts ranking you on those terms, so you should choose them the way you’d choose any keyword, and track them.

But don’t write the card for the algorithm. It still has to rank, that part’s real, but the card is also advertising, and the person reading it has to actually believe in the feature. Users want to know what this app and new feature does for them, what they actually get out of it.

Let’s say that “Track your calories” is the feature. Okay, cool. But what it’s really about is eating better, feeling healthier, the lifestyle they’re reaching for. That’s the part that makes someone tap.

So you balance the two: enough of the feature and the keywords to rank, but brand-first, with a real benefit and a bit of emotion behind it.

This one is judgment-heavy, keywords on one side and brand on the other, and it’s the piece I trust least to shortcuts. You’re checking which terms are actually performing in each geo and organizing the event around both at once.

4. Submit for featuring

Submit the app and the event for featuring.

If you’re on a big app, your odds are genuinely better, so use that. Do it early even when the timing feels too tight, and don’t count yourself out just because you’re late. As long as the event is live, there’s a chance.

Here’s a great resource that I’ve shared here before, created by Anastasia Radek, The App Store Pitch Guide.

5. Push it with Apple Ads

When the new page is live, put some weight behind it. You get a nice compounding effect: the ASA traction, the fresh metadata, and the new creative all landing in front of new eyes at the same time.

Where I’d spend:

• Generic keywords related to the new feature.

• Competitor keywords, both competitors who don’t have this feature yet, as well as the ones who do, on which you were not converting before.

• Custom Product Pages, if you planned them. And if you didn’t, consider this your reminder for next time.

• If you have the traffic, split ad groups by audience, one catch-all, then new & returning. Send returning users to a CPP that leads with the new feature, that’s the thing that gets them to open the app again.

The AI stack I used, and exactly how

All of the above is the what. Here’s the how.

• Keywords and metadata: Astro MCP + Claude Code.

This is where AI does the most.

You know I love Astro, as I keep recommending it everytime there’s an opportunity to do so. There’s also a reason for it: it’s simple yet complete. No limits on keywords, and a MCP that dropped months before anyone else did (still waiting on the big platforms to share theirs).

For this project, I drive Astro through Claude Code and ask it to translate and localize the US keyword set for each geo, pull the right generics for the market, and flag the competitor keywords actually performing there.

Then tag everything for each geo, using a simple set of rules, so you can see at a glance what’s worth keeping and what’s not:

• Low volume? It goes. A keyword nobody searches for isn’t worth keeping.

• Good volume but the rank is very low (50 or lower)? It goes too. If you’re realistically not competing for it, holding it just wastes space you could use elsewhere.

• Good volume and a decent rank? That one stays, and you protect it and try to improve.

Everything that fails those rules is a slot you can hand to the new feature keywords. Then I export the whole thing in seconds, analyze it clearly, share with the team, and decide what to keep for the new metadata.

🎯 You Might Also Like:

• Localising screenshots and CPPs: Canva MCP + Claude.

[…] Want to read the rest?

Become a free or paid subscriber to get access to this post and other subscriber-only content.

A paid subscription gets you:
⭐️ Deep-dives & exclusive issues
⭐️ Tools, guides and access to archives
⭐️ Regular Q&As

📣 Good ideas are better shared. Send this to your team, or anyone building an app right now, they might find it useful!

Splitting this into three parts: ASO, Apple Ads, and GEO.

🎯 You might also like:

ASO: the growth lever most teams still aren’t pulling

Here’s the number that got me: 90% of apps never A/B test their icon, 90% never test a video, screenshots fare only slightly better, with 35% test twice or more in a year.

If you’re running Apple Ads without ever having tested your creatives, you’re paying to send traffic at a conversion rate you’ve never actually tried to improve.

The pattern holds across ranks too: the top 20 apps by downloads refresh their metadata far more aggressively than everyone else, and same for top games which update screenshots around 12 times a year.

Correlation isn’t causation, but whether they test more because they’re ranked higher, or rank higher because they test more, the top of the store simply doesn’t stop iterating.

My take: if you haven’t tested anything in your last 12 months on App Store Connect / Play Console, it’s worth flagging. Start with metadata, which is the easiest and requires the least resources. Use a tool like Astro to follow progress on keywords ranking, and be patient.

Apple Ads: New placement and pricier Discovery

Apple rolled out additional ad placements in search results on March 25, 2026. AppTweak compared the 28 days before and after across ~2,800 US advertisers, and Discovery campaigns saw the biggest reach gains by far: +44.9% impressions, +41.6% taps, +35.8% installs. Generic, Competitor and Brand campaigns moved in the single-to-low-double digits.

Here’s the part that’s more interesting than “more inventory = more volume”: cost per tap dropped across every campaign type, including Discovery (-1.9%). But cost per install actually rose for Discovery (+4.3%), the only campaign type where CPI went up. Brand, Generic and Competitor CPI all fell (-2.6% to -3.8%).

My theory: this isn’t only about the second placement. Maximize Conversions campaign dropped around the same time as the new ad placement. It also requires pairing it with a Discovery / Automated Search Match ad group, which pulls in a wave of low-volume, previously-untapped keywords.

[…] Want to read the rest?

A paid subscription gets you:
⭐️ Deep-dives & exclusive issues
⭐️ Tools, guides and access to archives
⭐️ Regular Q&As

📣 Good ideas are better shared. Send this to your team, or anyone building an app right now, they might find it useful!

• Store search is semantic

• Web stops rewarding volume

• AI now answers directly based on users’ profile

Three surfaces, one shift.

On top of this, we thought it’d be even better if we actually looked at real apps, and applied the theory to reality.

We picked the US Health & Fitness category. Health apps live with regulated metadata, health-data and medical-claim restrictions, and review scrutiny other categories do not have.

But the sharper problem is on the web, and it’s a double penalty. Fitness, nutrition and wellness content is what Google calls YMYL “Your Money or Your Life” content that can affect someone’s health or finances. Google holds it to a far higher E-E-A-T bar: visible expertise, real credentials, trust signals. Most fitness apps run thin, anonymous, SEO-by-numbers blogs that fail that bar completely.

Here’s the trap: the thin content fails twice. It doesn’t rank on Google, and it doesn’t get cited by AI engines because both lean on the same authority signals.

🎯 You might also like:

The apps that seems to “have it figured out”

The most instructive audits aren’t on struggling apps, but on the ones that look like they’ve already won. 2 patterns keep showing up:

• The incumbents built a discovery machine 5 to 8 years ago, it worked, and they’ve barely touched it since. You can watch the slow decay in the charts as newer apps eat their edges.

• The newcomers have the opposite problem: they’re winning right now, on product and virality, with nothing underneath.

Four apps, all winning but building discovery debt.

Ladder (#7 US Top H&F Charts) with the abandoned blog

Ladder did the hard thing right. A real content blog, written by a credentialed coach (NASM-certified CPT) which is precisely the E-E-A-T signal health content needs and almost nobody has.

Then they stopped.

The most recent post is dated February 2022. And the whole thing sits on a subdomain, blog.joinladder.com, walled off from the main site. So whatever authority it earns passes almost nowhere useful. They nailed the part most apps can’t do, but did not push there.

Subscribe now

BitePal (1M+ users), the newcomer with no foundation

BitePal wins in 2026: slick product, viral hooks, a million-plus users, and a homepage that’s a hero, 3 feature cards and 3 weight-loss testimonials.

Credit where it’s due as the site is server-rendered, so AI can actually read it. That’s actually more than most.

But on the downside, there’s no blog, no informational content, nothing for Google to rank or an LLM to cite beyond the brand name itself.

The weight-loss app is making specific before-and-after claims, with no named expert anywhere on the site, and has a “Sources” page that is, well, pretty basic.

The science meant to back the product lives on a simple page, almost hidden . It passes zero authority and barely counts as content at all.

Right now they don’t need the web. But the real question is how long that lasts.

Unrot got a launch sprint, then… nothing

Unrot is an app, pretty similar to Bitepal in its branding, which is meant to replace doomscrolling with healthy habits.

It has more content that Bitepal, and even opened a blog with 5 posts, answering exactly the trend-driven queries its audience types into Google and ChatGPT: “What is brainrot?”, “Dopamine detox doesn’t work”, “Rebuilding your attention span”.

All five went up in a single week in early July 2025, then nothing. And as of writing, only 2 of the 5 are actually indexed. In a topic moving this fast, that’s the worst possible place to stop since AI engines tends to cite fresh content. A blog that stops doesn’t hold and fades.

But the worst finding is in a file they’ve probably never opened. Unrot’s robots.txt blocks GPTBot, ClaudeBot, Google-Extended, CCBot and Applebot-Extended. Except they didn’t do it… Cloudflare did. The block sits inside # BEGIN Cloudflare Managed content, injected at the edge, on by default, most likely in a dashboard toggle they’ve probably never seen.

And it’s the wrong half. Those aren’t the crawlers that cite you: OAI-SearchBot, PerplexityBot, Googlebot and Applebot are all still allowed. Those are the crawlers that train on you. The default is shaped with publishers in mind: protect the IP, keep the citation traffic. Sensible, if you monetize pageviews, like medias. For apps, it’s a different story.

They wrote the right content, for the right queries, and then instructed the machines not to learn from it.

LeanBites, a goldmine locked

LeanBites is an app that scores 30,000+ menu items across the chains people eat at: Chipotle, Starbucks, McDonald’s, Chick-fil-A.

That dataset is the single most valuable content asset in its category. “Healthiest order at McDonald’s”, “Lowest-calorie Starbucks drink”, “Best macros at Chipotle”… These are enormous search volume and AI-answer queries, exactly the kind of specific, comparative, data-backed question a model loves to have a source for.

The thing is that every bit of it is sealed inside the app: no blog, no indexable page whatsoever. LeanBites is invisible for the exact questions it was built to answer.

🎯 In case your missed it:

So how do you actually build E-E-A-T?

E-E-A-T is mandatory, especially in health where it’s non-negotiable. So here’s a quick checklist if you have an app in the category:

1. Put a human on it: a named author with real credentials, a bio, a photo, and a link to their LinkedIn or professional profile. Ladder is the only one of the four that did this.

2. Mark that human up: person schema with sameAs pointing to their credentials. This is you telling the machine who said it, in the machine’s own format. Without it, a model reading your page has an anonymous claim.

3. Cite your sources on your own domain, build internal linking to support your own content and facilitate crawlers’ work.

4. Show your work: original data beats another listicle every time. LeanBites is sitting on 30,000 scored menu items and publishing none of it.

5. Keep going, as freshness is a ranking and citation signal.

Conclusion

None of these 4 are failing companies. That’s the entire point of this article. I’ve been working with many mature apps, that are struggling to stay on top. Competition is tougher than ever, and old organic strategies are eroding quickly. Paid-and-viral might work for a time, but won’t run forever if you’re missing solid foundations.

Organic discovery is getting more opaque, more crowded, harder to differentiate in. The apps that win the next cycle won’t be the ones with the most keywords. They’ll be the ones that are present, clear, and everywhere intent forms.

✨ Want to stay ahead in Growth for Mobile Apps & Games?

Subscribe now & get the latest trends delivered straight to your inbox!

Subscribe now

The most useful tools I’m using now aren’t built by platforms, but by senior practitioners, shipping with AI what used to need a funded company and a roadmap. So here’s my list, from well-known ASO tools to new ones built by other growth marketers that just needed it.

If you’re building something that you think is worth sharing, feel free to reach out and I’d be happy to feature it in a future edition!

The no-brainer: Astro ASO

The best price-to-value ASO tracker on the market, full stop. Unlimited keyword tracking, every geo, and competitor tracking, so you’re watching the whole landscape, not just your own app in a vacuum.

No keyword caps, all in one place and it’s only $108 Billed annually. You can also connect it to Claude Code via MCP and unlock the full experience. I wrote up the exact setup here.

TRY ASTRO

🎯 You might also like:

Visual work: screensdesign + before.click

ScreensDesign is one of my favorite tool. It’s both a flow and screenshot reference library. Recently, they also added a newer AI feature that drafts onboarding flows from scratch. I used the AI screen creator on a recent client onboarding and it’s a genuinely good starting point. Both the library and the generator earn their keep for onboarding/paywall iteration.

Get 50% with the promo code NEOADS50

TRY SCREENSDESIGN

While ScreensDesign might be locking some content behind a paywall, before.click is the reference layer next to it: a deep archive of onboarding, screenshot, and paywall patterns - all for Free.

EXPLORE BEFORE.CLICK

The indie AI scene

This is the part worth paying attention to. Senior Growth Practitioners are now shipping real tools, answering to real needs, and they share them all for free.

NeoAds SOV Analyzer

I vibe coded this tool for my clients, to see exactly how visible you are on the Apple Ads keywords that matter. Share of voice by keyword and geo: where you’re dominating, where you’re invisible. 1st or 2nd position, trends by search terms, overview by geo or keywords.

Try the SOV Analyser

Review Retriever for iOS

Shared by Dash Antonova, Independent Head of Growth. Extract any app’s pain points straight into your LLM: draft CPP or onboarding copy, surface metadata angles, see what users like or don’t from your competitors app, and get clear user-feedback summaries without reading thousands of reviews.

TRY REVIEW RETRIEVER

App Store Visibility Checker

Produced by TopPanda, the tool shows where you actually appear on the AppStore: which "you might also like" and Browse placements are driving impressions, and whether a featuring is live.

TRY APP STORE VISIBILITY CHECKER

The pattern under all of this: tools are being built for genuinely specific use cases now.

Where we used to wait months for some big platform to maybe ship the feature we needed, we can now build our own, or grab the one a practitioner already built, in less time than it takes to get a sales call booked.

That’s the real shift: the gap between “I need a thing” and “the thing exists” has basically collapsed.

Subscribe now

Julie @ NeoAds - freelance Growth Consultant specialising in Apple Ads, ASO and CRO.

But if you walked away with “the App Store has a header now”, you missed it. The real move is quieter: Apple started routing discovery away from the listing, toward Siri, toward AI, toward intent. The storefront got a facelift, while the foundation moved.

The placement everyone’s talking about isn’t the one that matters

The search results asset is the one I’d build first. It’s the creative that decides whether someone taps at all at the top of funnel, where around 60% of installs still come from.

And for the first time you can break the screenshot format: landscape on a portrait app, a single big image or video where you used to get a cramped triplet. You can even pull a creative already winning on Meta or TikTok and test it straight in search.

The header doesn’t have to show your app

For a decade screenshots had one rule: show the product. That’s been quietly shifting as more apps treat the first screenshot as branding, not a feature shot. But the header is now officially free of it, carrying brand and mood.

The job of your screenshots is now becoming sharper: put the features and copy up top, readable above the fold, before anyone scrolls.

Two things I’d assume before the specs land:

• There’s a safe area: the system paints your app name over the header, so keep your own text and logos out of that zone.

• A header video is not a preview video. The preview shows the app working; the header sets a mood, often with no UI at all. Apple finally let you use for something other than the product.

Half of this is really about Ads

• The Today tab finally earns its premium. The placement that’s looked the same since 2022 is now larger, fully visible the second the store opens, CPP assets animating behind it. I’ve spent years telling clients the Today tab was an awareness placement with limited creative freedom. That’s over, and apps should now re-test it with a brand new strategy.

  

• CPP + header is a message-match play, and the easy win to set up in month one. Your ad creative, the header on the CPP it lands on, and the page itself can finally be one continuous visual.

• The Asset Library brings it together: pre-approve a batch of assets, then sync creative to a campaign’s timing. Think seasonal push, event launch, a reactive moment… different use cases to deploy new assets and align even more your paid and organic efforts.

🎯 You might also like:

Further than the store

A custom product page is the destination for whatever you run: Meta, Google, TikTok, your own site, a link in this newsletter. Now the header carries the story through, with the same creative people saw at the source.

And the part no one expected: Apple acknowledged that web traffic matters. Their walkthrough used a website selling yoga classes, linked to a CPP, the header reusing the site’s visuals, a deep link straight into the relevant in-app content. Web > store > in-app as one continuous journey.

So whether it’s paid social, search, your own site, or increasingly whatever an AI assistant decides to surface, you can finally control the first frame people hit on the store. Build a CPP and a matching header per major source.

Discovery is leaving the listing

This is the one.

App Intents is the only way an app surfaces in Siri, and the new Siri runs on a custom Google Gemini model. Apps that haven’t adopted App Intents are simply invisible to it. We’ve got a 2-to-3 year migration window, but the clock starts when iOS 27 ships this autumn.

Someone asks Siri “what’s the best app to track my workouts?” and Siri decides which one fulfils the intent. No browsing, no scrolling a results page, just a request and an answer. And it’s personalised, so two people asking the same question can get two different apps.

We’ve spent years optimising a listing for search. Part of the game is now optimising for an assistant that never shows the listing.

The question I can’t answer yet: do we get a “Siri” source in App Store Connect to measure any of it? If discovery moves to intent fulfilment and we’re flying blind on attribution, that’s a hole right where the growth is going.

And the part most takes this week will skip… in the EU and China, the new Siri isn’t available at launch because of DMA. EU developers can’t even test it in their own apps during development. So for a big slice of our market, this whole shift arrives late, with all the planning mess that implies.

The store is now recommending you to people who never searched

Siri isn’t the only thing deciding for users, the same shift is happening inside the store. Apple’s rolling out Personalized Collections: curated groupings like “Because You Play Backyard Birds > Mellow Atmospheric Games” on the Apps, Games and Search tabs, each with a line on why it’s relevant to that person, and evolve with usage and downloads.

Affinity-based discovery, Apple putting you in front of people because of what they already do, not because they typed your keyword.

It’s algorithmic and the “why” copy belongs to Apple, so there are no direct levers, but think of it like “You Might Also Like” scaled up, bringing extra Browse impressions. Limited to select markets and a few apps for now, expansion promised. Read it next to Siri and it’s the same message twice: discovery is moving toward what your app does and who already plays in your lane.

Native AI is coming for a few of your apps

The Camera and Photos are getting a real AI layer with generative editing, object removal, reframing a shot as if you’d moved the camera. Apple is openly pointing it at AI photo editors, and the same logic threatens calorie trackers, bill-splitters, background removers, anything whose core loop is “point your camera at a thing, get structured info back.”

I’d hold the panic, though. This might land like Currency Conversion in the Calculator app: native, genuinely useful, and somehow never as widely adopted as the “this kills X” headlines predicted. Native features win on convenience. They rarely win the people who actually care about the category.

If your app is a feature, worry. If it’s a workflow, you’ve got room.

If your users are under 18… well good luck

Child accounts are now mandatory for under-13s and stay active until 18. Parents pick which apps a child can use, Ask to Buy gates App Store and in-app purchases, and Apple is handing developers new APIs to tailor experiences by age. This is Apple answering regulators.

If your audience skews under 18 (games, social apps, entertainment..) this is real exposure. Acquisition, monetisation and even who can install you start running through a parental layer. I’d be auditing that funnel now and getting comfortable with the age-range APIs before they’re table stakes.

If your growth plan for the next year is “better screenshots”, well you might a be a bit outdated… The apps that win will treat the page as a living surface, get on App Intents early, and experiment with new assets and available features.

What are you all thinking? 👀

Julie - freelance Growth Consultant specialising in Apple Ads, ASO and CRO.

If someone forwarded this to you: subscribe here.

I love experimenting, testing new things, building new workflows, making them better. I also have an incredible partner who’s an even earlier adopter than me, so between the two of us I’m constantly reaching my Claude subscription limits and connecting new MCPs. So yes, here’s more MCP content, not only because I like it, but also because people engage with it.

The headline for this edition: I finally integrated AppsFlyer MCP! I said I would in a previous article, and here we are. But that’s just one piece of it.

This edition is about the full acquisition stack: what’s working, what’s still rough, and what I’ve started building on top of it. I’m managing 4 apps and 3 games across multiple clients. This is how I run it.

ASO + ASA: what’s changed since the last two articles

The Astro MCP setup is documented, as well as a first look at linking Apple Ads API.

What I want to talk about here is what daily use across several apps actually looks like. Because the multi-app context problem is real with constant switching between clients. So now the workflows are running, with frequent organic gap analysis, cannibalization checks, localisation updates, moving organic search terms into ASA… All of these that used to take few hours, now takes only few minutes. Results are there, clients are happy, apps are growing.

Subscribe now

Connecting the Apple Search Ads API: the short version

Apparently that’s something you’re interested in, so the full ASA API setup guide is coming! In the meantime, you need to prepare 3 things:

• Apple Ads API credentials

• A dedicated folder for Claude Code to run from, for privacy and security purposes

• A .env file with your keys.

Start Claude Code from inside that dedicated folder. From there, Claude can read campaign data, pull keyword performance, update bids or daily budgets, create ad groups, add keywords in bulk… all via prompt.

I’ve started slowly at first, with read-only actions until I fully trusted the setup. Then I asked it to generate CSVs with my changes so I could upload them directly into ASA. Once I realised this was working like a charm, I’ve started using it to bulk rename ad groups, update CPT bids for certain type of keywords only, and upload translated terms in dedicated ad groups.

The full guide with downloadable setup from auth flow, error handling, security config and the prompts I use will be dropping soon. Stay tuned!

The intelligence layer: markdown files, Claude.md & Claude Projects

Before I integrated AppsFlyer, I needed to solve a more fundamental problem. Claude has no memory between sessions, and every new chat starts from zero.

I found the solution in different places, from a workflow shared by Andrea Fuertes, to asking around and seeing how most people used structured files as a base and built from there.

The principle: before you connect any MCP or run any analysis, you create structured .md files for each client. That gives Claude everything it needs to work on that account from the first message, from account structure to ROAS targets by geo and time window, scaling rules, current priorities, historical context.

• In Claude.ai, you create a Project per client and upload those files as project instructions. Every session opens with full context already loaded.

• In Claude Code, the equivalent is CLAUDE.md: a plain text file that lives in your project folder and gets read automatically every time a session starts. No prompting, no re-explaining. Whatever you put in it becomes Claude’s permanent starting point: your rules, your account structure, your naming conventions, what it can and cannot do without asking.

I’ve also added a change log, saving automatically each session. Every action gets appended to a memory file with date, campaign, what changed, before and after values. Come back 3 weeks later, ask “what did we change on May 19th?”, and you get an answer.

Write it once. It applies forever, and evolves with the work.

It took a few hours to build properly. It’s saved far more than that. (Full breakdown in the Apple Ads API setup guide - coming soon).

How to set up CLAUDE.md for Apple Ads

CLAUDE.md is a plain text file Claude Code reads automatically at the start of every session for permanent context.

Step 1 - Create a dedicated project folder on your computer with one folder per client. /!\ Don’t run Claude Code from your home directory, keep it isolated.

Step 2 - Create the file and fill it in

## Required context files
- SKILL.md — role definition and analysis framework
- kpi-benchmarks.md — KPI targets and thresholds for this app
- output-templates.md — required output formats for reports

## Key facts
- App Store ID: [YOUR_APP_STORE_ID]
- Bundle: [your.bundle.id]
- Org ID: [YOUR_ORG_ID]
- Token refresh: python3 APPNAME-asa-auth.py (expires hourly)

## Rules
- Never make campaign, keyword, or bid changes without explicit approval
- Always save output files inside reports/, never the workspace root
- Follow bulk upload CSV format defined in output-templates.md

Three companion files live alongside it:

• SKILL.md defines how Claude should analyse and think about the account

• kpi-benchmarks.md holds your targets and ROAS thresholds (following Andrea Fuertes guide)

• output-templates.md locks the format for bulk upload on ASA so output is ready to use if API is not working (create keywords, update bids, add negative keywords…)

Step 3 - Start Claude Code from that folder

cd /your-client-folder
claude

To summarise, CLAUDE.md gives Claude facts, while SKILL.md gives it a framework for thinking about the account. The combination is what makes sessions feel like a continuation rather than a cold start everytime.

The token refresh command in Key facts is a small detail that saves real friction, since Apple Ads API Token has to be updated hourly. Claude knows to flag when auth is about to expire without having to track it manually.

AppsFlyer MCP: finally integrated!

In the last piece I said integrating a MMP was the next step for me.

AppsFlyer MCP works in Claude.ai desktop, and I’ve been using it daily for geo and campaign level ROAS analysis across client accounts.

For one of the account I manage, it’s pulling D1 & D7 ROAS across 40+ markets and read it against the context I’ve provided in my markdown files with objectives, goals, thresholds. That part has changed how I approach campaign reviews and reporting. Performance and targets live in the same window, and Claude can pull it clearly in front of me before I make any decisions.

The only caveat is that I’m not using it in the same session as the Apple Ads API, since that one lives in Claude Code, while AppsFlyer lives in Claude.ai. I could potentially do it, but the setup requires an AF Admin access.

So in practice:

1. I pull the ROAS picture in Claude.ai via the AppsFlyer MCP, spot a pattern, go deeper either via the MCP or in the dashboard directly

2. Act on it in Claude Code via Apple Ads API, giving it instructions on how to optimize (drop / increase bids, move keywords, change budgets…)

3. Report automatically in a Notion database so I can follow it whenever without having to connect

The full loop I described last time, from keyword to install to downstream attribution, all in one workflow, isn’t there yet. That’s still the next step, but I’m really not sure when I’ll be able to achieve it.

Building tools: this is the part I’m most excited about

There’s a category of work I’ve always done in Excel that I’ve never loved: templates. You know, template for performance analysis, template to look at share of voice overtime, template to review Custom Product Pages…

The output was never pretty, always hard to hand off, and needs fifteen minutes of explanation before anyone else can use it.

So I’ve started building my own, and replacing these templates with actual tools that are interactive, intuitive, properly shareable.

And guess what… the first one is almost ready!

It’s for Apple Ads, looking at Impressions Share of Voice and easily comparing First vs All Ad Positions. Because yes it’s exciting to have more inventory, but also a pain point if you had a strategy in place and don’t want to mess it up. Here’s a snapshot, with dummy data of course!

It’s free, and it’s dropping soon.

🎯 You might also like:

Conclusion

Let’s be honest… my work has been changing drastically in the past few months. Optimizing Apple Ads used to take most of my week, manually updating bids, running Excel scripts and creating templates to make my work slightly easier. And I’m not even starting with ASO, where keyword research across geos was simply a very large sheet of translations taken from Google.

Now the manual layer is delegated: keyword reviews, performance pulls, geo analysis, campaign restructures… everything handled. What’s left is the work that actually requires judgment and strategy: where to focus, which markets to push, what to test next.

At the end of the week, I do not work less, but I optimize more often. Campaigns are more efficient. I build tools that I can use to scale in less time that I would have taken to just run an analysis. That’s a real shift.

✨ Found this useful? Share it with someone building or growing an app.

Share Neo Ads

Quick update on where NeoAds is going

This newsletter has been free since day one. It’s staying that way… mostly!

I want to be honest about why this is changing: I’m spending hours building flows, testing tools, and figuring out how to optimize my work. I believe some of that is worth more than a free article, as it actually generates real growth and revenue.

That’s why I’ve decided to introduce paid tiers, while keeping the free one as it is.

Free: nothing changes. Weekly newsletter, everything that’s always been here, and some bonuses coming for the people who’ve been around since the beginning.

Paid subscribers: deeper breakdowns, early access to tools and guides, and regular Q&As where you can drop your app, your challenge, your numbers, and get actual specific advice back.

NeoAds+: limited to 3 spots per month. You pay for a year upfront, get everything in NeoAds+, plus a 1-hour 1:1 consulting session with me, either to answer any question about this specific setup, or get a live ASO or ASA Audit. I’m running a limited offer at $199 for the first cohort - Book before it’s too late!

Subscribe now

A lot going on, and more coming. It’s exciting, and honestly, I’m struggling to keep up sometimes. But I want to thank all of you for the time you take every week to read, share, and comment. I’ve realised this is only the beginning, and I have big plans for what’s coming.
Best,
Julie

Apple still hasn’t moved though, but the Almedia team shared they’re “continuing to work with Apple on iOS reinstatement”, so the door isn’t fully closed.

Same app, same business model.

Two platforms and potentially two very different conclusions.

Before we get into what that means practically, let’s roll back to what actually happened.

🎯 You might also like:

About reputation management.

If you work in mobile gaming, you already know Freecash isn’t a scam. The model is familiar as rewarded UA platforms have been part of the stack for years. You’ve either bought traffic from Almedia, competed with them, or know people who do.

What happened wasn’t a compliance review. It all really went down when TechCrunch wrote a piece on the app, and contacted Apple for comment. Apple gave no comments, however the app was gone the same day. Google followed 24 hours later.

[…] Want to read the rest?

Subscribe to NeoAds to unlock the rest.

Become a free or paid subscriber to get access to this post and other subscriber-only content.

Subscribe now

A paid subscription gets you:
⭐️ Deep-dives & exclusive issues
⭐️ Tools, guides and access to archives
⭐️ Regular Q&As

I connected the Apple Search Ads API to Astro via Claude Code. No third-party, no internal tooling team, just an API key, a MCP setup, and some time.

It’s not finished yet (more on that in a moment), but what’s already working has changed how I think about the ASO × ASA relationship.

So I wanted to share where I’m at.

The problem I was trying to solve

ASO and ASA have always been run in parallel. Different tools, different reporting, sometimes different teams entirely. But they’re looking at the same thing: what users search for, and whether your app shows up.

As a Growth Consultant, I either manage both or just one of them, but I always try to link them in some way or another.

The gap between them is where opportunities get missed.

• Keywords converting in ASA that never make it into your metadata.

• Organic ranking drops that nobody connects to a change in paid strategy.

• Competitor gaps spotted in ASO tooling that never get actioned in Search Ads.

The data exists on both sides, but it requires different access and setup to connect them. And I’m not even mentioning the time it takes to analyse the data and make something out of it.

That’s what I set out to fix.

This project was inspired by Kevser Imirojullari Claude Code setup connecting Apple Ads API, MMP, Astro, and Figma into a single workflow.

What the Astro MCP + Apple Ads API setup actually does

Astro is an MCP-native keyword tracking tool. When you connect it to Claude Code alongside the Apple Search Ads API, you can run both datasets in the same conversation, with no switching tabs or exporting reports manually.

Last month, I wrote an article on how to set up Astro MCP (you will need this if you want to link it with ASA):

So here’s where linking both gets useful:

• ASA to ASO: Keywords that are converting in paid are validated search intent. You don’t need to guess whether Japanese users search for a term, since your ASA data tells you. From there: push high-converting terms into your localised metadata or your screenshots copy. What’s working in paid should be informing your organic strategy.

• ASO to ASA: Astro surfaces keywords rankings, popularity, competitors in top terms... When you run that through Claude alongside your current ASA keyword list, gaps become obvious fast. For example, terms competitors are ranking for organically that you’re not bidding on, or even competitors that appears on your top terms that you haven’t added to your campaigns yet, geos where you’re missing some top terms, keyword clusters you haven’t structured campaigns around yet.

Here’s a real example of what a session looks like in practice:

Claude checks which app is in Astro, confirms 378 keywords across 12 stores, and immediately offers options: pull current rankings, get keyword suggestions, extract competitor keywords, or add keywords directly from ASA search term mining reports. All from one prompt to link Astro MCP with Apple Ads API for a specific app.

What the workflows actually look like

Once both are connected, Claude can suggest concrete workflows combining both datasets.

The most immediately useful is the organic gap analysis: pull rankings from Astro, filter for keywords sitting between positions 5 and 20, and those become your priority ASA bid targets. You’re already relevant for those terms, just not top 3 organically yet, so bidding there makes sense.

The flip side (and one of my fav) is the cannibalization check, keywords where you’re already ranking #1-3 organically, but still bidding in ASA. Claude flags them automatically, to avoid paying for traffic you’re already getting for free.

Then there’s the loop between paid and organic tracking: take your converting search terms from ASA, push them into Astro, and monitor whether paid spend is actually lifting organic rank over time.

And on the competitor side: extract keywords your competitors rank for in Astro, cross-reference with your ASA campaigns, and you’ll find terms you’re not bidding on that are already proven in your category.

It can take actions, not just give analysis

Because the Apple Ads API is connected, Claude isn’t limited to surfacing insights as it can also act on them. Here’s a workflow I’ve been using during an account restructure:

1. Pull keyword opportunities directly from Astro or legacy campaigns

2. List them with current bids across campaigns / geos

3. Organise them in clusters to separate them in different ad groups (for CPPs later)

4. Add them to newly created campaigns / ad groups, re-using the same bids as before

All of this, without touching the dashboard once. Well actually I did create the campaigns manually, since I’m still early with this and going very slowly with automating campaigns management. Let’s not forget that Claude having access to the API can have real consequences.

But the direction is obvious and the tedious parts of ASA management (keyword mining, bid reviews, restructures..) are becoming prompts.

I’ve also been connecting other tools like Superwall or Mixpanel into Claude to get a clearer picture of what’s happening after the install. More on that in a future issue.

The current MMP gap

The piece that would complete the loop is MMP integration of course, being able to tie ASA keyword performance back to downstream events. That’s where the real performance optimisation is.

I’ve spend some time trying to implement with Adjust, but it has proven difficult to connect so far.

Andrea Fuertes has written about this and has a setup worth following, so I’ll be testing her approach with AppsFlyer later this month.

Until the MMP piece is in, the workflow is powerful for keyword strategy and opportunity identification. It’s not yet a full attribution loop, more of an optimization one for now.

Want the full setup guide?

It took me a while to get this working, as I wanted to create a secure workflow, running Claude in a separate folder on my computer and not accessing all my files, making sure the API connections were setup correctly, configuring the MCP well, defining the prompting structure, etc.

If you’d like a full guide, reply to this email or leave a comment. If I get enough responses, I’ll build it out.

In the meantime, you can still check out my previous article on how to setup the Astro MCP:

✨ Found this useful? Share it with someone building or growing an app.

Subscribe now

MobileAction dropped their 2026 ASO Report: Keyword Trends, Visibility Benchmarks, and Top Apps in the US App Store.

In 6 out of 20+ major categories, “AI” is now the single most-used keyword in app metadata. Here are the occurrences of the word “AI” in each category, and if it appears in top 10, top 5 or even #1 for these categories.

Also, some striking notes:

• Productivity: AI is #1 keyword, above “notes,” “tasks,” and “daily”

• Photo & Video: #1 also, above “photo,” “video,” and “camera”

• Health & Fitness: #2, just below “health” itself

• Education: #2, just below “learning”

• Graphics & Design: #1, above “design” and “create”

• Lifestyle: #1, above “time,” “daily,” and “life”

• Utilities: #1, above “device,” “phone,” and “manage”

• Entertainment: #1, above “TV,” “content,” and “subscription”. Yes you read that correctly, “AI” now outranks “TV”.

What’s actually happening

There are two distinct dynamics here.

Dynamic 1: AI as a genuine utility signal.

In categories like Productivity, Photo & Video, and Utilities, AI is describing real, differentiated features: on-device processing, generated outputs, adaptive personalization.

• When Notion puts “Notes, Tasks, AI” directly in its title, that’s precise positioning.

• When Canva leads with “AI Video & Photo Editor” it’s describing something users can find immediately in the product.

In these contexts, AI is functioning as a capability keyword with relatively clear user expectations.

Dynamic 2: AI as a credibility badge with no substance.

In categories like Entertainment, Lifestyle, and increasingly Health & Fitness, “AI” is being used as a quality signal without any underlying specificity. It means “we have better recommendations” or “our algorithm is smart”. But these things are mostly invisible to users and indistinguishable from competitors.

This is the keyword equivalent of every app calling itself “the best”.

It once meant something, but now it’s noise.

The ASO mechanics behind the saturation

From a pure keyword strategy standpoint, here’s what the saturation means in practice:

1. Indexation is not the problem, ranking is.

   Ranking for “AI” in your category is easy, but getting in the top results requires authority, volume, and relevance signals. If you’re a mid-sized Health & Fitness app competing against MyFitnessPal or Calm, both of whom have massive download velocity and engagement, your “AI” keyword placement does essentially nothing.

2. Conversion risk is real.

   A user searching “calorie tracker” who lands on a page that leads with “AI-powered health companion” faces a messaging gap. The more generic your AI framing, the less it matches the specific intent that drove the search. This is the exact kind of metadata inflation that impacts Conversion Rate negatively.

3. AI saturation is real.

   The irony of AI saturation is that specific AI use cases, even with low volume, stay relatively crowded. “AI calorie counter”, “AI photo editor” editor for reels,” “AI habit tracker”… these carry a lower volume that the same keywords without “AI” in it, while having as much difficulty shown (see Astro screenshot).

🎯 You might also like:

The categories where AI is absent

What’s equally interesting is where “AI” doesn’t appear in the top 10: Games, Travel, Shopping, Navigation…

These are the categories where user intent is either highly transactional (Shopping, Travel) or real-time utility-driven (Sports, Navigation, Weather). Developers in these verticals are correctly identifying that AI framing doesn’t match the search behavior of their users.

This gives you a contrarian signal:

• In categories where AI saturation is highest, specificity is the edge.

• In categories where AI is absent, early mover advantage exists for apps where AI genuinely changes the product experience; think AI-powered flight deal alerts in Travel, or AI injury prevention in Sports.

Three apps, three different ways to handle “AI”

Gauth puts AI front and centre: “Gauth: AI Study Companion.” The screenshots they quantify the claim: “Solve 98.99% of questions”, “Live AI tutor explains in real time.” This is an app where AI is the entire product mechanism. The 1.4M ratings and #2 Education chart position suggest users are finding it exactly when they search for AI-powered homework help. AI in the title is honest, specific, and converting.

Canva, with 3.3M ratings and an established brand, the AI addition is a repositioning signal, not a discovery tool. They’re telling existing users the product has evolved, not trying to win a keyword. Notice they’re also running an In-App Event specifically about their AI launches, using every App Store surface available to make the claim feel current rather than generic.

Notion does something subtler, leading with the outcome. AI is third in the title and completely absent from the first screenshot, which reads: “Your life, beautifully organized”. The outcome leads. AI is positioned as a feature inside a broader product promise, not as the product itself, and captures the AI search intent without making AI the brand identity.

With or without AI

Looking at the health & fitness category now, especially around Calorie Tracker / Counter. What’s interesting, is of course, that this category has been completely disrupted by Cal AI. So I looked at competitors here, Lifesum & MacroFactor.

Lifesum takes the middle path: “Lifesum: AI Calorie Counter” with a screenshot headline that reads “Healthy eating. Simplified by AI”. The AI is present but attached to a concrete outcome. “Simplified by AI” makes a claim about the user’s experience, not just the technology stack. Still, it’s a competitive keyword landscape: per MobileAction’s data, Health & Fitness has 28,973 AI mentions in metadata, second only to “health” itself.

MacroFactor, on the other hand, doesn’t mention AI anywhere visible, and yet it sits at #54 in Health & Fitness with a 4.8 rating on 13K reviews, higher rated than Lifesum’s. MacroFactor is genuinely AI-powered under the hood, since they still use the word “Smart”, but the positioning leads with what the user gets, not how it works.

🎯 You might also like:

Use CPPs to fix the exact problem we just described

If “AI” as a standalone keyword is too saturated to rank organically, Custom Product Pages let you target AI search intent directly through Apple Search Ads, without touching your default listing.

The mechanic is straightforward: in Apple Search Ads, you assign a specific CPP to a specific ad group, gathering keywords by themes. For example, when someone searches “AI calorie counter” or “AI study app”, they see screenshots built for that exact intent.

MobileAction’s data shows 22,636 CPPs were created across the US App Store in 2025. Health & Fitness alone produced 1,792. Education hit 1,498. These are the exact categories drowning in generic AI metadata.

TL;DR

The question was never whether to use “AI” in your metadata. It was always whether your app needs it.

The data from MobileAction’s 2026 report makes the saturation impossible to ignore: 44,000 Productivity apps, 36,000 Photo & Video apps, 28,000 Health & Fitness apps all using the same word to say different things.

The apps that cut through aren’t doing anything complicated, they’re just attaching AI to something specific like a task, an outcome, a user who searches for exactly that thing. Otherwise, they’re leaving the word out entirely and competing on the strength of what the product actually does. Both strategies work. The one that doesn’t work is using “AI” as a badge and hoping users connect the dots themselves.

🎯 You might also like:

✨ Found this useful? Share it with someone building or growing an app.

Subscribe now

Missing ROAS, minimal reporting, features that feel like they were built ten years ago, no granular signals, no transparency into what’s happening.

And you know what? They’re not wrong.

For the story, I joined Apple in 2020, peak Covid & peak mobile gaming. I spent 3 years working with some of the biggest games and apps in EMEA. Puzzle, Casual, Action, Sports Betting… Brand you’d recognise, having big launches, wild budgets, CPIs higher than most campaigns daily budgets.

And I learned something that changed how I think about the platform entirely.

What I saw on the inside

The highs are genuinely high

Working at Apple means your actions have scale.

In my first year, I led a project that drove double-digit millions in revenue impact in a single quarter. That kind of leverage doesn’t exist at many places.

The clients are also the largest apps and games in the world, and when trust is established, the relationship shifts. You’re not just selling a platform, you’re advising. And the advice has to work, because everything is measurable, and performance doesn’t lie. You can’t push a product that doesn’t deliver.

You’re also learning from some of the best minds in the industry, and the people around are genuinely curious, ambitious, and extremely talented. I learned to sell, pitch, visualise data, present, network, lead a project, mentor… the list goes on.

But the size cuts both ways.

At a company of Apple’s scale, things move differently. An issue can be flagged, validated, understood by most… but it’ll still take months to become a product change. Not because people don’t care, but because the organisation is enormous, and alignment across that many functions, regions, and legal constraints is genuinely hard. It’s a structural reality.

So why doesn’t Apple Ads have ROAS?

This is the one we used to get asked about most. The honest answer: user privacy.

It is not just a marketing position. The same philosophy that drives App Tracking Transparency, SKAdNetwork, and Privacy Nutrition Labels on the App Store is the same philosophy that governs what Apple Ads can and cannot expose as reporting signals.

Connecting an ad impression to a downstream revenue event, which is what ROAS requires, involves user-level attribution at a depth that Apple has fundamentally decided not to provide. Not because the engineering team hasn’t thought of it. But because doing so would contradict the privacy architecture that Apple has made a defining product principle across the entire ecosystem.

🎯 You might also like:

Why new features take so long

A startup can ship a feature in two weeks. Apple cannot.

When you’re operating inside one of the most scrutinised companies in the world, every product decision carries implications that extend far beyond the feature itself. Think about it: the App Store is one of the most visited app ever. Every change means privacy review, legal & regulatory exposure across markets.

That’s why things like new ad placements, new reports, or new campaign controls move slowly. The bottleneck is the weight of operating at that scale with that level of scrutiny.

When I was there, we saw the launch of Search tab, then Today tab & Product Pages placements, then Custom Product Pages. Landmark features.

But we, the account managers talking to clients, often had no idea when they would ship. We knew just enough to answer basic questions. That’s not a complaint, it’s just how it works at that scale. The teams building these products are good. But they’re operating within a system that was not designed for fast iteration.

What this means in practice

If you’re running Apple Ads, there are three things worth keeping in mind:

1. Work with the signals you have, not the ones you wish existed. Impression share, tap-through rate, conversion rate by keyword group… these are enough to be directional. Yes it’s frustrating we don’t have more, but it is what it is. Apple knows these limitations, and be certain they'll do their best to deliver what's actually possible within the constraints they've set for themselves.

2. The organic-paid loop is the real advantage. What Apple Ads does well is the relationship between paid performance and organic ranking. Custom Product Pages let you test ad creatives by intent and potentially report it to your ASO team. Redownload campaigns consistently outperform on ARPU vs new install campaigns (we wrote about the Miniclip example recently), so it's worth exploring what's possible on the organic re-engagement side too. The platform has always rewarded teams that think holistically.

3. Manage expectations properly when reporting upward. Apple Ads doesn’t give you ROAS, and that’s a constraint everyone shares. What matters is documenting your methodology, reading properly your post-install data and communicating clearly about this constraint. And here’s something most people aren’t doing yet: you can now connect your Apple Ads and MMP APIs directly to your preferred AI platform and go deep on the analysis, without necessarily paying for an external automation tool. I’ve done it with Astro, and I’m rolling it out across my other tools. It saves hours, and the depth of insight you get back is genuinely surprising. More on that soon.

🎯 You might also like:

The honest verdict

Apple Ads is not the best-reporting, most transparent ad platform in the market.

It’s not designed to be.

It’s designed to work within Apple’s privacy architecture, move at Apple’s speed, and serve Apple’s long-term platform goals, which aren’t always perfectly aligned with yours.

But it reaches users at the moment they are literally searching for an app to download, with intent signals that no other platform can replicate. That’s still a structural advantage worth building around.

The frustration is valid, and the explanation is structural. Knowing the difference means you can build a better strategy around the constraints rather than fighting them.

🎯 You might also like:

✨ Found this useful? Share it with someone building or growing an app.

Subscribe now

This morning I opened the JP App Store and noticed LINE Manga had changed their app icon to sakura. A small detail, but that’s exactly the kind of intentional move that separates local Japanese apps from external ones. Most teams worldwide just translate their screenshots and call it localisation. But if you really want to take on this market, you’ll have to do more than that.

Experiencing sakura season in Japan is really something else. Beyond the aesthetics, being here has reminded me of something I used to tell clients every spring when I was at Apple Ads in London: Golden Week is one of the highest-value activation windows of the year. This year, it runs April 29 – May 6, which means you have roughly four weeks to prepare.

Here’s what to actually do with them.

1. Spend some time browsing the Japanese App Store

Open the App Store set to Japan and browse the Today tab, the Apps & Games tab, and your category. Screenshot what you see and start a library with inspirations.

Competitors running seasonally-themed creatives, curated editorial picks, in-app events… this is free competitor intelligence. Do it now and again closer to April 29th to track what’s shifting.

Explore the JP App Store Now

2. Don’t just translate, actually localise

Running Apple Ads ? Then start with keywords that are already working there. Those are validated intent signals. Run them through a JP localisation pass such as AppFigures’ Translation feature, which is useful for spotting which of your existing keywords are underperforming in JP compared to other locales; or use Astro MCP + Claude to surface keyword opportunities at scale.

🎯 You might also like:

3. Align visuals with what Japanese users actually respond to

Translating screenshot text is the first step, then you need to adapt your creatives too. Japanese App Store creatives tend to be character-forward, high-contrast, and emotionally expressive.

Seasonal cues land well. Think sakura, holiday family time, the feeling of having a week off, taking time to relax…

Here are some great examples I’ve found, across several categories.

4. Seasonality is perfect for In-App Events

If you haven’t submitted a Spring or Golden Week in-app event yet… well it’s time to think about it.

Apple needs review time so even a simple seasonal event or limited-time offer creates an additional surface on the App Store product page and signals editorial relevance to Apple.

5. The real ASA opportunity is before Golden Week starts

This is the one most teams miss. The window that matters most is 14 to 3 days prior to the actual peak, so roughly April 15th to 26th.

CPTs climb fast once Golden Week begins so here are my tips: get your JP campaigns structured, bids ready, budgets allocated and creatives approved now.

If you’re not active in JP Search Ads at all yet, even a broad match campaign, or a “Maximise Conversion” will compound over the holiday. Don’t wait for the week itself.

If you have extra budget, here’s something I would test:

• Search tab for re-engagement. If you have an existing JP user base, Search tab is an underused surface for bringing them back. It shows up just before users have typed anything. Pair it with a seasonal angle, a spring or Golden Week screenshot, and you’re reaching people in browse mode right when they have a week off to actually use your app.

🎯 You might also like:

Why this market, why now

The underlying context: in 2024, Japan was the world’s third-largest consumer app market at $16.5B (Adjust data: consumer app spending Japan 2024).

As tens of millions of Japanese travelers and shoppers turn to mobile ahead of or during the holiday rush, it’s expected to see install and engagement spikes across travel, food, entertainment, gaming, and e‑commerce apps.

• People finally have a full week off so entertainment and streaming sessions surge: entertainment installs were up 14% YoY in H1 2025, with video streaming averaging 55 minutes per session (Adjust × Sensor Tower Japan App Trends 2025).

• On May 1st during Golden Week 2021, player spending in Japan reached $137.7 million in a single day, the second highest day for game revenue that entire year, behind only January 1st (Sensor Tower, Top Mobile Games in Japan 2021 by Revenue & Downloads).

Golden Week is when multiple high-value categories spike simultaneously. So it’s most likely the opportunity you’ve been waiting for to unlock it.

🎯 You might also like:

✨ Found this useful? Share it with someone building or growing an app.

Subscribe now

This newsletter is the full breakdown: what it is, how to set it up, and the exact prompts I’m using now.

First, what is this actually?

Most ASO workflows look like this: Pull data from tool > copy to spreadsheet > think > go back to tool > make change.

MCP collapses that entire loop. Your AI and your data live in the same conversation.

Three things you need to know:

• Astro: my ASO keyword tracking tool for the App Store. It's lightweight, quite affordable ($108 billed annually) with unlimited keywords and apps.

  If you're not on it yet:

GET ASTRO

• Claude is Anthropic’s AI. You need a paid plan (Pro or above) for MCP to work. claude.ai/

• MCP (Model Context Protocol) is the protocol that connects them. Think of it as giving Claude a direct line to your Astro database. It reads your data, takes actions, and reports back, all inside the chat.

How to set it up

The docs here make it look simple. It mostly is but there are a few walls you might hit. Here’s the real sequence:

1. Enable MCP in Astro

• Open Astro > Settings > MCP Server.

• Toggle on Enable MCP Server.

• Note the port — default is 8089, leave it as is.

• Astro must be open every time you want to use this.

• The MCP server starts and stops with the app.

2. Make sure you have Node.js

Open Terminal (Cmd + Space > type Terminal) and run:

node --version

If you get a version number back, you’re good.

If not, download it from nodejs.org and grab the LTS version.

3. Install Claude Code

Claude Code is Anthropic’s CLI tool. It’s what connects to Astro’s MCP server.

In the terminal, run:

npm install -g @anthropic-ai/claude-code

⚠️ You might hit a permissions error here. It looks like:

EACCES: permission denied.

Don’t panic. Run these 3 commands in the terminal :

mkdir -p ~/.npm-global

npm config set prefix '~/.npm-global'

echo 'export PATH=~/.npm-global/bin:$PATH' >> ~/.zshrc && source ~/.zshrc

Then try the install again.

This is a one-time fix that solves npm permissions permanently.

4. Connect Astro to Claude Code

Open terminal and run:

claude mcp add --transport http astro http://127.0.0.1:8089/mcp

You should see: Added HTTP MCP server astro with URL: http://127.0.0.1:8089/mcp

5. Open Claude Code

Still in the terminal, type:

claude astro

You’re in. Astro is connected. Start asking.

⚠️ Google Drive / Sheets integration: I’m still working through the auth flow for writing reports directly into Google Docs via MCP. Will share a full tutorial once I have a clean, tested process. Watch this space.

What I did in my first session + the exact prompts I used

→ Filtered by popularity + difficulty to find real opportunities

Instead of manually scanning columns in Astro, I described what I was looking for in plain English. Claude reasoned through which keywords were worth testing, based on the criteria I added.

Prompt: For my app in the US, find keyword suggestions with low difficulty, relevant to my app and high popularity above 5.

→ Tagged keywords without touching Astro once

I asked Claude to apply colour-coded tags to a set of keywords based on themes, for future Custom Product Pages. It did it directly via MCP.

Prompt: For my app, look at all tracked keywords and group them into thematic clusters that could each support a dedicated Custom Product Page. For each cluster: suggest a CPP theme name (e.g. ‘crypto investing’, ‘neo bank’, ‘eSIM’), list the keywords that belong to it, show average popularity and current ranking for that cluster, and tag each keyword in Astro with the CPP theme name using a colour-coded tag. Prioritise clusters where we have both ranking traction and high popularity keywords.

→ Found non-US opportunities

FR, DE, JP.. great markets where we did not take the time to localise the metadata. I always knew international geos were an opportunity, but with US-focused clients and limited bandwidth, it always got pushed. Now I can surface the low-hanging fruit in minutes and actually act on it.

Prompts: Find high-popularity keyword suggestions for my app in FR, DE, JP in non-english-language keywords; they all should be relevant to my app. Show me the best opportunities with my current rankings, and add the most promising ones to tracking in their respective stores.

→ BONUS: Astro’s official prompt library

Astro has also published their own collection of ready-to-use prompts specifically built for this MCP.

My picks from their library to start with: keyword clustering by intent, ratings monitoring by country, and the competitor gap analysis prompt. They complement the ones above well.

The honest limitations

I don’t want to oversell this. Here’s what doesn’t work yet, and the workarounds I’ve found.

1. Astro is keyword rankings only

This is the biggest structural gap. Astro tracks keyword positions, that’s it. It doesn’t pull impressions, conversion rate, downloads, revenue, creatives, or Apple Search Ads data. So while Claude can do a lot with keyword intelligence, a full ASO report still needs data from other sources. The MCP doesn’t magically bridge that.

Workaround: I still pull conversion and impressions data manually from App Store Connect and Apple Ads. The goal is eventually to have multiple MCPs running in parallel: Astro for keywords, another for store performance data.. but that’s not plug-and-play yet.

2. It won’t suggest keywords you’re already ranking for

If you want to understand your current keyword footprint before doing research, Astro’s MCP doesn’t show you the full picture of where you’re already indexed.

My workaround: I pulled the app’s existing rankings from MobileAction, every keyword the app was ranking for in the top 100 and copy-pasted that list into Astro first, then used Claude for research on top of that baseline. It works, but it’s a manual step I shouldn’t have to do.

This is actually a feature request I’ve sent to the Astro team. A ‘current ranking import’ flow would make this setup significantly more powerful.

3. Ratings data is limited

Astro does support ratings tracking via MCP and you can ask Claude to pull ratings by store and country.

But there’s a catch: historical data only starts from the day you add the app to Astro. If you added an app last week, you have one week of ratings history. Nothing before that.

For new apps or recently added apps this is fine. For established apps where ratings trends matter, it’s a real gap. You’ll need App Store Connect or a tool like AppFollow for proper historical ratings analysis.

What’s next: ASO weekly reports, automated

The end goal is actually automating my weekly ASO reporting workflow entirely. What that looks like in practice:

• Claude pulls weekly keyword ranking changes from Astro

• Flags movements worth flagging (drops > 5 positions, new top 10 entries, competitor overtakes)

• Writes the narrative analysisL what moved, why it likely moved, what to do about it

• Send it to myself or clients via Slack or email

• Outputs directly into a Google Sheet

I’m not there yet as the Google Drive/Sheets MCP auth is still something I’m working through. Before connecting anything, I want to properly review the privacy implications and limit what data Claude actually has access to. That's the part I'm most careful about: I don't want any private client information ending up somewhere it shouldn't.

🎯 You might also like:

✨ Found this useful? Share it with someone building or growing an app.

Subscribe now

Apple rolled out the 2nd ad placement on Search Results globally. Simultaneously, the company dropped a significant Insights & Custom Reports update. New metrics, new reports, native data that we used to stitch together from external tools.

Two launches, one day.

I want to break down exactly what’s new, what it actually means for your workflow, and, more importantly, what you should do with it this week.

Let’s get into it.

🎯 The Report We Asked For

Daily & Weekly SOV: First Ad Position

This is the metric marketers have been requesting since the news has been out, and Apple has delivered it.

During the test in the UK & Japan, impression share data lumped all ad positions together. With the 2nd placement rolling out globally, that was becoming a real problem, because SOV from position 1 and SOV from position 2 are not the same thing strategically.

• First position = top of Search results page. Maximum intent, maximum visibility, premium CPT territory.

• Second position = high reach, lower intent, different audience behaviour.

An aggregated SOV number that blends both tells you very little. A 60% SOV could mean you dominate position 1 on your brand terms. Or it could mean you're getting broad reach from the 2nd placement with almost no presence at the top. For Brand Protection, this could be a real issue.

Now you can tell them apart.

What to do with it

1. Run the Daily SOV (First Ad Position) report and look at your branded keywords. If your first-position SOV on your own brand terms is below 80%, you have a brand protection problem, and a competitor is appearing above you on your own name. Depending on your strategy, it might be ok or not.

2. Compare First Position SOV vs. All Positions SOV side by side. The gap between the two numbers tells you how much of your visibility is coming from the 2nd placement. A large gap isn't necessarily bad, but it should be a conscious strategic choice.

3. Monitor this daily during any competitor campaign activity. First Position SOV is now your real-time brand defence signal.

🎯 The One We Did Not Know We Needed

Monthly Search Term Rank Report

This one is genuinely new territory. And I think it's going to change how a lot of teams approach keyword strategy. Here's what you get:

• Search popularity score for any keyword, from 1 to 100 (overall, across all categories)

• Search popularity score within your specific genre/category

• The keyword rank within that genre

• All of it native, monthly, free

To appreciate why this matters, you need to remember what we had before: a blunt 1–5 popularity band inside Apple Ads, with no category context.

A score of "4" could mean extremely popular within Productivity apps, or moderately popular across all of iOS. So you either paid for an external tool to get proper keyword intelligence, or you were essentially guessing.

That just changed.

What to do with it

1. Look into your category’s search landscape: Filter by your genre, sort by Rank in Genre ascending. You now have a priority-ordered list of the highest-demand search terms in your category. Start at rank 1 and ask: am I bidding on this? Am I competitive? Do I even have this in my campaigns? Work your way down to find gaps.

2. Spot cross-category opportunity terms. Compare a keyword’s overall popularity score vs. its genre score. If a term scores 85 overall but only 60 in your genre, it’s pulling search volume from outside your category, potentially underpriced because your direct competitors aren’t bidding too much on it. This might be a great keyword opportunity.

3. Validate your strategy with data and use report to check your existing keyword lists. Something you deprioritised scores 70 in your genre and you've been leaving volume on the table. Cross this data with your SOV, and you can see if there’s any additional volume you could catch.

⚠️ Important Caveat ⚠️

Before you go all-in on this data, something critical to flag, as spotted by Maxim Melnik.

The Monthly Search Term Rank Report currently doesn't show search popularity for terms scoring below 44. This isn't new, it's been an ongoing issue since October 1st, when Apple effectively disabled Search Ads Popularity (SAP) data collection following changes to how this data was surfaced.

What this means in practice: the report gives you excellent visibility on high-volume, high-popularity keywords. But medium and low-frequency search terms, which are often where great long-tail opportunities live, are currently sitting in a blind spot.

Third-party tools like AppTweak have built prediction models to estimate popularity for these terms based on historical data, but by their own admission, accuracy has taken a hit since the October change. It's directionally useful, not definitive.

So the honest picture is this: the Monthly Search Term Rank Report is a genuine step forward, and for high-popularity keyword strategy it's immediately valuable. For long-tail work, you still need external tools, and even those are working with imperfect data right now.

📉 The Nice Insights Update

Embedded data visualisation in the dashboard

Last but not least, charts are now live directly inside the reporting interface. Spend, installs, CPA, impressions, all visualised natively without exporting to Sheets (!!!).

I’ll be honest: I remember asking for this feature when I was working at Apple few years ago. So yes, it was about time.

But I’m genuinely happy it’s here. The time saved adds up fast: no more exporting and uploading to an external tool that I’m also paying for, just to see a trend line I could have seen in 30 seconds. In practice, it’s the kind of friction removal that makes you actually check your data more often, which is always a good thing.

What to do with it

1. Start with an account overview chat. Build top-level reports covering Spend, Installs, CPI, TTR and CR, broken down by geo and/or campaign. This becomes your morning dashboard.

2. Go granular on your top campaigns and take your highest-spend or highest-priority campaigns and drill down: by ad group, by CPP, by keyword. The chart view makes it much easier to spot which layer is underperforming.

3. Build a focused report for your current priority. Working on a specific goal, geo, or campaign this quarter? Create a dedicated report for it and keep it as your reference point. Instead of rebuilding filters every time you check in, you have a living view of exactly what you’re optimising for.

🎯 You might also like:

✅ Action List for this Week

1. Set up Daily SOV (First Ad Position) report for your Brand campaigns.

2. Run First Position SOV vs. All Positions SOV and compare them side by side. The goal is to understand where your visibility is actually coming from now that 2nd placement is live.

3. Pull the Monthly Search Term Rank Report filtered to your genre and build your keyword gap list.

4. Cross-reference your current keyword list against genre rank to find what you’re missing.

That’s it for this week. A lot landed recently so I’ll keep watching how the 2nd placement rollout affects SOV data in practice and report back as the data comes in.

See you next Friday,

Julie 🫰

✨ Found this useful? Share it with someone building or growing an app.

Subscribe now

Hard paywall vs soft paywall

Let’s get into it right away. Something really stuck with me, and it’s all about paywalls.

Freemium apps convert a median 2.1% of downloads into paying users within 35 days. Hard paywall apps convert 10.7% of downloads.

That’s a 5× gap in how many downloads actually turn into revenue.

Look at the spread too. The freemium box plot is tight and low, which means you’re unlikely to find a freemium outlier that performs like a hard paywall app’s median.

The hard paywall distribution is wide, which means the variance is real, but so is the ceiling. The best hard paywall apps are converting nearly 4 in 10 downloads into paying users within a month. The best freemium apps top out at 8.2%.

The users who push through a hard paywall are pre-qualified. You’re trading volume for quality, and the LTV math strongly favours quality.

This chart should be in every monetisation strategy, because the freemium “get everyone in” logic has a cost: you’re paying to acquire, retain, and re-engage users who were never going to pay.

A hard paywall filters, keeping only users who convert despite the friction as they usually pick annual plans and churn less.

At the same time, Adapty released their own report “State of In-App Subscriptions 2026”, which added a second dimension: hard paywall users generate 21% higher 1-year LTV. Higher conversion rate AND higher revenue per user.

Edit: flagged something important in the comments. Adapty’s 2026 report actually shows the opposite, with soft paywalls that outconvert hard ones by nearly 50% (4.85% vs 3.34%, paywall view to payment).

This directly contradicts what RevenueCat highlights above.

Here's why they might diverge: RevenueCat's data covers 115k+ apps vs Adapty's 16k+, which means broader representation. Also, depending on which app categories are most represented, the data will be biased too. But more importantly, they're measuring different funnel moments. RevenueCat tracks download-to-paid over 35 days. Adapty tracks paywall view-to-payment.

What will most likely work for your app? Well you won’t know for sure unless you keep testing, testing, testing & you have your own data.

The Biggest Missed Opportunity

The other chart that stood out is the offer usage by category. Across all categories, only 9.3% of apps run any promotional offers. Health & Fitness leads at 14%, still barely a tenth of the category. Gaming sits at 3.6%, the lowest of all.

Vahe Baghdasaryan, founder of Tangent Growth, breaks promotions into four types in the newly launched startapp.school, in ascending order of sophistication:

• Extended free trial (7-day to 14-day, no discount) to use when trust is the primary blocker and users need more time to feel value.

• Trial + discount combo, which outperforms giving either alone. The trial removes risk, the discount removes price friction. Both together = maximum conversion.

• Intro offer (deep discount first period only), which works for acquisition but requires careful LTV modeling as you’re buying the relationship at a loss and betting on retention.

• Win-back offer, the most underused of all. Hit churned users within 10–15 minutes or at next session after cancellation, that’s when intent is still live.

And remember: when someone cancels, it does not necessarily mean “I’m not interested”. It could also mean “not right now” or “not at this commitment level”.

Localising your paywall

One of the sharpest practitioner finding in RevenueCat report came from the Mojo team. They tested paywall layouts by geography, not just pricing.

“At Mojo, we’ve seen different designs outperform depending on the market.” - Michal Parizek, Senior Growth PM

• +20% conversions in Japan vs US, by adding a long-scroll paywall with social proof & Free vs Pro comparison.

• +30% trial start rate and +10% yearly take rate Latin American from re-anchoring annual plan to monthly equivalent, with zero price change.

Most teams localise currency and stop there. The Mojo data is giving context and stakes: design, layout, information hierarchy, social proof format, and pricing presentation are all geo-specific levers almost nobody is pulling.

🎯 You might also like:

Looking beyond the paywall

Here’s the thing most teams get wrong: they look at paywall conversion in isolation. But every number on your paywall is the result of everything that happened before it, from your ads to your App Store page and your onboarding.

The full funnel looks like this:

Ad creative → CPP / Product Page → Onboarding → Paywall → Trial → Paid

Where the drop off happens is an indication on what to analyze and optimize.

Drop-off Ad → Install (low App Store CVR): Message mismatch between ad and product page. Your ad made a promise; your product page is telling a different story. You can fix it by building CPPs aligned to each creative angle.

Drop-off Install → Onboarding completion (high D0 churn): Expectation gap. They installed expecting something, but the app opened with something else. Fix it by auditing your top 3 performing creatives + your product page and check whether your onboarding’s first screen reflects those exact promises.

Drop-off Onboarding → Trial start (low trial conversion): Timing or trust failure. Was motivation high at this exact moment? What friction did you introduce? Did you earn trust quickly enough? And does your paywall headline continue the narrative from onboarding, or does it start a new one?

Drop-off Trial → Paid (low trial-to-paid): This is most likely a product problem, not a paywall problem. The paywall already did its job since the user started a trial. They tried the product and weren’t convinced. Redesigning your paywall won’t fix this, so now it’s time to look at activation: did users hit your core value moment during the trial period?

The copywriting test: pull your best ad headline, your CPP headline, your onboarding first screen, and your paywall CTA. Read them in sequence.

Do they tell one coherent story? Most apps fail this test badly.

5 things to check ASAP

1. Audit your paywall UI against the benchmark

RevenueCat Paywall UI Element Adoption by category is a good way to check if you’re missing anything on your paywall. The data highlighted pricing ~75% adoption (expected), testimonials 6–17%. Cancel assurance, feature list, free trial messaging… check all. Missing 3+? Fix the basics before running any A/B tests.

2. Check your funnel for message disconnects

Pull your best-performing ad, your CPP, your onboarding first screen, your paywall headline. Can you draw a straight line between them? If not, you might have a trust failure somewhere in the funnel and it’s costing you more than any paywall tweak can fix. Discuss with your UA & ASO teams and align the messaging to make sure it’s consistent.

3. Break your paywall metrics by geo

What’s your conversion rate in your top 5 markets individually?

The variance will surprise you. For underperforming markets: is it price sensitivity (try monthly-equivalent anchoring for annual) or a layout/trust issue (try more social proof)?

Each country need different treatments. Ask UA teams what’s their strategy and knowledge about different markets, and make sure you also go granular there.

4. Turn on a win-back offer if you have zero promotions running

You’re almost certainly in the 90.7%. Start with a win-back: trigger a 30–50% annual discount at next session after trial cancellation. Lowest risk (the user already left), clearest intent signal, fastest way to see if offers work for your audience.

5. Define what you’re actually optimising for before the next experiment

Trial start rate, paid conversion within 7 days, or ARPU? These require different paywall designs. Pick one primary objective, set guardrail metrics on the others. Without this, you’ll debate experiment results forever and the paywall will keep changing for the wrong reasons.

The Real Conclusion

The best paywall isn’t hard or soft. It’s the one your specific user, in their specific market, at their specific moment of motivation, is most likely to walk through, because the friction is calibrated, the timing is right, and trust has been earned from the very first ad they saw.

Most teams are running one global paywall, optimising button colours, using zero promotional offers, and not speaking to their UA or ASO colleagues.

The data from this week shows it clearly: the opportunity is enormous, the playbook is clearer than it’s ever been, and the market is still maturing there.

✨ Found this useful? Share it with someone building or growing an app.

Subscribe now

What Happened

Anthropic had signed a $200M contract with the Pentagon in July. The terms they’d negotiated included two hard limits: no use of Claude for fully autonomous weapons, and no mass domestic surveillance of American citizens. It was non-negotiable.

The Department of Defense refused to budge, as they wanted AI available for “all lawful purposes”, and told Anthropic to drop its restrictions or lose the contract.

Dario Amodei, CEO of Anthropic, made his position clear: “These threats do not change our position. We cannot in good conscience accede to their request”.

Trump’s response was quick, and ordered every federal agency to immediately cease all use of Anthropic’s technology. Defense Secretary Hegseth went even further, designating Anthropic a “Supply-Chain Risk to National Security”, a label normally reserved for foreign adversaries.

Then came the move that changed everything.

Hours after Anthropic’s blacklisting, OpenAI announced it had struck a deal with the Department of Defense to deploy its models on classified networks. Medium Sam Altman moved fast... maybe too fast. By Monday, he publicly admitted OpenAI “shouldn’t have rushed” its deal, calling it something that “looked opportunistic and sloppy”.

But users noticed, and they responded in the only way that actually hurts a tech company: with their behaviour.

The Charts Don’t Lie

Before the government standoff even happened, Claude was already having its best month ever.

Their Super Bowl ad in early February drove 170K downloads in a single day, already well above their 62K daily average. By February 18th, they’d hit 225K, an all-time high, per Appfigures. The trajectory was already pointing up.

Then the standoff hit. Claude’s downloads surged 37% day-over-day on February 27th, then a further 51% on Saturday. Claude climbed to #1 on the U.S. App Store, rising over 20 ranks in under a week. For the first time ever, daily Claude downloads surpassed ChatGPT’s.

Meanwhile the opposite was happening at OpenAI. U.S., as uninstalls of the ChatGPT app jumped 295% in a single day. One-star reviews surged 775%, and LinkedIn became a real-time feed of people announcing their switches.

This is what values-driven churn looks like at scale: it’s people who were already slightly dissatisfied, handed a reason to finally act.

The QuitGPT campaign formalised it too, with about 1.5 million sign-ups at launch, reaching 2.5 million by March 3rd, with physical protests outside OpenAI’s San Francisco HQ.

, one of the most widely-read AI writers on Substack, published “Why I Deleted ChatGPT After Three Years”. The cultural signal was as loud as the data signal.

Claude briefly crashed from the surge, with the company’s servers buckling under the weight of a values moment.

This Is Not a Coincidence

Let me connect this back to what I wrote few weeks back.

The apps winning right now, Proton, Bluesky, Vinted, Back Market, StoryGraph, are winning because users are making active choices about who they trust, what they consume, and who they support.

Anthropic drew a real line, with real financial consequences, and users rewarded them for it in the most direct way possible: downloads.

Anthropic updated its competitor-switching feature, making it easier to import preferences and context from other LLMs, and directly capturing the wave of users arriving from ChatGPT. That’s smart, reactive product growth.

🎯 You might also like:

The Uncomfortable Nuance

I want to be honest here, because this newsletter is not a PR piece for Anthropic.

Anthropic is STILL a $35B company backed by Google and Amazon, two of the largest data collection operations on earth. The “values vs. Big Tech” framing gets complicated when the cap table is Big Tech.

Anthropic’s own investors didn’t publicly back them during the standoff. In a meeting between Amazon CEO Andy Jassy and Defense Secretary Hegseth, Jassy reportedly declined to take Anthropic’s side.

So what do we make of this? Anthropic isn’t a scrappy indie. But in this specific moment, on this specific issue, they held a line their competitors didn’t. And in the absence of meaningful legislation, Americans are largely left relying on a tech company CEO to protect privacy, which is exactly the kind of fragile, unsustainable situation that makes you realise how badly proper legal frameworks are needed.

This doesn’t make Anthropic perfect, but it makes them a better alternative right now. And at the moment, that’s what users respond to.

What This Means for Growth

Here’s the part I can’t stop thinking about.

Anthropic hit #1 on the App Store with values decision, made under pressure, with significant financial cost. It became the most effective growth lever in the company’s consumer history.

This is the logical endpoint of everything I wrote few weeks ago. The funnel started with trust, with a public stance that users could see, verify, and act on.

For growth teams, the lesson is uncomfortable: you cannot manufacture this. You can only build toward it, by being genuinely aligned with your users’ values long before a moment like this arrives, so that when it does, your App Store page, your community, and your brand story are already telling the right story.

Anthropic’s App Store page didn’t change last Friday.

Their ranking did, because everything they’d built before that moment made the choice obvious for users looking for an exit from OpenAI.

That’s the new growth strategy: being more legible about what you stand for.

One More Thing

I predicted last week that this category was accelerating. That values-aligned apps were winning ground. That the App Store charts were becoming a mirror of something much larger than product preferences.

I didn’t expect the proof of concept to arrive within seven days. But here it is, in the data, in the charts, and in 2.5 million people who decided their AI assistant was now a political choice.

✨ Found this useful? Share it with someone building or growing an app.

Subscribe now

A few months and a lot of data later, Apple is now rolling it out to all accounts. And the timing couldn’t be better… or more deliberate.

In March 2026, Apple Ads is also introducing additional Search Results placements. Ads will now appear further down in search results, not just at the top. More inventory, and likely more affordable inventory since lower positions will be less contested. No action needed since existing Search Results campaigns automatically qualify for all positions.

This changes the Maximize Conversions feature significantly: more surface area for the algorithm to explore, potentially lower CPTs on new placements, a broader learning pool from day one. Apple is expanding the auction at the exact moment it’s handing advertisers an AI tool to navigate it.

Regarding third-party tools like SplitMetrics, MobileAction, or AppTweak, they aren’t going anywhere. What Apple is launching is a basic CPA automation tool, on a new type of campaign. These platforms offer rule-based automation across all campaign types, MMP-connected post-install event bidding, and far more granular control.

So here’s everything I’ve learned over the past few months, what works, what kills campaigns, and how to set it up properly from day one.

🎯 You might also like:

What it actually is (and isn’t)

Maximize Conversions is a discovery and learning tool first. Apple has finally joined every other major ad platform (Google, Meta, Unity) in offering native AI bidding, but let’s be honest… it’s years behind.

Right now it optimizes purely for installs, which means no trials, no purchases, no retention signals. Keep that scope in mind before treating it as a magic solution.

The core mechanic shift: CPA cap was a hard ceiling per keyword, as the bid never exceeded your cap, no matter how valuable the query, which meant missed opportunities on your potential best terms.

Target CPA is an average across the campaign. The algorithm bids €8 on a term worth €8 and €2 on a term worth €2, aiming to average your goal. More opportunity and less granular control.

The additional ad group: what you can actually control

When you create a Maximize Conversions campaign, Apple automatically generates a locked Automated Ad Group. Leave it completely untouched since that’s your exploration engine and it needs maximum reach.

You also have the option to create an additional manual ad group on top.

Here’s what you can and can’t do in it, because this is frequently misunderstood:

• You can: change audience settings (new vs. returning users), use a Custom Product Page, select specific keywords to guide the auto-bidder, pause specific keywords you don’t want to run, add negative keywords.

• You cannot: set or manage individual keyword bids, as there are no CPT bids in a Maximize Conversions campaign. If you need per-keyword bid control, you need a separate Manage Bids campaign (classic one).

The practical implication: when you add keywords to the additional ad group, you’re not bidding on them manually but you’re telling the algorithm “use these as anchors to find related queries”. The auto-bidder still sets the actual CPT, so it means this optional ad group is useful for keyword scoping and CPP segmentation, not for bid optimization.

My recommendation: use the additional ad group for Custom Product Page testing on your highest-intent audience (new users only) + intent alignement. Don’t forget to add negative keywords at campaign level to prevent overlap with your existing Exact match campaigns.

Calculating budget & CPA

If your campaign don’t get enough data for the algorithm to learn, your campaigns will most likely fail. So here’s our to calculate your minimum & recommended daily budget:

Apple explicitly states the campaign needs at least 5 conversions per day to function. That’s the floor, so below 5 conversions/day, the model does not have enough room to learn.

Honest take: if you can’t hit the recommended budget (10× CPA/day), you’re not necessarily blocked from running it, but you should expect a longer, messier learning phase and less reliable performance. The model stabilizes proportionally faster with more daily conversions. A campaign getting 50 installs/day learns in 2 weeks what a campaign getting 5 installs/day takes 10 weeks to learn.

Also important: Apple says daily costs may be above or below target on a given day, and weekly/monthly averaging is how performance should be read. Don’t react to single-day CPA spikes in the first two weeks, that’s the learning phase.

Recommendations

• Set your Target CPA at or slightly above your blended CPA from existing campaigns. Starting too aggressive means the algorithm can’t find inventory and delivery will be low. You can tighten CPA gradually once the model stabilizes, but not before.

• Add campaign-level negative keywords before you launch. Any terms already running as Exact in your other campaigns must be excluded here. Otherwise you’re competing against yourself in the same auction and inflating your own CPTs. This is probably the most overlooked setup step.

• Leave the Automated Ad Group completely alone. Don’t restrict the audience, don’t add keywords, don’t touch it. Every restriction you add slows learning.

• If you create the additional ad group, use it for CPP testing. If you have Custom Product Pages, this is where they belong, paired with a specific audience segment (new users, a specific theme / intent) where you want cleaner measurement. This is probably the highest-leverage use of that optional layer.

• Wait a full 2 weeks before drawing conclusions, and Week 3 before adjusting anything. Changing Target CPA during the learning phase resets the model.

• Mine your search terms report actively from Week 3 onward. The real long-term value of this campaign type is discovering high-intent query variations you’d never have bid on manually. Export them, validate performance, and promote the winners into dedicated Exact match campaigns with manual bidding. The lifecycle: Auto > Learn > Monitor > Move to Exact > Scale.

• Track post-install quality from day one. The limit here is that the algorithm only optimizes for installs. If you don’t have your MMP properly connected and you’re not looking at D1/D7 ROAS, retention, trial rates, or purchase rates by source, search term or even ad group, you have no idea if the installs are valuable. Cheap installs from irrelevant queries will look fine at the CPI level but negatively impact your ROAS.

What to avoid

• Don’t set a Target CPA below what’s actually achievable in your category. The campaign will most likely underdeliver. If delivery is weak after 2 weeks, raise the Target CPA before assuming the strategy doesn’t work.

• Don’t run it without sufficient budget. An under-budgeted Maximize Conversions campaign enters a spiral: low spend > few conversions > weak signal > worse bidding > low CTR... Manual bidding will outperform it in this scenario every time.

• Don’t look at keyword-level CPA within the campaign and panic. Individual keywords will frequently appear above Target CPA. That’s by design as the system averages across the campaign. If campaign-level CPA is on target, keyword-level inflation is noise. Evaluating at the wrong level leads to bad decisions. What you should do is always look at post-install data.

• Don’t expect it to work for pre-orders. Apple is explicit: post-launch only. No conversion history = no learning = no performance.

• Don’t use the additional ad group to try to replicate manual bidding. If you need per-keyword CPT control, you need a Manage Bids campaign running separately. Maximize Conversions and bid management are two distinct campaign types so don’t conflate them.

• Don’t treat Week 1 volatility as failure. Higher CPAs, unexpected queries, aggressive spend, strange tap-through rates… this is just the model calibrating.

Account structure that makes sense

Run Maximize Conversions as your discovery layer alongside (not instead of) your existing campaigns:

• Maximize Conversions campaign: discovery & keyword mining. Automated Ad Group untouched, campaign-level negatives in place, budget above 10× Target CPA. Optional additional ad group for CPP only.

• Exact match campaigns: proven terms at controlled CPTs. This is where you run BAUs campaigns and graduate the winners from Maximize Conversions.

• Brand campaign: if you decide to protect your Brand at all cost, keep always separate, always Exact.

The mistake is running Maximize Conversions as your only campaign. It’s not a replacement for structured Exact campaigns, but more like a feeder for them.

Bottom line

Maximize Conversions is genuinely useful if you go in knowing what it is: a keyword discovery machine that learns through spending.

The teams that will extract the most value are those who treat it as the top of a funnel that leads to structured Exact campaigns.

The teams that will waste budget are those who expect immediate CPA efficiency and start touching things before the model has learned anything.

I’m curious to hear how you’ll test this new feature: any thoughts or questions? Leave a comment!

🎯 In case your missed it:

Client FAQ — Maximize Conversions

1. “We already run ASA manually, why change?” This runs alongside your Exact campaigns as a discovery layer and finds converting search terms you’d never have bid on manually.

2. “What budget do we actually need?” Target CPA × 10 per day, minimum. Apple needs 5 conversions/day for the model to learn, below that you’re wasting budget. €5 CPA = €50/day.

3. “Will CPA spike at launch?” Yes. First 7–10 days will most likely look bad, with higher CPAs, unexpected queries, messy spend. This is the learning phase so don’t touch anything for 2 weeks and judge by Week 3.

4. “Can we control which keywords it runs on?” Partially as you can add keywords in the optional second ad group. You’re telling the algorithm “look for queries like these”, but keep in mind that no CPT bids exist in this campaign type. If you want bid control, that’s a separate Manage Bids campaign. What you can do: add negative keywords to prevent overlap with your existing Exact campaigns.

5. “How do we know if it’s working?” Evaluate at campaign level only, as individual keywords will regularly appear above Target CPA and that’s by design. If campaign-level CPA is trending toward target by Week 3, it’s working. Then look at your MMP: the algorithm only optimizes for installs, so post-install quality is entirely on you.

6. “What about our CPA cap campaigns?” CPA cap is being deprecated (finally..) migration is forced. The shift: CPA cap was a hard ceiling per keyword. Target CPA is an average across the campaign. Individual bids will flex higher on valuable queries, lower elsewhere.

7. “When should we NOT run this?” Budget too small to hit 5 conversions/day. Pre-launch. Teams who can’t resist optimizing every week.

✨ Want to stay ahead in Ad Tech for Mobile Apps & Games?

Subscribe now & get the latest trends delivered straight to your inbox!

Subscribe now

It made me start asking bigger questions: Can we actually grow out of Big Tech? Is digital sovereignty even possible for a normal person? How do we reclaim something we gave away so gradually we didn’t notice it was gone?

Because here’s the thing: I’m not just observing this from the outside. As someone who’s always felt a bit like an outsider myself, I’ve always been keen on using tools that are better for the planet, for the users, for the culture. I use Bandcamp and gave hundreds to support indie artists. I’m also on Bluesky, as I left Twitter and everything I built for years when it got acquired. I only buy secondhand, either thrifting or using apps. I test smaller apps, support independent artists, pay for things that feel worth paying for.

And somewhere between being a user who cares and a growth consultant who thinks about the App Store for a living, I started wondering: what if being indie, being small, having actual values, is a growth strategy? Turns out it is. So here we are.

Update - March 1st, 2026

Since publishing this piece, the topic got even more interesting, and faster than I expected.

Anthropic refused to allow the U.S. government to use Claude for mass surveillance and autonomous weapons. Trump responded by banning them from all federal contracts. OpenAI immediately opened their arms to fill the gap.

Users noticed and Anthropic hit #1 on the App Store charts, in front of OpenAI, for the first time ever. As I write this, over 1.5 million people signed a petition to cancel their ChatGPT subscriptions at quitgpt.org.

Everything I wrote below? It just happened.

I’m covering this in full in the next issue.

Let’s start with an interesting chart that came across my feed few weeks ago: 60% increase year-over-year in iOS app releases as of December 2025.

All this linked to AI and the rise of agentic codic. Now everyone can build and launch an app, for a relatively low fee. Then agentic coding tools dropped in early 2025 with Cursor, Claude Code, Replit Agent, Bolt.. and the App Store basically broke its own ceiling.

I briefly covered it in my 2026 Predictions, and now it’s time to look into how this shift is affecting growth & product strategies.

The big discomfort nobody can ignore

There’s also a slow-burning anxiety building among app users.

I covered part of this last month when TikTok’s U.S. restructuring triggered an unexpected reshuffling of the Social Networking charts. Users didn’t necessarily delete TikTok overnight, but they paused and looked around. In that window of attention, apps like Skylight and Upscrolled surged, because they were present and credible when trust cracked.

🎯 You might also like:

The triggers are stacking:

• EU–U.S. data transfer tensions, and sovereignty is becoming a real question. It’s not just about governments and enterprises, but about everyday users who are starting to ask where their data actually lives, and under whose laws.

• The creeping realization that most of our digital infrastructure runs through 5 American companies aka the Big Tech. The dependency was always there, but what's changed is that people are now aware of it.

• A post-ATT world where Apple itself normalized the question “who owns your attention?”. When the most powerful platform in mobile forces a consent prompt on every app, it changes the conversation and users started asking questions they never thought to ask before.

People are actively looking for alternatives, and for the first time in a decade, those alternatives are genuinely good.

The Alternatives are winning ground

The indie and ethical apps winning right now aren’t worse than their Big Tech counterparts. In several cases they’re similar or even better, built by people who cared about the problem. Here are the ones worth knowing.

Productivity & Utilities

Proton Mail — The Swiss Fortress

The end-to-end encrypted alternative to Gmail that turned privacy from niche paranoia into a mainstream feature. Proton passed 100 million accounts in 2024. The Swiss jurisdiction angle, your data lives outside U.S. and EU law enforcement reach, has gone from a footnote to a selling point. The product is competitive, not just principled.

Le Chat by Mistral — The European Answer to ChatGPT

Probably the most strategically important non-U.S. AI consumer product right now. Mistral raised at a €6B valuation in 2024. Le Chat runs on models that compete at the frontier, and it carries the rare credibility of being European-built, European-funded, and European-regulated.

In a post-GDPR world, that’s a structural advantage. For French users especially, there’s a pride dimension driving organic growth.

The Music Stack: a full counter-ecosystem

The music industry has quietly assembled the most complete set of alternatives of any vertical.

Bandcamp remains the gold standard for artist-first distribution, where artists keep around 80–85% of revenue. In here, the model is the product and drives growth, as artists happily share links to the platform.

Deezer is making a genuinely contrarian bet: leaning into human-curated, AI-free music as a positioning story. In a world where Spotify is known for leaning heavily into AI, “a human who loves music picked this” is becoming a real differentiator.

SoundCloud is THE platform for DJs, producers, and electronic music communities. With 40 million creator-uploaded tracks, it remains the most democratic music platform ever built. No streaming service has replicated what it does, because what it does was always about sharing culture.

Discogs is the analog internet. A marketplace and database for vinyl, CDs, and cassettes with 15 million+ releases catalogued and a community that treats music as something you possess, not stream. In a world where your Spotify library can vanish with a licensing change, Discogs represents the place where you get physical media, permanent ownership and zero algorithmic intermediary between you and the record.

Second-hand is here to stay

The secondhand economy has quietly become one of the most values-driven categories in the App Store, as well as one of the fastest growing.

Vinted crossed 100 million users in 2024 across 20+ European markets, while Depop became the default wardrobe marketplace for Gen Z. Together they represent a direct rejection of the fast fashion model: one secondhand purchase saves roughly 3kg of CO₂ compared to buying new. That’s not marketing copy anymore, it’s a purchase motivation.

Back Market is arguably the most quietly radical app on this list. The entire premise is a direct challenge to the consumer electronics industry’s core business model, always wanting the latest device. Founded in Paris in 2014, raised at a $5.7B valuation in 2022. The fact that it’s French-built and European-first makes it particularly resonant in the sovereignty narrative: here’s a company that challenged Apple’s and Samsung’s upgrade pressure, from France.

The ownership thread running through this section, own your vinyl, own your hardware, own your wardrobe history, is the physical-world version of the same instinct driving LeChat and Proton. Sovereignty isn’t not only a data question, but also a consumption one.

What Agentic Coding actually changes

For 10 years, the moat for Big Tech wasn’t just money or data, but engineering velocity. A two-person team couldn’t match a 200-person product org on shipping cadence. Agentic coding tools are collapsing that gap in real time: a solo developer with a clear product vision can now ship, iterate, and maintain an iOS app at a speed that would have required a team of 5–8 engineers two years ago.

The 60% surge in App Store releases is the first data signal of what happens when the production cost of software drops by an order of magnitude. The question is which alternatives will have the product-market fit, the community, and the trust to stick around.

What this means for ASO and Growth

And this is where I want to be direct, because most growth teams are misreading the implications. More apps means more noise. In a noisier store, the apps that win are the ones with earned trust and clear values, not the biggest install budgets.

The pattern I keep seeing: ethical alternatives win on conversion.

Users arrive already half-convinced as they’ve probably read a Reddit thread, seen a post on Bluesky, heard about the app from someone who actually cares about privacy. Their intent is high and the App Store page is the last filter. And right now, let’s be honest, most indie apps are failing that filter.

• Metadata that leads with values, not features: title and subtitle should answer “why does this exist and why should I trust it”. Proton doesn’t say “Email App” It says “Encrypted Email & Calendar”.

• Screenshot 1 is a trust signal: When users arrive from a privacy-concern context like a tweet, a news article or a moment of institutional distrust, the first screenshot needs to immediately confirm they’re in the right place.

• Keywords are getting longer-tail and more values-driven: Search queries like “private email no ads”, “open source social”, “ethical music app” are growing. Low volume but extremely high intent.

• Apple Editorial is supporting values-aligned apps: Apple is actively curating a narrative around the App Store as a place for thoughtful, user-respecting software. StoryGraph, Too Good To Go, Proton… these apps are often featured. So if your app is genuinely ethical, your store page should reflect this for both users and Apple’s editors. Also when you pitch it for Featuring, always make sure to highlight these elements.

• Community is your retention layer: The apps winning here, from Bluesky, Bandcamp, StoryGraph to Discogs, all have communities that actively evangelize. That organic word-of-mouth is bringing growth and trust, and that’s something you can’t buy. You build it by being genuinely aligned with your users’ values from day one.

The growth takeaway: fewer, more committed users beat higher volume of disengaged ones, especially in a trust-driven, privacy-aware acquisition environment.

🎯 You might also like:

Takeaway

The window is open right now. Trust is fragile, agentic tools just collapsed the cost of building, and Apple is actively featuring apps that tell a better story.

The brands and growth teams that move now and that build values into their metadata, their screenshots, their community before it becomes table stakes, will own positions that will be almost impossible to displace in 18 months.

The ones waiting for volume to come back are optimizing for a world that’s already gone. So instead of focusing on growth, volume, and acquisition, just look back and decide what values are you deciding to align with. This could be your biggest growth move.

✨ Found this useful? Share it with someone building or growing an app.

Subscribe now

For a long time, gamification was a feature.

It showed up as streaks, badges, points, progress bars… small UX layers added to increase engagement, retention, or habit formation. Something you sprinkled on top of a “real” product.

That era is officially over.

What we’re seeing now is gamification as a product strategy. Entire apps, and even full segments, are being built around game logic, as the core reason for users to show up every day.

Apps are no longer borrowing from games, they are becoming games in their own right.

🎯 Read my 2026 Predictions:

From features to actual products

The first wave of gamification was functional and was mostly about behavior control: come back tomorrow, don’t break the streak, complete the task.

But once every app started doing the same thing, it stopped working as a differentiator. Streaks became invisible, badges lost meaning and engagement flattened.

That’s when product teams pushed further. Instead of asking “How do we make this task more engaging?”, they started asking something much more radical:

“What if the task itself was play?”

That question changed product strategy entirely.

Apps started shipping real games

Duolingo is the most obvious example, as it has always been gamified with XP, leagues, leaderboards, streaks. But its recent move into chess marks a new phase.

Chess isn’t a learning mechanic layered onto language education. It’s a standalone game, embedded inside a learning app. It extends session length, increases emotional engagement, and creates a reason to open the app that has nothing to do with language learning.

Duolingo isn’t just teaching anymore, it’s slowly becoming a playground for the brain.

What’s particularly interesting is how they shipped it.

Instead of launching a standalone Chess app, Duolingo integrated chess directly into the existing product. On the App Store side, they differentiated using Custom Product Pages, with screenshots showcasing the chess experience served specifically to users searching for “chess”.

In other words, the product page adapts to user intent. Someone searching for languages sees one story. Someone searching for chess sees another.

From an ASO perspective, it’s a smart move: Duolingo leverages its existing brand, distribution, and install velocity, while testing a new game-driven use case without starting from zero.

That said, it’s a debatable choice.

Launching a separate app would arguably make more sense from a positioning and product clarity standpoint. But it would also mean rebuilding acquisition, rankings, and trust from scratch.

When games deserve their own app

NYT Crossplay follows a different but equally telling strategy.

While a Crossword game is available within NYT Games app, the team still decided to launch Crossplay as a standalone app, with its own identity, positioning, and growth path. By spinning it out instead of embedding it, NYT is treating games not as a feature to support news consumption, but as a product category worth owning independently, effectively stepping into the role of a game publisher rather than a media company.

What’s interesting is that this isn’t a sudden pivot. I actually reviewed this strategy back in August 2025, when The New York Times crossed 11.6M subscribers. Even then, it was already clear that games were doing more than driving engagement as they were becoming a meaningful revenue lever. Looking at Crossplay today, that trajectory feels even clearer. NYT keeps growing its subscriber base, and games are taking up more space in the value proposition, not as a side benefit, but as a core reason to subscribe and stay.

One last signal that’s hard to ignore: when searching for “NYT” on the App Store, the first result is now NYT Games, not the main The New York Times news app. That’s a reflection of user behavior, engagement, and download velocity. Games are increasingly the primary entry point into the NYT ecosystem.

For many users, the relationship with the brand now starts with play, not news. That shift in discovery perfectly mirrors what we’re seeing at the product and revenue level: games aren’t a side experience anymore, they’re becoming the front door.

🎯 You might also like:

Turning daily routines into games

Health, fitness, and productivity apps are going through the same transformation.

Step counters like Sweatcoin or WeWard turned physical movement into a game, where effort is translated into points, rewards, and progress rather than goals or guilt.

Productivity and habit apps followed a similar path with the most emotional evolution of gamification is happening in self-care. Apps like Finch or Habit Rabbit ask users to take care of something by nurturing a character through daily habits.

This is Tamagotchi logic applied to modern life, and it works because it removes pressure. The product is all about attachment, softness, and care. Once users care emotionally, retention stops being a metric and becomes a feeling.

The most successful products in these categories no longer frame themselves around discipline or optimization, but around participation and continuity.

The most successful products in these categories no longer frame themselves around efficiency or self-optimization, but around participation and continuity. By shifting from performance to progression, they turn wellness and productivity into ongoing experiences rather than checklists, borrowing directly from casual game design to make showing up feel lighter, not harder.

What’s also telling is how explicitly this shows up in App Store positioning. Many of these apps openly use “gamified” in their metadata, sometimes in the description, sometimes in screenshots, and occasionally even in the title or subtitle.

That’s a strong signal: gamification is no longer just a design tactic, but it’s a selling point users actively search for.

Games exported their playbook

While games mastered retention, monetization psychology, progression pacing, and emotional stickiness, apps brought massive audiences, real-world relevance, and daily utility.

Once those two worlds merged, the outcome was inevitable:

• That’s why learning apps ship games

• Why news companies run puzzle ecosystems

• Why self-care apps feel like tamagotchis

Gamification moved from being a feature designed to increase engagement, to becoming the core structure of entire products and entire segments.

✨ Want to stay ahead in iOS Growth?

Subscribe now & get the latest trends delivered straight to your inbox!

Subscribe now

This kind of movement rarely comes from a single viral post or a clever launch.

It usually happens when something external shifts user behavior all at once.

That shift is happening right now.

🎯 Read my 2026 Predictions:

TikTok U.S. has now officially moved to the U.S.

Last week, TikTok confirmed the creation of a new U.S. based entity, backed by American investors, designed to comply with U.S. regulation and avoid a potential nationwide ban.

From a corporate and regulatory standpoint, this was positioned as a stabilizing move with local ownership and clearer governance.

But moments like this are never neutral for users.

Structural transitions come with rewritten terms, updated privacy policies, and subtle changes in how responsibility, access, and control are defined. Even when the product experience stays the same, the relationship between the platform and its users shifts.

This wasn’t a loud announcement in the app. It was rather quiet, very legal and procedural.

Which is exactly why people paid attention.

Why users are uneasy

The reaction to TikTok’s transition has been driven by new terms in the privacy policy. What surfaced repeatedly across Reddit threads, complaint forums, and press coverage was the explicit introduction of precise location tracking, which hadn’t been part of the policy before.

Adding to this, some users also pointed to wording that appeared to allow broader use or sharing of sensitive attributes, including citizenship or immigration-related status, depending on jurisdiction and enforcement context.

In the current U.S. climate, with heightened immigration enforcement, political tension, and growing skepticism toward institutions, precise location tracking does feel worrying and personal. It raises questions that go beyond product features and into trust, exposure, and control.

Most users already assume large platforms collect data. What unsettles them is when a platform asks for more; especially when that request coincides with a major governance change.

This is what makes this moment different from previous TikTok controversies. The reaction is situational. People are saying that they’re not comfortable agreeing to this now.

When trust cracks, charts move

When users hesitate, they don’t always delete an app immediately. Instead, they pause, they delay updates, reduce usage, and eventually.. they start looking around.

That’s when charts move.

Moments of trust uncertainty create a brief window where users are open to alternatives, because they are actively willing to explore.

In those moments, the App Store becomes less about loyalty and more about availability.

This is exactly the kind of environment where smaller apps can surface unexpectedly.

Attention temporarily fragments; and fragmentation is enough to reshuffle rankings in a category as competitive as social.

🎯 You might also like:

What Skylight & Upscrolled get right

1. Skylight, the open-source alternative to algorithm-heavy social feeds

Skylight positioning matters more than any individual feature.

Launched just last year and backed by investors, the app is built on the AT Protocol, the same decentralized foundation that powers Bluesky, which now counts more than 42 million users.

Skylight positions itself as a short-form, vertical video app reminiscent of TikTok, but with a radically different posture: open source, lightweight, and transparent by design. That positioning suddenly resonated as it signals restraint, visibility, and alignment with user interests at a time when opacity feels risky.

Following TikTok’s U.S. transition, Skylight crossed 380,000 total users, with growth accelerating sharply over a single weekend. According to the team, the app added around 20,000 new users in just a few days, pushing January monthly active users to roughly 95,000.

Skylight simply needed to be present as a credible option when people were ready to look elsewhere.

2. UpScrolled, the transparent & censorship-free alternative to mainstream platforms

Upscrolled benefited from a similar dynamic, but at an even more extreme velocity.

According to data from Appfigures, Upscrolled recorded approximately 41,000 downloads between last Thursday and Saturday, immediately after TikTok finalized its U.S. deal. That short window alone accounted for nearly one-third of the app’s lifetime installs.

On January 29th 2026, it reached the #1 spot in the U.S. Social Networking chart, ahead of apps like Threads and WhatsApp.

It happens because users are actively sampling alternatives.

Upscrolled’s appeal is subtle but well-timed: free expression, support for different media types, a simple feed, no visible AI positioning, and a general feeling of being lighter and less intrusive.

Growth Opportunities

Both Skylight and Upscrolled did the hard part already: they were present when demand leaked. Now the question is both simpler and harder: how do you lock in trust at the store level?

Skylight: strong positioning, under-leveraged store page

Skylight’s biggest strength is clarity of intent. Open source, decentralized tech, and a TikTok-like format without the opacity. Unfortunately, it does not translate well inside the App Store.

Right now, the page still leans a bit too much on what the product does, and not enough on why it exists; which is exactly what users care about in a trust-driven moment.

ASO opportunities:

• The title and subtitle could be sharper about differentiation: openness, transparency, independence.

• The screenshots are the biggest opportunity: frame-1 should immediately answer the question users are asking when they arrive: “Why is this safer / different?”

  • Call out open-source explicitly

  • Explain what Skylight does not do

  • Reinforce simplicity and user control

• The description is way too short and conservative, while it could be more opinionated & confident.

Upscrolled: massive momentum, but fragile conversion layer

Upscrolled’s growth is explosive, which comes with risk.

When an app jumps from niche to #1 in days, the App Store page becomes a bottleneck. Every ambiguity costs installs and every vague message increases bounce.

Upscrolled’s product philosophy is actually very compelling: free expression, multiple media types, no visible AI layer, and a feeling of being quieter and less extractive.

But that story isn’t yet obvious enough at first glance.

ASO opportunities:

• The title and subtitle currently undersell intent: expression, simplicity, or independence should be mentioned.

• The screenshots should move faster and do narrative work. Users need to understand why the feed exists.

  • Copy should be bigger

  • Highlight values more than features: “freely express thoughts, share moments, and connect with others”, as mentioned on the app’s website

  • Reinforce privacy and user control

• App icon feels slightly heavier and more “engineered” than the product experience itself.

This matters more than it sounds as during spikes, users scan fast and visual friction hurts conversion.

The growth takeaway

This moment isn’t about TikTok failing, but about trust becoming more important than ever.

Governance changes, privacy updates, and policy shifts are no longer background noise. Users notice them, discuss them publicly, and factor them into their behavior.

For growth teams and founders, the takeaway is simple but uncomfortable: distribution leaks at moments of institutional change.

If you’re building in social, wellness, finance, or any category that touches identity or personal data, your privacy posture is part of your funnel. Your values need to be legible in seconds. And your App Store page needs to confirm a story users have already heard elsewhere.

Skylight and Upscrolled didn’t create this demand, but they were ready when it appeared.

✨ Want to stay ahead in iOS Growth?

Subscribe now & get the latest trends delivered straight to your inbox!

Subscribe now