We ran that experiment on the twenty companies that reached the bottom of our funnel between 2022 and 2025 — not the casual passes, but the ones we took all the way into due diligence, where we spoke to customers, pressure-tested the market, and argued about the decision internally. These are the closest calls we have, which makes them the only passes worth studying. And the finding that mattered most was not about any single company. It was about the scoreboard itself: most of it is still blank, and the parts that have resolved largely vindicate the discipline that produced them.

That is the honest headline, and it is more instructive than a simple tally of wins and misses would be.

The Scoreboard Is Mostly Unresolved — and That’s the Point

Of the twenty companies, sixteen are still operating, two were acquired, and two have shut down. Read quickly, a 90% survival rate sounds like a fund that dodges disasters. Read correctly, it is the most ordinary outcome in early-stage venture, and it is the first thing that should humble anyone doing this exercise.

Survival is not the same as vindication. A large share of the cohort is still sitting at seed stage years later, alive but capital-light, raising bridges rather than breakouts — companies like LimeLoop and Venku that continue to operate on modest funding without any clear venture-scale trajectory. The industry backdrop makes this the base case, not the exception: seed-to-Series A progression has slowed materially, with a record share of seed deals now structured as bridges. So when a fund looks back and sees that most of its passes are “still alive,” the correct conclusion is not that it narrowly avoided greatness. It is that the verdict hasn’t been rendered. Most of these decisions cannot be scored as right or wrong yet without leaning on hindsight — and hindsight is exactly the bias a retrospective is supposed to guard against.

This is the uncomfortable truth underneath every “missed unicorn” narrative in venture. The companies that generate headlines by raising a mega-round are visible; the far larger number still searching for repeatable scale are not. A fund that judges its own discipline by the loud minority is measuring the wrong thing.

Most “Misses” Weren’t Quality Calls. They Were Fit Calls.

The most common reason we passed on companies in this cohort was not “this is a bad company.” It was valuation, or fund economics — some version of this doesn’t fit the return math of our fund. And several of the companies we passed on for that reason went on to raise perfectly real rounds afterward.

That looks like a miss only if you collapse two very different judgments into one. There is a world of difference between “this is a good company” and “this is a good investment for our fund,” and the discipline of venture lives in that gap. Bastazo is the cleanest illustration. We redirected it on valuation and fund-economics grounds — the premium didn’t fit Fund I’s structure and exit horizon. Eighteen months later it raised a $5.3M seed led by strong sector investors. The company is doing well. And our decision was still defensible, because industrial cybersecurity is capital-intensive with long, acquisition-driven exit cycles that genuinely sat at odds with our fund’s timeline. The pass wasn’t a misjudgment of the company. It was a correct judgment about portfolio construction.

Naming that distinction out loud matters, because “we passed and they raised” is the kind of fact that produces sloppy self-criticism — or, worse, sloppy over-correction, where a fund abandons its discipline to chase the next company that merely looks like the last one that got away. Discipline means being able to watch a company you passed on raise capital and still stand behind the reasoning, because the reasoning was never about whether the company was good. It was about whether it fit the fund.

What Actually Predicted Momentum

If the dataset had a genuine signal — something that separated the companies gathering real momentum from the ones treading water — it was not the size of the market on the pitch deck. It was distribution.

The companies pulling ahead consistently had a channel, not just a category. Fitmatch, which we passed on and which has since raised more than $26 million, did it by leveraging patented body-mapping technology out of retail fitting and into healthcare and sports science — riding a distribution and technology adjacency into higher-margin verticals. Canopie raised its seed with a strategic co-lead tied directly to its customer channel in maternal health. NasaClip moved from concept to commercialization on a mix of device sales and institutional and grant support. In each case, the repeatable path to the customer was more predictive of momentum than the TAM narrative that led the deck.

This is a signal we can weight more deliberately going forward. Early-stage diligence rewards the story of a large market because it’s the easiest thing to get excited about in a room. The data says the harder, less glamorous question — how, repeatably, does this company reach the next thousand customers? — is the one that actually forecasts who breaks out. That question now sits closer to the center of how we underwrite.

Where the Framework Held

Where we had a clear thesis and applied it, the framework did the job it was built to do: preserve capital and avoid structurally difficult businesses.

The clearest validation is Dora Maar, a luxury-resale platform we passed on where diligence flagged weak scalability and an expense-to-revenue ratio well above 100%. The company ceased operations in 2024. Wellfound Foods — a vending-machine model we declined for lack of a defensible technical moat — raised nearly $3 million and still could not escape the economics of an operational business without differentiated IP; it shut down in late 2025. And we held our thesis line on companies like Stem Lingo, which have performed respectably but sat outside our core focus, because portfolio coherence is itself a form of discipline. These weren’t lucky dodges. They were the framework identifying exactly the fragility it was designed to catch.

Sharpening the Instrument

A retrospective is only worth running if it changes how you underwrite the next twenty companies. This one pointed to three refinements, each of which makes the framework more precise rather than more cautious.

The first is to treat distribution as a first-class diligence question, not a footnote to market size. The companies with momentum in this cohort had a repeatable channel; the ones treading water often had a beautiful TAM slide and no proven path to the customer. Weighting channel evidence more heavily — partnerships, embedded distribution, institutional buyers — is the single highest-leverage change the data supports.

The second is to tie valuation judgments explicitly to ownership and dilution math rather than instinct. When a pass rests on fund economics, the discipline is sharpest when it’s expressed as a number: the ownership we’d hold at entry, the ownership left after the next round, and the exit valuation required to return the fund. That turns “too expensive” into a defensible, repeatable standard.

The third is to let valuation tolerance flex by sector. Strict SaaS-style multiples don’t translate cleanly to medical devices, healthcare infrastructure, or critical-infrastructure cybersecurity, where regulatory milestones can justify pricing that would look absurd in software. NasaClip is the instructive case — a company we anchored against a 460x revenue multiple that has since grown revenue quickly enough to reframe what that multiple meant. A sliding scale of valuation tolerance in regulated, capital-intensive sectors keeps the framework from mistaking a high multiple for a bad deal.

Why We’re Publishing This

Most funds run this exercise privately, if at all. We’re sharing it because the lesson generalizes past our own file. The instinct in venture is to measure yourself by the winners you caught and torture yourself over the ones you missed. The more honest measure is whether your process is disciplined enough to defend and precise enough to improve — whether the reasoning that produced a pass still holds up when the company goes on to raise, and whether the exercise teaches you something you can apply to the next decision.

On both counts, this cohort was reassuring. The framework preserved capital where the economics were structurally weak, held its thesis line under temptation, and surfaced a clearer signal — distribution — to weight more heavily going forward. At High Street Equity Partners, the goal of a retrospective was never to eliminate missed opportunities; those are the cost of doing this at all. It’s to sharpen the process that will identify the next generation of them — and to make sure that when we’re right, we know why.

This analysis was led by Krisha Chheda, drawing on a company-by-company review of High Street Equity Partners’ 2022–2025 late-stage diligence decisions.

There’s an old line about venture: show me your coinvestors and I’ll tell you what you believe.

We’ve been sitting with our own coinvestor map lately — the full list of firms, funds, angels, and strategics who’ve written checks alongside us across the portfolio — and it turns out to be one of the more honest artifacts we have. Not a pitch. Not a thesis deck. Just a record of who else looked at the same founders we did and said yes.

A few things jumped out.

The best founders really are everywhere — and so is the smart money.

We’ve said for a while that geography isn’t destiny, that the highest-upside companies aren’t clustered in two zip codes. It’s easy to say. It’s more convincing when you look at who’s actually on the cap tables next to us.

The map runs from coastal institutional names to heartland and Arkansas-ecosystem funds, from specialist healthcare investors to climate accelerators, from a former NBA owner to state-backed venture programs. These aren’t investors slumming it outside their comfort zone. They’re sophisticated capital following founders to wherever the founders happen to be building. When a Chicago consumer fund, a Notre Dame angel network, and a sports-licensing strategic all land on the same collegiate-economy company, that’s not luck. That’s a signal that the opportunity was legible to a lot of serious people at once.

Coinvestors are a portfolio’s second balance sheet.

Capital is the obvious thing a coinvestor brings. It’s rarely the most valuable thing.

Look at the pattern by category and you see what each syndicate is really for. A digital-health founder surrounded by specialist healthcare VCs and a mission-driven strategic isn’t just funded — they’re plugged into clinical networks, regulatory instinct, and distribution they couldn’t buy. A climate-transition company with an accelerator, an impact fund, and a utility-affiliated strategic on board has a path into the exact institutions it needs as customers. The syndicate is the go-to-market, if you assemble it on purpose.

That’s the part we think about hardest when we invite people into a round. We’re not filling an allocation. We’re building a founder’s second balance sheet — the one that doesn’t show up in the wire.

A good syndicate is a portfolio of perspectives, not a monolith.

The thing we’re proudest of in the map is how little it rhymes with itself. Institutional funds sit next to angel networks. Corporate strategics sit next to equity-crowdfunding communities. Impact-first capital sits next to growth-stage generalists.

That mix is deliberate. Homogeneous syndicates tend to agree with each other — which feels great until the company hits a decision where consensus is exactly the wrong answer. A founder is better served by a table where a payments strategist, a gender-lens impact investor, and a seed generalist would each stress-test the same plan differently. Productive disagreement is a feature. We try to underwrite for it.

Why we’re sharing any of this.

Partly because it’s a useful mirror for us. A coinvestor map is one of the few documents in venture that can’t be spun — either serious people showed up next to you, repeatedly, or they didn’t.

But mostly because it’s the cleanest evidence we have for the thing we actually believe: that conviction travels. When you back founders on their merits rather than their zip code, you end up in good company — literally. The map is just the receipt.

We invest alongside partners who share that conviction. If you’re building something that doesn’t fit the map anyone drew for you, that’s usually where we want to be.

High Street Equity Partners invests in, supports, and scales high-growth technology companies across Work, Health, and Emerging Tech — guided by the belief that the best founders are everywhere.

A couple of years ago we started spending real time in Arkansas — an emerging innovation hub that fit our thesis and happened to be home for one of our co-founders. We wanted specifics: how many companies, at what stages, in which sectors, funded by whom. So instead of guessing, we started keeping a list. That list turned into a map of 120+ startups, and then a second map of the investors and accelerators active across the state.

We’re sharing what we learned not because we’ve got it all figured out — we don’t — but because the picture that emerged is interesting.

We’ve catalogued more than 120 active startups across the state, sorted by stage and by sector, and we’ve built a companion map of every investor, accelerator, and venture studio that’s actually putting capital and support to work here. We did it the slow way — sitting with founders, showing up at accelerators, and having hundreds of conversations that don’t scale and aren’t supposed to. What came out the other end isn’t a hunch about Arkansas. It’s a dataset.

And the dataset says something the coasts are missing.

What the map actually shows

Two things jump out when you see the whole board laid out.

First, the shape of the market: roughly 70% of Arkansas startups are at pre-seed, and about 20% at seed. That’s an overwhelmingly early-stage ecosystem — which is exactly where patient, disciplined capital has the most leverage and the least competition. When you map the 100+ companies by our theses, they cluster cleanly into the areas we already have conviction in: Future of Work, Health (our Future of Care thesis), and Emerging Tech, with a healthy tail of everything else.

Second — and this is the part that quietly demolishes the biggest objection to investing here — the capital is real, and it’s growing. Arkansas went from $17M in total VC in 2015 to $288.5M in 2021 and $334.2M in 2022. In 2024, the state saw 13 deals at $177.2M, averaging $13.6M per deal. That’s not a nascent market anymore. That’s an investable one, and the lighter-volume years historically set up price-disciplined investors to compound.

We know this because we built the second map too: the one showing who’s actually here. Entrepreneur support organizations, pre-seed and seed funds, Series A+ funds, venture studios — the connective tissue of a functioning ecosystem. The picture it paints is a market with real infrastructure and real momentum, just without the valuation frenzy that infrastructure usually attracts everywhere else.

The objection everyone raises (and why the data answers it)

Whenever we talk about Arkansas, someone asks the same thing: “Sure, but is there enough follow-on capital? What happens when your company needs a Series A?”

Fair question. Here’s the data. In 2024, 20 investors participated in Arkansas deals — only 4 of them in-state. Louisiana had 54 (15 in-state). Oklahoma had 30 (13 in-state). Translation: syndicates form here even when the money is headquartered elsewhere. And just south of us sits Texas, the anchor of Southern venture (peaking at 1,125 deals and $13.56B in 2021) — a deep downstream pool of capital and talent we tap for our companies when their traction earns it.

So the follow-on story isn’t “hope someone shows up.” It’s “originate high-signal deals locally, then assemble the corridor syndicate when the numbers justify it.” The map tells us exactly who to call.

Why the geography actually matters

Here’s the part that makes Arkansas more than a valuation-arbitrage play. It’s not just cheaper — it’s customer-rich in a way that’s unusually good for the kind of companies we back.

Northwest Arkansas is home to Fortune-scale anchors. That concentration creates something most early-stage ecosystems desperately lack: real enterprise buyers, sitting right there, willing to run pilots. For a capital-efficient B2B company, that’s gold. It means a founder can land a pilot, get genuine buyer signal, and turn that pilot into paying revenue — all without relocating to a coast and burning eighteen months of runway on customer discovery.

For us, proximity to that dynamic is a sourcing edge. We get faster diligence backed by real buyer signal, early reads on founder quality, and the ability to make targeted introductions that shorten the distance from “interesting product” to “signed customer.” That’s not something you can do dialing in over Zoom from Sand Hill Road.

Presence, not tourism

The difference between an investor who “covers” a region and one who’s actually part of it comes down to showing up. We have seven team members and advisors living and working in the state. Our Managing Partner is on the ground several times a quarter, and one of our main state partners allows us to have an active footprint at Onward HQ in Bentonville — a hub where founders and investors actually collide. Our advisors, partners, and team are embedded across the state.

To a local founder, that means we’re not a coastal fund parachuting in for a demo day and disappearing. We’re a neighbor — present before the round, during it, and after it. To our LPs, that presence is the whole point: it produces deal flow other funds simply can’t access, because you can’t access what you don’t show up for.

Take Sober Sidekick, one of our Arkansas bets in the behavioral-health space. We didn’t wire money off a pitch. We spent months as an early diligence partner — digging into product and retention, making warm introductions to regional providers and employers, and helping shape the go-to-market narrative and pricing. That’s the model: patience, proximity, precision.

The bigger idea

There’s a lesson here that outlives any single geography, and it’s the thing we most want founders to take away.

The best opportunities in venture are rarely the obvious ones. They’re the ones hiding in plain sight, in markets everyone has decided in advance aren’t worth the flight. The edge doesn’t come from being smarter than the next fund on the same deal — it comes from doing the unglamorous work of actually mapping a market before you deploy into it. Research first, capital next.

Arkansas isn’t a side bet for us. It’s a cornerstone — the clearest proof of a thesis we’d apply anywhere: that discipline plus depth beats hype plus herd, and that the future gets built in the places everyone else overlooks.

We don’t chase the coasts. We map the overlooked, we show up, and we let the data do the convincing.

High Street Insights is where we share what we learn backing early-stage founders across the Future of Work, Future of Care, and Emerging Technologies. If you’re building in a market the coasts ignore — or investing in one — we’d love to compare maps.

Here’s how the mistake usually looks. A founder raises a pre-seed on a beautiful number — 97% model accuracy, a treatment that works in a pilot, users who love the demo. Great. Then they spend the next eighteen months polishing that same number, walk into a seed or Series A pitch leading with it, and watch the room go politely quiet.

Nothing went wrong with the metric. It just stopped being the point.

The thing nobody tells you is that KPIs aren’t fixed targets you hit once and frame on the wall. They’re a moving conversation, and the question underneath them changes as you grow. Early on, every metric is quietly answering “does this thing actually work?” A little later, it shifts to “is there early proof this becomes a real business?” And by Series A, it’s the coldest, most expensive question of all: “does this make money in a repeatable way?”

We invest at seed, so we live in the middle of that arc — the stage where “it works” has to start becoming “it works as a business.” So let’s map all three stages, across the sectors we spend our lives in, so you can see the ground move before you’re standing on it.

The three questions, in plain terms

Before the sector detail, internalize the shape of the whole journey:

• Pre-seed proves feasibility. Can the model perform? Do users engage? Does the treatment succeed? These are validation metrics. They answer whether you’ve built something real. Growth percentages here are mostly noise — going from $1K to $10K MRR is 900% growth, impressive as a number, irrelevant as a benchmark.

• Seed proves early economics and stickiness. This is our stage. The single most important thing at seed is proof that customers want your product and will keep using it — measured through retention, engagement, and willingness to pay. Investors also start scrutinizing capital efficiency hard: your burn multiple should be under 2.0x (every dollar you burn generating at least $0.50 in new ARR), and top-performing seed companies keep 18+ months of runway at all times.

• Series A proves repeatable, scalable economics. Series A readiness in 2026 means roughly $1–2M ARR, NRR above 110%, LTV:CAC of 3:1 or better, CAC payback under 12 months, and gross margins above 70% — and critically, evidence the growth engine is repeatable and not dependent on founder-led sales alone.

That’s the whole game. Notice the middle stage — seed — is where the story changes from “cool product” to “leaky bucket or compounding machine?” A company growing 15% month-over-month with 10% monthly churn is a leaking bucket, and seed is exactly when investors start checking whether you’re filling the bucket or bailing water.

Now let’s watch this play out where it actually bites.

Future of Work: from “look how accurate” to “look how sticky”

If you’re building workplace AI or automation, your pre-seed story is technical. Your accuracy rate (or “hit rate”) and your Human-in-the-Loop ratio — how often a person has to step in and clean up after your model — are what matter. High accuracy paired with a falling HITL ratio says your system works and is getting smarter. Latency, uptime, and false-positive rates round out the “it functions in the real world” case.

At seed, the question becomes adoption and retention. Are people using it daily, or did it become shelfware? Time to Value (how fast a new customer feels the benefit) and real engagement metrics — DAU/WAU, NPS — become the signal that you’ve found a wedge into actual workflows. This is also where you start instrumenting willingness to pay and early cohort retention, because those are the seeds of the unit economics story you’ll need next.

By Series A, nobody’s impressed the product works — they assume it does. Now it’s the CAC:LTV ratio (roughly 3:1 is the “you have a real business” line), CAC payback under a year, and proof that efficiency gains repeat across many customers, not just your favorite three. That HITL ratio you tracked at pre-seed? It should now be on a sustained march downward, proving your automation matures as you scale instead of secretly requiring an army of humans behind the curtain.

The same arc governs upskilling platforms (pre-seed: people complete the training → seed: the training produces measurable competency and behavior change → Series A: provable ROI, retention, and revenue-per-employee impact) and hybrid-work tools (pre-seed: people adopt the tool → seed: the tool creates focus time and completed work → Series A: measurable, repeatable organizational productivity).

Future of Care: from “it’s clinically credible” to “it lowers the cost of care”

Healthcare punishes founders who confuse these stages, because the credibility bar is higher and the money is tied to outcomes.

In value-based care, your pre-seed proof point might be HCC recapture — evidence you’re documenting patient conditions accurately. At seed, you need to show sustained clinical engagement and the early signal that outcomes are moving — patients staying enrolled, adhering, showing measurable improvement. By Series A, investors want the Medical Loss Ratio trending down and hard evidence of fewer hospital readmissions and ER visits. Translation: you graduate from “the model is sound” to “we measurably reduced the total cost of care for real patients at scale.”

Longevity and women’s health follows the same arc: pre-seed is clinical robustness and credibility; seed is retention and durable engagement (are people staying, month after month?); Series A is the LTV:CAC ratio clearing that 3:1 bar. Specialty care moves from treatment success rates (pre-seed) to consistent outcomes plus early payer traction (seed) to the unglamorous but decisive economics — payer mix, reimbursement rates, and a falling cost per encounter as you scale (Series A).

The theme across all of Care: you graduate from proving the medicine works to proving the business around the medicine works — and seed is where that hand-off begins.

Emerging Tech: from “the demo is magic” to “the unit economics survive”

Frontier tech is where the gap between stages is most seductive and most dangerous, because the pre-seed demo is often genuinely dazzling.

If you’re building AI agents, pre-seed lives and dies on Success Rate per Task — high performers target north of 95%. At seed, the conversation adds workflow entrenchment and early cost discipline — is the agent embedded deeply enough that customers can’t easily rip it out, and are you watching your compute costs from day one? This matters more than ever now that AI-native companies face deeper diligence, with investors digging into compute economics, usage depth, and workflow entrenchment alongside a credible path to sustainable gross margins. By Series A, it’s a phrase every agent founder should tattoo somewhere visible: Inference Cost per Revenue Dollar. It’s not enough that the agent completes the task; the compute cost of completing it has to shrink relative to the revenue it generates, or you’ve built an expensive magic trick instead of a margin. Worth knowing: AI-native startups often show materially higher revenue per employee than traditional SaaS, and that efficiency signal is increasingly one investors want to see prominently.

Mobility and clean tech make the jump from utilization and uptime (pre-seed) to early proof of real-world reliability and unit-level margins (seed) to Asset Payback Period, where investors want capital back in 12–18 months (Series A). And in defense/RegTech, you move from clearing the compliance gate — SOC 2, FedRAMP, the price of admission (pre-seed) — to landing initial contracts and design partners (seed) to revenue visibility through contract backlog and Net Revenue Retention, where the strongest companies clear 120% (Series A).

The efficiency metric that now shadows every stage

One thing has changed the game across every sector, and seed founders especially need to hear it: the burn multiple has become, in the words of one 2026 benchmark report, the ultimate truth serum. In 2025, 56% of seed investors called burn multiple a critical metric in their evaluation process — a fundamental shift from 2021, when growth rate alone drove term sheets.

Translation for founders raising from us and firms like us: growth alone no longer clears the bar. A company growing 150% but burning $3 to earn $1 of ARR is less fundable than one growing 80% with a 12-month CAC payback. Efficient growth beats hypergrowth now, full stop. Start tracking this early, even before it’s flattering — you can’t show a clean burn multiple trend at your next raise if you only started measuring it the month before.

What to actually do with this

You don’t need to hit your Series A metrics at seed. That’s the trap in the other direction — founders who obsess over LTV:CAC before they’ve proven anyone will stick around, optimizing economics on a thing nobody’s committed to yet. Right metric, wrong stage.

The move is to know which conversation you’re in, and to see the next one coming:

• Pre-seed: prove it works. Nail the validation metric for your sector — accuracy, engagement, clinical success, task success rate — and be honest about what’s still directional.

• Seed (where we come in): prove they stay and you’re efficient. Retention, engagement, willingness to pay, and early capital discipline. Keep 18+ months of runway, keep your burn multiple under 2.0x, and show the bucket isn’t leaking.

• Series A: prove it scales profitably and repeatably. NRR above 110%, LTV:CAC past 3:1, CAC payback under a year — and evidence the growth engine doesn’t depend on the founder personally closing every deal.

• Watch the hand-off metrics. Some numbers — HITL ratio, cost per encounter, inference cost, burn multiple — show up at every stage, but the bar changes. Early on they just need to exist; later they need to be visibly improving. A flat line there is a quiet red flag.

• Lead with the right number for the room. The single most common self-inflicted wound we see is a later-stage pitch built on an early-stage metric. Read the stage, then pick the headline.

Metrics aren’t a report card. They’re a story about what you’ve earned the right to claim. Tell the story that matches where you actually are — and start quietly gathering the evidence for the story you’ll need to tell next.

High Street Insights is where we share what we learn backing early-stage founders across the Future of Work, Future of Care, and Emerging Technologies. If a founder in your life is staring at a dashboard wondering which number actually matters right now — send this their way.

The single biggest mistake we see founders make in an M&A process isn’t sharing too little. It’s sharing everything, immediately, in excitement, because a buyer seemed interested and asking questions felt rude. Diligence is not a trust exercise. It’s a staged negotiation, and the sequence in which you reveal information is itself a form of leverage. Give it all away on day one and you’ve got nothing left to build conviction with — and no protection if the “buyer” turns out to be a tire-kicker with a competing product.

Here’s the mental model that fixes this.

Buyers build conviction in layers. Your disclosure should too.

Think of a diligence process as three doors, each one unlocked by a stronger signal of commitment. You don’t hand someone the keys to door three when they’ve barely knocked on door one.

Door 1 — After the NDA: “Prove it’s real.”

The buyer just signed a confidentiality agreement. That’s a handshake, not a marriage. At this stage they need to validate that you have a real product, real traction, and a team worth betting on — without seeing anything they could weaponize if the deal falls apart.

So you share the story and the shape of the business, not the crown jewels. Your pitch deck. A two-page product overview. A high-level architecture diagram (the kind that shows boxes and arrows, not the kind that shows an attacker where the soft spots are). Your growth chart. Anonymized case studies. A cap table summary.

Everything sensitive stays masked: no customer names, no pricing specifics, no personal data, and absolutely no source code. Redaction here isn’t paranoia. It’s strategy. You are proving momentum while keeping your powder dry.

Door 2 — Pre-LOI: “Help me price it.”

Now the buyer is serious and circling a Letter of Intent. To put a real number on the table, they need to move from “this is interesting” to “I can underwrite this.” That means deeper proof: cohort retention and churn analysis, funnel metrics, your sales cycle, redacted samples of executed contracts, a real architecture diagram, operating financials with budget-vs-actual, and a headcount plan.

Notice what’s still protected — full unredacted contracts, source code, complete legal files, sensitive security detail. The buyer gets enough validated reality to structure an offer. They do not yet get the vault.

Door 3 — Post-LOI: “Confirm everything before we sign.”

The LOI is signed. There’s now real, written commitment on the table, usually with some exclusivity attached. This is when the vault opens — under tighter access controls, not looser ones. Full customer lists, unredacted contracts, bank statements, tax filings, cap table with all the SAFEs and side letters, full pen-test reports, IP assignments.

If a buyer needs to review source code, that doesn’t mean emailing them a zip file. It means a clean room: a time-boxed, limited, auditor-style review that lets them confirm what they need to confirm without walking off with your codebase. Even at the altar, you’re still managing risk.

The seed-stage part nobody tells you

Here’s the thing that trips up first-time founders: you are not expected to have the diligence package of a public company. You will not have audited financials. You probably don’t have SOC 2. Your “security program” might be a Notion doc and a lot of good intentions. That’s fine.

At seed stage, the buyer isn’t underwriting a fortress. They’re underwriting a story with evidence behind it: why you win, what traction is real, what’s genuinely risky, and what they’ll have to invest after the deal closes. The goal of a good data room isn’t to look big. It’s to be coherent, honest, and stage-appropriate — sharing what exists, and summarizing what doesn’t with enough candor that nobody feels lied to later.

Which brings us to the single highest-ROI move in this entire process.

Write the “here’s what’s ugly” memo before they find it

Keep one short document — call it a Disclosures & Exceptions memo — that lays out your warts proactively. The customer who churned and why. The one-off discount you gave to land a logo. The contractor whose IP assignment you never quite got in writing. The minor security incident from eighteen months ago.

This feels insane. Why would you volunteer the bad news?

Because surprises are what kill deals — and specifically, what triggers re-trades, the delightful moment when a buyer discovers something mid-diligence and uses it to renegotiate your price downward. A problem you disclose upfront is a footnote. The exact same problem discovered by the buyer’s lawyer in week six is a crisis, a trust rupture, and a very expensive conversation. Surface it early, frame it clearly, and you keep control of the narrative and the valuation.

The tl;dr

A clean M&A process comes down to a handful of instincts:

• Stage your disclosure. NDA gets the story. Pre-LOI gets the proof. Post-LOI gets the vault. Never skip ahead.

• Redaction is a tool, not an insult. Mask names, pricing, and personal data until commitment justifies revealing them.

• Match the effort to your stage. Seed companies don’t need public-company diligence. They need a coherent, honest, well-organized story.

• Set up three folders now — 01_NDA, 02_Pre-LOI, 03_Post-LOI — and pre-sort your materials so you’re not scrambling when the moment comes.

• Disclose your own ugliness first. It’s cheaper as a footnote than as a discovery.

An exit is the rare moment where being organized is worth actual money. The founders who run a tight, staged, honest process don’t just close faster — they close at the number they were quoted, instead of the number the buyer talked them down to in week six.

Build the room before you need it. Future you, sitting across from the quarter-zip, will be very grateful.

High Street Insights is where we share what we learn backing early-stage founders across the Future of Work, Future of Care, and Emerging Technologies. If this was useful, forward it to the founder in your life who’s about to open a data room and has no idea what they’re doing.

When tech shows up, it offers a shortcut and convenience.

You can use Google Maps to direct you somewhere without paying much attention to the surroundings.

You can use Claude to write your marketing copy and get a better-than-mediocre result the first try.

You can look for a gift on Amazon, pick the first match, and be pretty sure it’ll do the job.

Tech adoption often focuses on making things easier, simpler, and pre-decided.

And yet… we can also decide to use tech to do more work, insert more humanity, and amplify flexibility. We don’t try to get our time back, we try to figure out how to leverage the time we’ve got.

When a film director uses AI to create storyboards, it’s a chance to generate multiple approaches to a scene, not just one. When we sit with all the data Google Maps offers us for a trip, we might plan a less direct route, with more stops and detours, simply because we now know what our options are. And once we know what the mediocre and average marketing copy looks like, we put in the time (and take the risks) to go to edges we never would have had the resources to explore in the old days.

The best tech gives us a chance to work harder on the parts that matter to our customers and to us.

Here’s the simple fork in the road:

Professionals and organizations that use AI to save time, cut costs, and lay people off are taking a lazy road to failure and irrelevance.

Those who use it to do harder, braver, and more powerful work, who figure out how to create more value and charge more for it, and who end up hiring more people to do so, will be defining our future.

Single User Software - No Market, but Very Valuable

One of the things I didn’t fully grasp until I started building software for myself using AI coding assistants is the rapid growth in software applications for single users. What makes these products interesting to me is that they are so bespoke you would not have paid a software developer to build them for you because doing so would have been cost-prohibitive in the pre-AI era. I also doubt any developer would have pursued these products as a for-profit business because the market was too small.

We are in the very early innings of understanding what it means to have many applications with only a single user per product. That single user will continue to develop and support the product for as long as it remains useful and as long as the time and energy required to use it feel worthwhile. If that balance tips and it no longer feels useful, those products will die. I don’t believe there is a business to be built here, but this idea of single-user software really connects to another thing I have only recently begun to fully internalize as I’ve built roughly two dozen agents that I use regularly at Precursor.

So what does all of this mean for what software is worth? The picture is murky, and the impact depends on the product’s nature. Basic, utility applications with a narrow scope will face consistent downward pricing pressure from the threat that the end user will just build the product themselves. They might never build it, but the belief they could build it will make them more price-conscious and skeptical of paying more for products that deliver narrow value propositions. The result isn’t uniform deflation - it’s a split. Products that require maintenance, new feature development, integrations, and the ability to interact with the customer’s AI tooling and infrastructure will still be purchased, as buyers won’t want to do all that work themselves. The experience of being a junior builder will reinforce the value of this work to the buyer, but they won’t pay software-like margins for it; it will be valued, but less than what software vendors are used to getting from license revenue.

LeanSite is operating within one of the largest and most operationally fragmented markets in the economy. Facilities management represents a $700B U.S. market and a $1.7T global opportunity, yet many enterprise operators still rely on spreadsheets, emails, phone calls, and disconnected legacy systems to manage critical maintenance operations. This fragmentation creates costly inefficiencies across vendor coordination, asset management, compliance documentation, and reactive maintenance workflows.

What excites us most about Demi is her deep operational understanding of the problem. Before building LeanSite, Demi spent years working within cleaning and facilities operations, giving her firsthand exposure to the inefficiencies facility managers navigate every day. That operational proximity is reflected throughout the company’s product design. LeanSite’s Automated Maintenance System centralizes work orders, vendor dispatch, asset history, budgeting, compliance documentation, and maintenance workflows into a single operating layer designed to help enterprises automate and streamline facility operations at scale.

Today, LeanSite supports operations across more than 100 facilities and 27 enterprise customers, including brands such as PF Chang’s, LA Fitness, and PetSmart. The company has demonstrated strong early traction, recurring revenue growth, and customer retention while operating with a lean and capital-efficient team. We believe LeanSite is well positioned at the intersection of facilities management, operational automation, and enterprise AI infrastructure, particularly as labor shortages, compliance requirements, and operational complexity continue driving demand for modern facilities software.

Our investment in LeanSite is driven by the strength of the founding team, the scale of the operational problem, and the company’s long-term vision to modernize facilities management through automation and AI-driven workflows. We believe the facilities operations category remains significantly under-digitized relative to other enterprise functions, creating an opportunity for LeanSite to become a system of record for multi-location operators. Demi and her team have built a thoughtful platform within a massive market, and we are excited to support the company as it scales its enterprise footprint and advances the future of facilities operations.

Intellectible is an emerging leader within an $18B+ total addressable market of enterprise service providers whose revenue operations still run on PDFs, spreadsheets, email, and tribal knowledge. What excites us most about Jesse is his deep familiarity with the customers Intellectible serves. Over 17 years, he founded and sold an automation business, then co-founded an education and government supply company that grew from $0 to $10M ARR in under three years before selling to private equity, work that earned him EY Entrepreneur of the Year recognition in London. Reuben brings deep AI/ML and runtime systems expertise from his work creating Renderella, a leading point-cloud rendering system, and contributions to Apple Vision Pro. Rosie brings enterprise integrations and data systems experience from leading multi-million-dollar CRM and Salesforce implementations. Together, the founding team combines repeat operator experience, technical systems depth, and the integration discipline required to embed deeply into enterprise workflows.

The Intellectible platform is organized around four configurable AI Engines: a GovCon Engine that monitors SAM.gov and other government sources to deliver curated, right-fit opportunities; a Proposal Engine that parses complex RFP packages into compliance matrices, outlines, and draft response sections; a Pricing Engine that automates the multi-variable pricing workflows that historically take 16–24+ hours per bid; and a Knowledge Engine that turns prior proposals, SOPs, and past performance materials into governed, reusable libraries. Rather than functioning as a thin AI wrapper, the platform combines deterministic workflow logic with AI execution, producing the auditable, repeatable outputs that revenue-critical and pricing-critical workflows demand. Early customers including HHS, Oceus, and Office Design Group have already seen 95% reductions in manual RFP and opportunity parsing, with hundreds of millions in qualified pipeline identified across the customer base.

We believe Intellectible presents a compelling opportunity at the intersection of enterprise AI and the Future of Work. Federal contracted services spending has grown materially over the past decade, services have become a larger share of the U.S. economy, and service organizations cannot continue to scale revenue operations linearly with headcount. Our due diligence brought to light Intellectible’s significant growth trajectory, with ARR expanding from approximately $100K at the start of 2025 to more than $635K today, 15+ enterprise customers in production, and $200K+ of additional contracts in redline. This momentum, paired with a land-and-expand motion that begins with a low-friction GovCon Engine deployment and expands into higher-value Pricing and Knowledge workflows, reinforces our confidence in the scalability of the model.

Our investment in Intellectible is driven by the team behind the vision, the white space within a large and underserved market, and a business model that combines high-margin SaaS economics with deep workflow integration. We believe Intellectible is poised to become the default system of record and system of action for how enterprise service providers win, price, manage, and deliver work.

Here’s the headline that’s easy to misread:

• $10.4B in total pre-seed capital raised in 2025

• −1% change in dollars deployed vs. 2024

• 92% of pre-seed rounds used SAFEs

• −13% change in number of deals closed

Read those numbers fast and the story looks flat. $10.4 billion. Down 1%. Nothing to see here.

Read them carefully and something shifts.

The dollars barely moved. But the number of checks written dropped 13%. That gap, same money, far fewer deals, is where the real story lives. It means the average round got larger. It means fewer founders walked away with a term sheet. It means capital didn’t spread out across the market the way it did in prior years. It concentrated.

Investors didn’t get cheaper. They got more deliberate. And that distinction, concentration, not collapse, is the one that actually predicts who gets funded next.

What concentration looks like in the data

Among SAFEs raising at least $1M, the average deal size in 2025 was $1.4M, up from $1.1M in 2024. Median caps rose to $7.5M. Investors are paying premiums for conviction and ignoring everyone else faster, what founders are calling the “speed of No.”

That speed isn’t dismissal, but rather discipline. With smaller pools of conviction in any given quarter, investors have less to lose by passing quickly. Risk-precise, not risk-off.

The middle is thinning. On one end you still have small rounds around $250K friends-and-family checks moving as they always have. On the other, a smaller number of well-capitalized, lead-led rounds at premium caps. Party rounds are fading. Operators are the new scouts. Whoever writes your first check increasingly determines who’s even in the room for your second.

That last sentence is where the access question lives. Merit is only half the story.

Concentration rewards the pattern-matchers

When investors write fewer checks, they lean on the shortcuts they already trust. Warm intros. Brand-name pedigree. Proximity to networks they recognize. This is rational behavior. It is also where structural disadvantage compounds fastest.

Social capital functions like a key here. It unlocks rooms, the conviction meeting, the second take, the “I know someone you should talk to.” Founders who inherit those keys (alumni networks, parents who invested, friends who’ve exited) walk in pre-credentialed. Founders without them are running extra reps just to get in the door, before any conversation about the company has even started.

A founder building in Little Rock or Louisville doesn’t have less to offer than one building in SoMa. They’re operating in a market where the filter wasn’t calibrated for them. When capital concentrates, that miscalibration matters more, not less.

This is the part that gets flattened in the headlines. “AI is taking everything” is a louder narrative than “concentration is widening the access gap”, but only one of those framings actually predicts who gets a check next year.

What this means if you’re raising in 2026

A few things worth holding onto.

The bar is higher, but it’s specific. Pre-seed isn’t about product readiness or revenue readiness. It’s about evidence readiness, did you build something before the raise that the next investor can believe in? Earned insight (why this founder, why this problem). A real distribution wedge (a repeatable way to reach users where cost < value). Capital efficiency (what you built, with how little, how fast). Those three carry more weight than a deck.

The cap is the deal. At $1M–$1.9M raises, median dilution is 15.6%. At $5M–$5.9M, it jumps to 23.7%. Founders optimizing for the highest headline valuation without modeling forward dilution are making a compounding mistake every round after. Treat your SAFE like a priced round, because functionally it is one.

Networks matter; dependence is dangerous. Access to investors without relying on a single relationship. That’s the ask. Build the network now, before you need it, because warm intros aren’t a strategy you can backfill mid-fundraise.

AI is a feature, not the moat of a founder’s business. LLMs are changing how we work and build, but without intentional integration they become a liability. I call it “Scatter-shot AI”, deploying it broadly for efficiency without redesigning the system. You speed up one area, but the bottleneck just shifts to ops, quality, or customer experience. AI doesn’t fix weak systems; it exposes and accelerates them. Use it deliberately, or it will scale your problems as fast as your output.

High Street’s lens

Here’s where I’ll close in the firm’s voice rather than my own.

High Street’s conviction, we are intentionally building hubss of innovation around our core theses: the Future of Work, the Future of Care, and Emerging Tech. As structural shifts continue to reshape how people live, earn, and access care, we focus on the systems that underpin human productivity and well-being. AI is redefining how work gets done, health outcomes remain uneven, and the definition of human value is expanding beyond traditional metrics. These converging forces are not incremental, they are foundational. We believe they will drive the most consequential transformations in the years ahead.

We read this report not as a story of collapse, but as a map. Concentration tells you where the filters are tightest. It also tells you where the inefficiencies are largest. Both matter. My bet is that this opens a new landscape where top VC firms lack insight in rural areas, where transformation of work and hardware are key industries in the rust belt and midwest.

For founders, the work is staying capital efficient, building real evidence, and using the gap between pre-seed and seed as a focused window rather than a waiting room. For investors, SAFEs are priced rounds, sector frameworks need to flex (hardware and biotech don’t follow SaaS dynamics), and the AI-enabled vs. AI-adjacent distinction is going to define which 2025 vintage actually compounds.

Capital is risk-precise now. That’s the real shift. The question isn’t whether the market is open, it is. The question is whether your evidence, your network, and your cap math are calibrated for a market that’s paying premiums for conviction and ignoring everyone else faster.

Read the data carefully. Then build accordingly.

Written by Tony Lodge, HSEP Venture Fellow

____________________________________________________________________________

Sources: Carta, State of Pre-Seed 2025 in Review (Feb 2026)

The venture capital industry in 2026 is operating in two entirely different realities depending on whether you are building with artificial intelligence or without it. If you are in the former camp, valuations are setting records, check sizes are swelling, and the headlines look euphoric. If you are in the latter, you are navigating the most selective early-stage funding environment in more than a decade. Both things are simultaneously true, and the tension between them defines everything about how capital flows, or why it stalls now.

This piece is not an indictment of the industry. It is a challenge to founders and investors alike to interrogate their own assumptions before they deploy another dollar or sign another term sheet. The data makes clear that the playbook most people are still running was written for a market that has drastically changed.

More Money Into Fewer Companies

The headline numbers for venture capital in 2026 look almost triumphant. Total venture funding is on pace to exceed $120 billion for the year, up from roughly $97 billion in 2024. Global funding through the first half of 2025 alone hit $205 billion (up 32% year-over-year) and had the strongest half-year since early 2022, per Crunchbase. Two of the largest single venture financings ever recorded occurred last year: Scale AI’s $14.3 billion round in Q2 2025 and OpenAI’s staggering $40 billion raise in Q1. These are generational capital events. They are also deeply misleading as indicators of health for anyone operating below the late or mega-cap stage.

Strip away these huge rounds, and the picture inverts. Seed deal count was down 10 to 13 percent year-over-year through 2025. In Q1 2025 specifically, Carta tracked just 401 seed rounds — a 28% decline from the same quarter a year prior — with $1.2 billion raised, down 37%. The venture industry is not growing uniformly. It is concentrating, consolidating, and bifurcating in ways that make aggregate statistics almost useless as a guide to what founders and early-stage investors are actually experiencing on the ground in the U.S.

Cambridge Associates captured the structural problem precisely in its 2026 venture outlook: more than 4,200 venture funds have been raised in the United States since 2022, with 42% of 2024 vintage funds classified as micro-funds in the $1 million to $10 million range. This number is way up from just 25% in 2020. More funds, same number of breakout companies. The math is unforgiving.

Too Many Seeds, Not Enough Series A

Crunchbase data illustrates what practitioners have nicknamed “the Orange Crush” problem: over the past few years, seed round volume has grown by approximately 33% while the number of Series A rounds has declined by nearly 10%. What was a roughly 1:1 ratio of seed to Series A deals in 2008 has steadily deteriorated to approximately 2:1, and this gap continues to widen. You now have an ecosystem with roughly double the number of seed-funded companies competing for the same number of institutional Series A slots that existed a decade ago.

The median time between seed and series A has stretched to 616 days as of Q2 2025, which is more than 20 months and roughly two months longer than just two years prior, per Carta data. Series A investors have responded to the glut of candidates by raising their proof-point thresholds substantially. Revenue trajectory, repeatable sales motion, unit economics discipline, and team stability are no longer differentiators at this stage; they are the bare minimum.

“Only companies with the strongest competitive positions are attracting substantial funding. Investors are prioritizing companies with strong unit economics, growth, and defensible market positions.”

— Wellington Management, Harvard Law School Forum on Corporate Governance, December 2025

For founders, 18 months felt like a long runway in 2021. In 2026, it is a liability. Building for 24 to 36 months of capital is not paranoia; it is the new baseline for anyone who wants optionality in their fundraising process. The bridge round surge tells the story plainly: bridge financings now represent 16.6% of all venture capital deployed in Q2 2025, up from 11.8% a year prior. Many companies are not bridging from strength. They are bridging from desperation, extending SAFE agreements to buy time for milestones that institutional investors keep moving further out of reach.

Fear, Greed, and the 42% Premium

Let us be direct about what is happening with AI and venture capital: it is both the most rational allocation decision in the industry right now and a source of serious malinvestment risk. Those two things coexist, and shifting them in either direction by dismissing the AI opportunity or surrendering all critical thinking to the hype, is how both founders and investors lose money over the next cycle.

The rational case is formidable. PitchBook data shows AI startups captured 65% of all US venture deal value through Q3 2025, with more than half of all new unicorns built on AI innovation. The median age at first financing for AI startups is 65% lower than that of non-AI peers, meaning these companies are raising capital faster and progressing through rounds more quickly. This could also be a signal that AI companies At our seed stage specifically, AI companies command a 42% valuation premium over their non-AI counterparts. In the 95th percentile of seed rounds, post-money valuations have hit $80.5 million. This figure would have been mid-Series A territory just four years ago. At High Street, this growth in seed round valuations depletes the number of deals we can consider at our sweet-spot valuation of under $15 million.

The irrational case is also visible if you look. QED Investors, in their 2026 predictions, put it bluntly: “From the smallest local pre-seed fund to the largest megafund, VCs are people with a hammer seeing the whole world’s problems as AI nails. Though hints of skepticism have snuck back into whispers, without a core AI story it feels nearly impossible for a company to get funded. This has led to malinvestment, particularly at the earliest high-risk stages, where funds that previously would take fliers on non-consensus deals crowd into ‘also-ran’ AI companies.”

The AI Veneer Problem

Investors who cannot distinguish between genuine AI-native architecture and a traditional SaaS product with a GPT wrapper are making the same mistake that LPs made with “mobile-first” companies in 2012 and “blockchain-enabled” companies in 2018.

The category label is not the moat. The data advantage, proprietary training pipeline, switching costs, or network effect is the real moat. Diligence on this distinction will separate this cycle’s winners from its write-offs.

One investor quoted by Crunchbase framed the structural risk clearly: “The competition is not between AI and a business process outsourcing contract. It’s between two AI companies. VCs are missing the power of competition to drive margin down in the age of AI.”

The sector-specific data shows both extremes in sharp relief. Gaming companies represent the cautionary tale of the current environment: only 2.3% of gaming startups that raised seed in Q1 2022 made it to Series A within two years, per Carta. For traditional SaaS, the graduation rate collapsed from 37% in 2020 to a measly 12% by 2022. These are not marginal declines. They are category-level capitulations that should be informing how fund managers construct their portfolios right now.

Last Cycle’s Bets Aren’t Paying Off

Perhaps the most consequential and least discussed factor of the current environment is what is happening inside the funds that deployed heavily during 2021 and 2022. The short version: it is not good, and the timeline to resolution is longer than most LPs have been told.

Carta’s fund performance data shows the median unrealized IRR for 2021 and 2022 vintage funds remains below zero. Only 37% of 2019 vintage funds and 30% of 2020 vintage funds had made any distributions to LPs whatsoever by Q1 2025. These are not distressed outlier funds. These are representative data points across the industry’s most active deployment years. Meanwhile, 2022 vintage funds had deployed only 43% of their committed capital by the 24-month mark — the lowest deployment rate on record.

The downstream effect of this distribution drought on the seed stage is significant and underappreciated. Cambridge Associates noted that LPs are increasingly entering the secondary market as first-time sellers, with continuation vehicles estimated to represent at least 20% of distributions in 2026. Secondary special purpose vehicles surged 682% from 2023 levels through 2025. Liquidity is not returning through the traditional IPO and M&A channels quickly enough, and the alternative plumbing being built to compensate for that shortfall represents a fundamental shift in how the venture asset class delivers returns.

“The private for longer dynamic compounds the challenges facing current seed-stage investors. As average hold periods extend, and the bar to go public or achieve significant M&A becomes more elevated, winners may become rarer and more consequential for the asset class.”

— Cambridge Associates, 2026 Venture Capital Outlook

The concentration of LP capital into the largest firms adds another layer of structural pressure. Panelists at Startup Boston Week 2025 noted that 90% of all venture capital raised over the prior two years flowed to just three firms — Andreessen Horowitz, General Catalyst, and NEA. The implication for micro-funds and emerging managers is stark. Not only are they competing for the same deals as multi-billion-dollar platforms with brand advantages and deep portfolio networks, they are doing so at a moment when their own fundraising environment has become materially harder.

What Winning Looks Like Now

If the old playbook was “raise seed, raise Series A, raise Series B,” what replaces it? The honest answer: several different things, depending on the company. The venture industry’s most important intellectual task over the next 12 to 24 months is expanding its definition of a successful outcome beyond the Series A graduation event that the entire ecosystem has been optimized around.

Among the 15% that do graduate to Series A, we’ve noticed a consistent profile. Carta data shows these companies typically display the following traits:

• $100,000 to $500,000 in MRR

• Strong trajectory, with 10% month-over-month growth a common threshold

• Retention curves leveling at 80% or higher

• Evidence of inbound demand signals & genuine product-market fit versus pure sales intensity

• Sales motion must be repeatable without the founder in the room

For the 85% that will not graduate on the traditional timeline, the conversation needs to shift earlier. The most forward-thinking seed funds are already building frameworks for three distinct outcomes: Series A graduation, acquisition by a strategic buyer or private equity, and sustainable revenue business with no further institutional funding. That third option has been treated as a failure mode by the venture industry for years. In the current environment, it may increasingly represent rational capital allocation.

Visionary Shift

A growing number of seed managers are quietly repositioning around “optionality investing” — deliberately constructing portfolios where some percentage of companies are underwritten for Series A graduation, some for acquisition or strategic M&A, and some for capital-efficient revenue growth with no further dilution.

This trimodal outcome model requires different deal terms, different milestone frameworks, and different investor-founder conversations from day one. Though, it may be the most honest response to a graduation rate that is structurally settling around 15%.

Service-as-Software is also reshaping the addressable market calculus. PitchBook notes that enterprise software is shifting from seat-based SaaS to outcome-based models — a development that changes how TAM is sized and how revenue compounds at early-stage AI companies in ways that most traditional seed underwriting models have not yet accounted for.

Defense-tech and robotics are emerging as categories where the hardware cost curves and procurement visibility have finally converged in ways that make early-stage bets more underwriteable than at any point in the past decade.

WHAT FOUNDERS AND INVESTORS MUST RECKON WITH

Six Risks That Define the Next 12 Months

1. The solo founder surge is a due diligence gap.

Thirty-five percent of startups incorporated in 2024 had solo founders, a growing trend & more than double the rate from 2015, per Carta. Yet, only 17% of VC-backed companies are solo founded, meaning investors are systematically discounting solo teams at the point of institutional investment. More troubling is the co-founder attrition data: roughly one in four VC-backed two-founder teams loses a co-founder within four years, with that rate climbing to 35 to 40% by years six through eight. Teams that appear to satisfy the co-founder preference filter are more fragile than the statistics suggest.

2. The valuation gap between AI and everything else is creating a dangerous selection effect.

When 42% of all seed capital flows to AI companies and the 95th percentile of seed valuations touches $80.5 million, founders in every sector are being incentivized to attach AI narratives to products that may not fundamentally benefit from them. This is the growing “AI veneer” problem where a large portion of companies will face brutal renegotiations at Series A when institutional investors apply real revenue and retention scrutiny.

3. Geographic concentration is a structural disadvantage not adequately priced.

Pre-seed capital flows 38.4% to California startups, with the Bay Area, New York, Boston, Los Angeles, and Washington, D.C. dominating the distribution map. European founders face an even steeper climb: EU seed graduation rates are approximately 30% lower than US rates, with only about 11% making it to Series A by the same two-year window.

4. The down round risk has not disappeared.

Down rounds represented 17% of all rounds in Q3 2025, technically the lowest level in three years. But this follows a period in which down rounds exceeded 20% for seven of eight consecutive quarters from 2023 through Q1 2025. Median seed-to-Series A step-up multiples have compressed from 4.2x at the 2021 peak to 2.6x today, a 38% decline that mathematically reduces the room investors have to generate fund-level returns.

5. The bridge round normalization is masking a zombie company problem.

Bridge financings rising to 16.6% of all venture capital is not inherently alarming. The concern is that when bridge rounds are growing as a share of total capital deployed while the number of underlying deals is declining, it suggests that a meaningful portion of the seed ecosystem is in extended survival mode. This ends up burning through additional capital while waiting for a Series A environment that may not materialize on their timeline.

6. The power law is becoming more concentrated, not less.

Cambridge Associates cautioned that “missing” power law winners can result in a venture program that underperforms expectations across an entire fund cycle. Given that today’s IPO-ready companies require median trailing twelve-month revenue of over $500 million and 31% growth rates to access public markets, the probability that any given seed investment represents a company that can reach that threshold is genuinely small. The math of the asset class has always required power law outcomes. What has changed is how many funds are competing for the same small number of companies that might generate them.

High Street’s Advantage in this Market

In this changing landscape, High Street is actively navigating these risks. We pride ourselves in running a fund that only invests when our team can add real value. We also diversify across emerging tech, future of work, and digital health, which shows we only pay the AI premium in sectors where we see long-term advantages. We also apply many of the winning profile filters to our thesis as we ensure our portfolio companies have proprietary data, high retention, and crystal clear product-market fit. By exploring underserved regions and betting on underrepresented founders, we also have an advantage because High Street avoids competing in overinflated areas or deals based in hype. This allows us to block out the noise and treat every deal the same as it runs through our diligence process.

In a 15% market, you engineer success through quality instead of investment volume. Most seed funds built their models on abundance and are now facing harsh reality. We built our fund for a different thesis, and the current market is highlighting our strengths.

✦ ✦ ✦

The venture capital industry does not have a capital problem in 2026. It has a clarity problem. There is more money in the ecosystem than at almost any point in history, and yet it is flowing through narrower channels toward fewer companies, in fewer sectors, in fewer geographies. The funds that will outperform over the next decade are not the ones that identified AI as important. By now, everyone has done that. It’s likely the ones that built genuine frameworks for distinguishing the 15% from the 85% before the Series A process makes that answer obvious.

For founders, the implication is straightforward even if the execution is hard: the market will not reward you for raising capital. It will reward you for building a business that institutional investors cannot rationally pass on. That is a higher bar than it sounds, and achieving it requires treating the 616 days between your seed and your Series A as the most operationally important stretch of your company’s life, not as a fundraising interlude!

For investors, the data argues for honest underwriting about what graduation rates are achievable and what fund return profiles look like at 15% conversion rather than 30%. A fund built on the assumption that one in six seed investments reaches Series A is a very different portfolio architecture than one built on the assumption that one in three does. Right now, the gap between those two assumptions is where a lot of investor capital and a lot of founders’ time is quietly disappearing.

The “15% truth” is uncomfortable because it implies that most of what is being funded at seed right now will not produce institutional venture outcomes on traditional VC timelines. Discomfort is not a reason to ignore a data set. It is, historically, the best reason to pay attention to one.

Cheers!

Written by Jacob Teeter, HSEP Venture Fellow

_________________________________________________________________________

SOURCES & CITATIONS

Carta. State of Private Markets Q1–Q3 2025; VC Fund Performance Reports; State of Seed 2025. Data represents activity across 50,000+ startups and 2,500+ venture funds. carta.com/data

Crunchbase. Seed-to-Series A ratio trends; 2026 VC Outlook with commentary from Insight Partners and Menlo Ventures. news.crunchbase.com, December 2025.

PitchBook. 2026 US Venture Capital Outlook; early-stage deal activity through Q3 2025. pitchbook.com, December 2025.

Cambridge Associates. 2026 Outlook: Private Equity & Venture Capital Views. cambridgeassociates.com, December 2025.

Wellington Management / Harvard Law School. “Venture Capital Outlook for 2026: 5 Key Trends.” corpgov.law.harvard.edu, December 2025.

QED Investors. 2026 Fintech and Venture Capital Predictions. qedinvestors.com.

Startup Boston Week 2025 Panel. “Venture Capital Crystal Ball: What 2026 Holds for Startups and Investors.” startupbos.org, November 2025.

Duet Partners. European vs. US graduation rate comparison analysis.

According to the Solo Founders Report 2025 by Carta, more founders than ever are starting companies on their own. The share of startups founded by a single founder has grown from 23.7% in 2019 to 36.3% in 2025, a significant shift in how companies are being formed. Lower startup costs, cloud infrastructure, and advances in AI tools have made it easier for talented individuals to build and test ideas without needing a full founding team from day one.

But while more companies are starting this way, venture funding hasn’t fully followed the same trend. Solo founders make up roughly 30–35% of startups, yet they receive only about 14.7% of venture capital funding. Investors often point to familiar concerns: reliance on a single person, limited leadership bandwidth, and the lack of complementary skill sets that a founding team can bring. As a result, solo founders often bootstrap longer and show more traction before raising capital.

Interestingly, the data also shows that solo founders tend to hire earlier than teams, bringing on their first employees in about 399 days on average compared to 480 days for founding teams. In many cases, these founders replace the role of co-founders with early hires, building out the capabilities they need as the company grows.

The High Street Mindset

At High Street Equity Partners, we view this shift less as a change in who builds companies and more as a reflection of how much individual leverage has increased.

Today, a single founder can build products, reach early customers, and validate ideas much faster than even five or ten years ago. Tools powered by AI, scalable infrastructure, and global access to talent allow founders to move quickly and test bold ideas with relatively small teams. That dynamic aligns closely with our belief that exceptional talent and disruptive ideas can emerge from anywhere.

At the same time, the data also reinforces something investors have long recognized: while one person can start a company, scaling it almost always requires a team. The biggest risks investors see in solo-founder companies are key-person dependency, leadership bandwidth, and operational coverage which become more important as the company grows.

What we often see in the strongest solo founders is an awareness of this reality. The most successful ones don’t try to do everything themselves indefinitely. Instead, they move quickly to bring in the right people around them early hires, advisors, and operators who complement their strengths. Over time, these founders effectively build the functional equivalent of a founding team, even if the company started with just one person.

There are also moments where solo founders may have a real advantage. In the earliest stages of building a company, clarity of vision and speed matter a lot. Without the need to align multiple founders, solo builders can often move faster, experiment more freely, and make decisions quickly. In areas like Future of Work, Future of Care, and emerging technologies, that speed can make a meaningful difference in the early phases of innovation.

A Deeper Lens for Founders and Investors

For founders building companies on their own, the takeaway isn’t that being a solo founder is a disadvantage, it’s that investors will often look for slightly different signals. Execution speed, early traction, and clear market insight become even more important. Just as critical is showing that you can attract strong people around you, whether through early hires, advisors, or operators who strengthen the team as the company grows.

For investors, the rise of solo founders suggests it may be worth looking beyond the traditional “founding team” lens. Some of the most interesting founders today are individuals who move quickly, operate with strong conviction, and show unusual resourcefulness in the early stages. Evaluating how a founder builds momentum and attracts talent can sometimes tell us more than the initial team structure itself.

And for founders thinking about whether they should bring on a co-founder, the answer is rarely one-size-fits-all. A co-founder can be valuable when there is a clear gap in expertise or when shared leadership would strengthen the company. But bringing someone on simply to match a perceived venture model can create its own challenges. What ultimately matters is not whether a founder starts alone, but whether they can build the right team around the vision over time.

The Status Quo

Personally, I don’t think the solo-founder conversation should be framed as “solo vs. team.” What matters more to me is the founder’s clarity of vision, resilience, and ability to execute.

That willingness to take the first step and figure things out along the way often says a lot about how Solo founders operate as builders.

At the same time, building a venture-scale company eventually becomes a team effort. The founders who succeed over the long run are usually the ones who recognize when it’s time to bring in great people around them and build real leadership depth.

From my perspective as an early-stage investor at High Street Equity Partners, whether someone starts with a co-founder isn’t the most important signal. What I pay more attention to is how they think, how quickly they execute, and whether they can attract talented people as the company grows.

More companies are being started by solo founders because technology allows individuals to build more than ever before. But the companies that ultimately scale tend to succeed because those founders build strong organizations around their ideas.

In the end, the future of startups may not be about choosing between solo founders and founding teams. It may simply be about backing exceptional founders, wherever they are and however they start , who can turn bold ideas into enduring companies.

Written by Alexandra Strong, HSEP Venture Fellow

The evidence above points to a concept that would be prudent to consider during manager selection: the diversity dividend. The NAIC describes the diversity dividend as the “measurable financial and organizational advantage that results when the full spectrum of talent is given the opportunity to compete. Diverse leadership and investment teams bring differentiated insights, stronger decision-making, greater resilience, and outcomes that consistently translate into superior returns.”

LPs have a fiduciary duty to try to capture the dividend to maximize risk-adjusted returns for beneficiaries. Excluding diverse and emerging managers—through AUM thresholds, fund sequence requirements, or brand-recognition heuristics—suppresses returns. Institutional screening processes, the Colibrí Institute notes, “systematically exclude alpha-generating fund managers at the phase where their advantages are strongest.”

For LPs, the research from the BCG, Cambridge Associates, the Colibrí Institute, the NAIC, and StepStone, should trigger reassessments of screening criteria, program design, and manager selection—not as a DEI response, but as a performance imperative

Practically, Institutional investors can:

1. Evaluate capability and configuration, not proxies. AUM thresholds, minimum fund sequences, and brand recognition predict institutional embeddedness, not performance. Assess instead whether the manager’s portfolio construction choices (check size, breadth, stage focus, pacing) align with their resource environment and capabilities. That alignment is a better predictor of performance outcomes, according to the Colibrí Institute, given that over a decade ago, “NexTier Consulting observed that “the single greatest misconception plaguing the selection process is that AUM is an accurate proxy for a firm’s ability to manage operational, financial, compliance, and other business risks.””

2. Integrate, don’t isolate. Emerging manager programs that operate as separate, compliance-driven portfolios held to different standards than core allocations produce weaker results and create skepticism. The data supports treating diverse and emerging managers as part of the core portfolio, with inclusion grounded in quantifiable returns goals and strategic investment objectives.

3. Act on first-mover advantage. From 2021 – 2023, the number of diverse funds grew at a 21% CAGR, according to StepStone, who notes that there ~530 and ~1,000 diverse GPs and funds, respectively, in 2024. Most remain significantly undercapitalized relative to demonstrated performance. LPs who establish relationships now can gain access to the top managers before the market prices in their track records. In VC, early relationships govern access to future allocations, capacity, and information advantages across investment cycles.

High Street’s Perspective: Living Proof of the Research

High Street Equity Partners (HSEP) is a values-driven, Black-led venture capital firm founded in 2022. We invest at the seed stage in post-revenue, tech-enabled companies across emerging innovation hubs (the Mid-Atlantic, Arkansas, and other regions), with a thematic focus on the Future of Work, Health Technology, and Emerging Tech (including Clean Tech, Cybersecurity, and AI). We target companies raising $100 – 300K seed rounds, aim for 5–15% ownership, and are building a target portfolio of 20 companies in our $15 million inaugural fund.

We are, by the Colibrí Institute’s definition, a multidimensional emerging manager: (Fund I, firm age 5 years, AUM $100M, diverse founding team). We are proof of the research that managers meeting three or more emergence criteria have an amplified IRR advantage and view our profile and position in the market as an edge, not a limitation.

We are also testament to another observation from the the Colibrí Institute’s research on diverse managers, which is that while “early-sequence funds can signal higher perceived uncertainty” they can also “reflect a team with unusually strong investing experience, even if the track record sits in prior roles rather than prior vehicles.” We are deploying Fund I, but have unusually strong investing experience.

A Track Record That Predates the Fund

Prior to founding HSEP in 2022, Managing Partner Mitch Brooks began angel investing and advising startups in 2012. His pre-fund track record across 16 investments spanning venture debt, seed, and Series A rounds produced a portfolio IRR of 52.66%, compared to the S&P 500’s 11.02% over the same period. Standout investments include Birchett (150% IRR, 2.5x MoM); Big League Logistics (107% IRR, 5x MoM), and ProteiosBio (73% IRR, 5.25x MoM).

Mitch’s angel portfolio demonstrates strong returns embedded in prior roles and vehicles in addition to thematic consistency—the majority of his angel investments align directly with HSEP’s thesis, as they are concentrated across emerging geographies, underrepresented founders, and sectors where diverse networks open an information advantage. This is precisely the kind of capability evidence (investing experience, not yet reflected in a fund-level track record) that the Colibrí framework calls on LPs to interrogate. That said, and as we’ve previously analyzed, our inaugural fund ranks in the 90th percentile for its vintage across several key metrics, including a TVPI of 2.24x and an annualized IRR of 22.7%.

Our Fund’s performance thus far reflects three pillars: 1) values-driven diligence that underwrites to resilience over pedigree; 2) thematic discipline concentrated in sectors where HSEP has outsized conviction and sourcing advantage; 3) and intentional portfolio construction that paced capital deliberately and avoided overexposure during inflated cycles. Our choices affirm the emerging manager strategy-constraint configuration playbook described by the research; we continue to execute it to deliver outperformance.

The Moment

The structures and mechanisms of GP exclusion in VCs are understood, and the costs are measurable. The opportunity to invest in the most consistently outperforming, yet most undercapitalized category of private market managers is accessible to allocators willing to digest the data and adapt as needed. We look forward to discussing with any institutional investors who are ready to capture differentiated sources of return.

The following data points are drawn from institutional research and analyses, audited indices, and peer-reviewed research. They form a connected argument, each one illuminating a different dimension of the same underlying truth.

1. 7.2% IRR Outperformance

After analyzing 2,471 US venture capital funds raised between 2000 and 2024, the Colibrí Institute found that emerging managers outperform established peers by an average of 7.2 percentage points in IRR, 0.34x in total value creation, and 0.18x in unrealized value.

Again, declining to allocate to emerging managers in institutional portfolios is not a neutral decision. Over a fund’s ten-year life cycle, a lack of exposure to emerging managers’ outperformance translates to approximately $72 million in foregone returns over a fund’s ten-year life cycle, as the structural misallocation to emerging managers compounds with every vintage year. At $500 million, returns ceded grows to $360 million.

According to the Colibrí’s Institute, emerging managers outperform because they have to build their firms and portfolios under constraints (smaller check sizes, broader sector exposure, earlier stage focus, and measured capital pacing), which translate into discipline.

The Colibrí Institute extends the NexTier framework, which defines emerging managers as firms excluded from traditional institutional search processes, reinforcing that “emerging” is a structural category, not a demographic label.

Importantly, emerging managers’ outperformance is amplified when multiple definitional criteria apply. Funds meeting three or more criteria show an IRR advantage of nearly 14 percentage points and significantly higher unrealized portfolio value. The upshot is that multiple constraints force greater discipline.

Takeaway: LPs are leaving money on the table by not allocating to emerging managers. Institutional screening processes systematically exclude alpha-generating talent based on criteria such as size, track record length, and brand recognition that do not reliably predict performance.

2. 90.5% of vintage years saw diverse managers outperform the benchmark

From 1998 through September 2024, the NAIC Private Equity Index, which represents diverse-owned PE funds and was compiled by GCM Grosvenor and audited by KPMG, recorded a net IRR of 16.0%, net TVPI of 1.6x, and DPI of 0.65x. Over the same period, the Burgiss private equity median benchmark produced a net IRR of 9.0%. The NAIC index outperformed the Burgiss median in 90.5% of vintage years studied. The NAIC index also outperformed the Burgiss median in 90.5% of vintage years, posting a median TVPI of 1.62x versus the Burgiss median of 1.31x.

Consistency matters as much as magnitude. The NAIC Private Equity Index produced first- or second-quartile performance roughly 66.1% of the time, and top-quartile performance approximately 35.7% of the time. It outperformed the Burgiss upper quartile in 8 of 21 vintage periods, a result that indicates structural edge, not cyclical luck. The edge is defensible, as the NAIC index spans 25-plus years of vintage data and is among the most rigorous performance analyses available for any segment of the private markets.

Takeaway: Consistency is the hallmark of a structural advantage. Diverse PE and VC funds have outperformed the industry median in the majority of years studied; this should inform how LPs think about portfolio construction.

3. 1.7x TVM vs. 1.4x benchmark

The StepStone Group’s proprietary analysis of diverse PE and VC managers found they produced a net total value multiple (TVM) of 1.7x, compared to a benchmark median of 1.4x over the same vintages. Outperformance was most pronounced among smaller diverse managers raising under $2 billion, where the TVM reached 1.8x. At the $2B AUM level, diverse managers’ net IRR exceeded 21%.

The concentration of outperformance at smaller fund sizes, where most diverse managers operate, is a feature of how their portfolio configuration. As the Colibrí Institute’s research confirms, emerging managers operating with disciplined check sizes ($5–15M) are more likely to outperform, as funds that attempt to compete with established platforms on check size often dilute value. Diverse and emerging managers lack the syndicate relationships and reputational capital that allow large platforms to justify concentrated positions. This constraint becomes the competitive advantage via disciplined, differentiated deployment into investment opportunities.

StepStone’s data presents a direct challenge to LPs who apply minimum AUM thresholds that effectively exclude “the long tail of compelling small-cap diverse funds” generating meaningful alpha.

Takeaway: The highest net multiples are concentrated in the lower end of the market where many diverse emerging managers operate and where institutional capital is most scarce. Under-allocating to diverse managers risks forfeiting enhanced returns.

4. ~30% exclusive deal flow

BCG and Cambridge Associates analyzed over 84,000 PE and VC deals and found that “approximately 30% of deals completed exclusively by diverse-owned firms are not accessed by nondiverse firms.” This exclusive deal flow represents 7% of all private market deals and provides a genuine portfolio diversification advantage that cannot be replicated through capital alone.

The mechanism behind this data is worth understanding. Diverse managers, by virtue of their networks, geographies, and lived experiences are connected to founder and operator communities that nondiverse managers have little or no access to.

In the deals that BCG and Cambridge Associates analyzed, they found that 35% of nondiverse PE and VC firms have not coinvested with diverse asset managers. This is because asset managers tend to invest with networks of professionals who have similar backgrounds.

LPs aiming to mitigate sector, manager or deal concentration risks in their portfolios can turn to diverse managers’ differentiated, uncorrelated deal flow, particularly as the opportunity to do so expands. From 2018 to 2022, according to BCG and Cambridge Associates, the value of private market deals led by diverse PE and VC firms grew at a 25% CAGR from $33B to ~$80B, almost twice the growth rate of nondiverse firms’ deal value. Over the same period, diverse firms’ deal count grew at a 14% CAGR versus nondiverse firms’ 12% CAGR. As diverse managers increase their share of transactions, more of the market becomes exclusively accessible through them.

Takeaway: diverse managers have access to high-quality, uncorrelated deal flow. For LPs who rely exclusively on nondiverse managers, the implication is that no amount of capital will provide access to the 7% of deals that diverse managers see exclusively. As the diverse manager universe grows and its deal share expands, the cost of non-participation compounds

5. 40% less first-time capital raised

StepStone points to research that indicates that minority managers have a harder time meeting first-time fund fundraising goals, raising about 40% less capital. Additionally, minority managers’ fundraising success for raising follow-on funds is three times more sensitive to past performance than that for comparable nondiverse managers.

Diverse managers are raising less capital than their performance warrants. The same track record buys less with institutional LPs if the manager is diverse. Again, the causes are here are structural. StepStone notes that diverse managers: 1) are subject to implicit bias, which is difficult to identify or overcome; 2) don’t always have access to the conventional funding avenues nondiverse managers have access to; 3) don’t always have traditional finance backgrounds and institutions struggle to see how nontraditional backgrounds enable investing acumen. The Colibrí Institute notes:

Where emerging manager programs exist, they are often designed in response to political, stakeholder, or diversity-related considerations rather than grounded explicitly in performance objectives. This framing creates a predictable failure mode in which fund managers are selected based on identity-adjacent proxies rather than capability, and the programs they manage are often held to different standards than their core portfolios. When underperformance follows, programs are eliminated, and the entire category is blamed, reinforcing the belief that “emerging” equals “concession” rather than “alpha.”

The solution is to evaluate diverse managers on the same standards, with the same rigor, applied equally to nondiverse managers. The data shows that diverse managers are meeting and raising the bar.

Takeaway: Current LP frameworks for evaluating diverse and emerging managers are often structurally flawed. Allocating capital to diverse managers should be understood as an exercise in fiduciary responsibility rather than a social gesture. Failing to engage in that exercise risks compounding performance or access to diverse and emerging funds as managers as establish reputation and desirability among institutional investors.

In 2021, the Knight Foundation found that 98% percent of all assets across every asset class in the US are managed by nondiverse asset management teams. If we drill down into private equity and venture capital, the imbalance persists: minority-led firms comprise only 5.1% of firms and 4.5% of assets. Women-led firms represent only 7.2% of firms and 1.6% of assets.1

On the surface, this disparity may appear to be a workforce representation issue, important, but separate from the question of investment returns. The performance data tells a different story. Allocating capital to emerging managers in institutional portfolios is not a social gesture; it is an exercise in fiduciary responsibility; failing to engage in that exercise risks data driven and proven underperformance and losing compounding performance that cannot be recovered.

Across various research and studies, the conclusion that diverse and emerging managers consistently outperform their established counterparts on standard financial metrics takes form. The Colibrí Institute’s analysis of 2,471 venture funds, for example, frames the stakes with precision: emerging managers outperform established peers on IRR by an average of 7.2 percentage points.2

In this article, we unpack some of the evidence behind diverse managers’ performance, interpret what it means for LPs and their fiduciary obligations, and explain why High Street Equity Partners (HSEP), an emerging fund focused on seed-stage venture investing across emerging innovation hubs, fits the pattern of leading performance in the data and represents a solution for alpha-hungry LPs.

Key Findings at a Glance

• Emerging managers’ IRR beat established peers’ by 7.2 percentage points, which translates to $72M in foregone returns on a $100M VC allocation.

• The NAIC index of diverse PE managers beat the Burgiss median in 90.5% of vintage years from 1998 – 20243

• Diverse managers’ net total value multiple (TVM) of 1.7x outperforms the benchmark median of 1.4x; for funds under $2B, diverse managers’ net TVM rises to 1.8x4

• ~30% of transactions by diverse-owned firms are not accessed by nondiverse firms, representing exclusive deal flow5

• Minority managers raise ~40% less capital for first-time funds; when raising follow-ons, their fundraising success is 3x more sensitive to past performance6

Charts

A clean deck, a tight narrative, and thirty minutes of high energy.

We understand why… That is the most visible part of the process.

But it’s not what we are actually evaluating.

At High Street Equity Partners, we’re not underwriting a presentation. We’re underwriting a person and team. The pitch is just a point in the process. What matters is how a founder operates across time, across pressure, and across repeated meetings.

Trust is not built in a single meeting… It grows, and we’re paying close attention to how that plays out.

From First Impression to Pattern Recognition

Over the past several years, we’ve met thousands of founders across different environments and life cycles. Formal pitch meetings, quick introduction calls, coffee conversations, even parties while traveling.

Early on, our process was more intuitive. We reacted to energy, narrative, and first impressions.

Over time, that changed.

We started documenting interactions, documenting case studies, and comparing outcomes.

We were looking at where our early impressions were right and where they were wrong.

What emerged was a shift from instinct to pattern recognition.

Not patterns based on background or pedigree, but patterns in behavior.

How founders communicate
How they make decisions
How they follow through
How they show up when things are unclear

That became the foundation of how we evaluate.

Founder Archetypes We See Often

There’s no single profile of a successful founder. But there are recurring archetypes in how founders show up and operate.

Most founders sit somewhere across these patterns.

1. The Operator

Close to execution, close to the customer, and grounded in reality.

They can walk through their decisions they made last week, why they made them, and what changed as a result.

Where they stand out: Strong execution discipline
Where they struggle: Sometimes they under communicate their vision or think too narrowly

2. The Market Builder

Customer-facing, commercially sharp, strong instinct for distribution, and demand.

They understand how buyers think and how markets move.

Where they stand out: Go-to-market clarity
Where they struggle: Narrative can get ahead of operational readiness

3. The Architect

Deep builder, product, technical, and/or systems-oriented.

They see how things should be built at a fundamental level.

Where they stand out: Depth and long-term thinking
Where they struggle: Translating complex ideas into simple and clear communication

4. The Narrative-Driven Founder

High conviction, strong communicator, can attract people, capital, and attention early.

They know how to tell a story that impresses.

Where they stand out: Vision and momentum
Where they struggle: Execution does not always keep pace with narrative

5. The Bottleneck Founder

Highly driven and carries the company through sheer energy.

They are involved in everything.

Where they stand out: Ownership and intensity
Where they struggle: Scaling beyond themselves

These are not labels, they are patterns.

What matters is not which archetype you are. What matters is how your strengths and risks show up over time.

The Signals That Actually Matter

Across all archetypes, a consistent set of signals separate top founders.

Proactiveness

The strongest founders do not wait for direction.

They anticipate, they move early, and they close loops without being asked.

You can see this between meetings, not during them.

Clarity Under Pressure

Anyone can sound clear when rehearsed.

What we look for is how founders communicate when the question is unexpected or the answer is not obvious.

Clarity in these moments is a strong indicator of how they think.

Feedback Adjustment

Strong founders listen carefully, but they do not lean on consensus.

They take in inputs, process them, and still maintain conviction where it matters.

That balance is difficult and it is also one of the most important signals.

Self-Awareness

Founders who understand their own gaps move faster.

They hire better, they delegate earlier, and they avoid predictable mistakes.

Lack of self-awareness is one of the easiest risk signals we see.

Follow-Through

This is where most differentiation happens.

What was said in the meeting matters less than what happens after.

Did they do what they said they would do
Did they come back with better answers
Did they close the loop

Consistency builds conviction quickly.

How Evaluation Actually Happens

Evaluation is not a moment, it is a sequence.

The First Interaction

We look at how founders frame problems and communicate their thinking.

It’s about clarity.

Deeper Conversations

We like to move beyond the narrative.

We ask questions that require real-time thinking.

We look at how founders handle uncertainty, pushback, and fill in incomplete information.

Diligence

We examine their underlying work.

• Decisions over time

• Data and materials created

• Product and operational knowledge

We are triangulating whether execution matches the narrative.

Between the Meetings

This is often the most important layer.

How founders communicate updates
How they respond to requests
How quickly they act

Patterns during this phase tend to be more predictive than anything said in a pitch.

What Founders Often Misunderstand

Many founders over-optimize for the meeting.

We optimize for everything around it.

We care about:

• How you respond when you do not have the answer

• How you communicate when things are not going well

• How you incorporate feedback across time

• How you operate when no one is watching

This is less visible, but far more telling.

Bringing Structure to a Human Process

Founder evaluation will always have a human element. It is not purely quantitative.

At the same time, we have worked to make our process more structured.

We have built internal frameworks that track:

• Behavioral signals

• Execution patterns

• Communication consistency

• Decision-making quality

We also use an AI Agent “Deborah” to organize and synthesize information across interactions.

Not to make decisions, but to surface patterns we might otherwise miss.

The goal is not to remove judgment, it is to make judgment more consistent.

Final Thought

There is no perfect founder profile.

We have seen different personalities, backgrounds, and styles all produce strong outcomes.

What consistently stands out is not the pitch.

It is the pattern.

Founders who:

• Show up consistently,

• Communicate clearly,

• Execute between conversations,

• And improve over time

Those are the founders that build trust. And ultimately, that is what we are underwriting.

Written by George Darden, HSEP Venture Partner

AI changes that.

Not because “AI is magic,” but because early architectural choices now say much more about whether a company can become a real business. In traditional SaaS, a rough early architecture was often forgivable. In AI-native companies, the design pattern often is the business model. It determines product reliability, gross margin, defensibility, regulatory risk, and whether the company is actually automating work or just demoing intelligence.

At pre-seed, founders are still selling a vision. Revenue is limited. The product is often part prototype, part promise. In that environment, the right question is not whether the AI sounds smart. The right question is whether the company has chosen an AI design pattern that can survive contact with real customers, real workflows, and real governance.

From the lens of someone who spent 20 years building back-office automation, role-based software, and audit trails: the best AI startups do not look like science projects. They look like durable workflow companies with a probabilistic reasoning layer inserted into the middle.

The First Principle: Architecture is Now the Thesis

In traditional software, architecture answered questions like where the business logic lives, how systems communicate, and how failure is handled. In AI-native companies, those questions still matter, but there is a new variable: some of the “logic” is no longer explicitly coded. It is generated by a model, influenced by prompts, context, tools, and orchestration.

When you invest in a pre-seed AI company, you are backing an answer to this fundamental question: Where can probabilistic software safely and repeatedly create value inside a real workflow? That answer shows up in the company’s primary AI pattern. Here are the five structures that matter most right now.

1. Retrieval-Augmented Generation (RAG) / Knowledge Copilot

• The Concept: Find the right context, then answer. The system retrieves relevant information from internal documents before generating a response.

• Why Investors Should Like It: Grounded outputs, easier updating than retraining, and a natural path to enterprise governance.

• The Red Flag: Weak retrieval masked by a polished UI. If they have no differentiation beyond “chat over documents” and lack a trust/citation layer, it’s a fragile demo.

2. Documentation Copilot

• The Concept: Capture → Summarize → Structure → Human Approves. Think transcripts turning into CRM updates or clinical notes.

• Why Investors Should Like It: Immediate ROI through time savings, incredibly easy user adoption (it drafts instead of deciding), and low risk because human accountability remains intact.

• The Red Flag: Draft quality that collapses in edge cases or a lack of real workflow insertion.

3. Workflow / State Machine AI (The Most Underappreciated)

• The Concept: A deterministic workflow with probabilistic steps. AI sits inside a defined process (classify, extract, route, approve) and performs bounded tasks.

• Why Investors Should Like It: Production reliability. This is often the highest-quality pattern in the enterprise because it mirrors how organizations actually operate, allowing for auditable compliance.

• The Red Flag: Startups calling themselves “agentic” when they are just a thin wrapper over a basic script, or systems with weak exception handling.

4. Agentic Tool Use

• The Concept: Understand goal → Choose tools → Act. The model plans and iterates across tasks spanning multiple systems.

• Why Investors Should Like It: Massive leverage and the potential for incredibly sticky product behavior in operations or research.

• The Red Flag: Too much autonomy too early. Demos that only work on the “happy path” without guardrails, boundaries, or clear failure behaviors easily become expensive theater.

5. Human-in-the-Loop Decision Support

• The Concept: AI assists, humans own the outcome. The AI recommends or scores, but a human makes the final call.

• Why Investors Should Like It: This is the only practical adoption path for high-risk markets (HealthTech, FinTech, Cybersecurity). It creates a massive data flywheel from user corrections.

• The Red Flag: Human review that is so heavy it destroys efficiency, or a lack of clarity on what triggers an escalation.

The Biggest Board-Level Mistake

A board member often sees a smooth demo and concludes the company has “great AI.” That is usually the wrong conclusion. A great demo often just means the prompt was strong, the example was curated, and the model performed well on a narrow case.

It does not tell you whether the company has solved data quality, permissions, auditability, latency, or regulatory constraints. In enterprise SaaS, the software people buy is not the same as the software that survives deployment.

Don’t ask, “How smart is the model?” Ask, “What design pattern is carrying the value, and can that pattern survive production reality?”

Spotting the Winners vs. The Science Projects

What Good Pre-Seed Companies Get Right

1. They choose a narrow insertion point: Specificity is a feature. They don’t try to solve “work”; they summarize clinical visits or extract contract terms.

2. They design for control: They understand AI is a system. Validation, role-based access, and fallback behaviors matter more than the LLM they use.

3. They know where the human belongs: They have a clear answer for what the AI should do alone, what it should draft, and what requires human escalation.

What Weak Companies Get Wrong (The Red Flags)

• “We’re building a horizontal AI employee.” (Translation: We haven’t found a wedge or a specific buyer.)

• “Our moat is the model.” (Translation: We don’t understand that true moats come from workflow ownership, proprietary data exhaust, and deep system integration.)

• “The agent can do anything.” (Translation: The product is unpredictable and un-deployable in a strict enterprise setting.)

A 6-Question Diligence Framework for Boards

When evaluating an AI-native startup, run this checklist:

1. What is the primary AI pattern? Force them to describe the architecture in plain English.

2. Where is the value actually created? Look for measurable workflow improvements (time, cost, compliance), not just “the user gets an answer.”

3. Where does failure happen? Mature founders know exactly how their system breaks and how customers correct it.

4. What is the control layer? How are permissions, logging, and validation rules handled?

5. What improves with scale? Does the product get better because of reviewer feedback and workflow coverage, or are they just waiting for OpenAI to release a better model?

6. Is the chosen pattern right for the industry? A fully autonomous agent in a highly regulated environment is the wrong product. A modest documentation copilot in that same environment is a massive business.

Final Thought

If I were advising a board or writing a first check into a pre-seed AI company today, I would ask one question over and over:

Why is this particular AI pattern the right one for this workflow, this buyer, and this industry—and what happens when it fails?

A founder with a serious answer is worth spending time on. A founder without one is probably still selling a concept. And in this market, concepts are abundant. Durable design is rare.

• TAM

• Traction

• Technology

• Team

• Terms

Not a checklist. A lens.

TAM

How big is the problem? Is it a must-solve or nice-to-solve? Are you obsessed with solving the problem? Not casually interested. Not indifferent. Obsessed.

We ask:

• What’s the moat?

• What gets stronger as the company scales?

• Can this realistically become a venture-scale outcome in 5 years?

Traction

Traction is evidence, not hype.

At the earliest stages, founders are operating on hypotheses. The best founders turn those early assumptions into repeatable insights backed by customer behavior.

We look for:

• Clear ICP understanding

• Paid customer validation

• Repeatable sales motion

• Strong retention and referrals

• Healthy pipeline relative to valuation

• Capital efficiency and disciplined runway management

Real product-market fit usually looks boring before it looks explosive.

Customers come back. They tell other customers. Growth compounds organically.

Technology

Technology should create leverage.

We’re less interested in “tech-enabled” and more interested in true advantage:

• Proprietary workflows

• Defensible product infrastructure

• Scalability

• Product roadmap clarity

The best founders know exactly why their product becomes harder to compete with over time.

Team

Early-stage investing is still heavily about people.

The founder traits:

• Resourcefulness

• Speed

• Resilience

• Clarity

• Intensity

The best founders create momentum before the business is fully figured out.

Terms

Great companies can still become bad investments at the wrong price.

We look for:

• Reasonable valuations

• Healthy cap tables

• Strong lead investors

• Clear understanding of revenue-to-valuation multiples

• Space for HSEP to add real value

At the end of the day, the 5 Ts help us answer one core question:

Can this founder execute through uncertainty?

Because venture outcomes rarely come from only the best storytellers. They come from founders who keep compounding insight, trust, and execution long after the pitch deck is closed.

-Mitch

We think about that story often, because the venture industry is still making Butler’s mistake. Not out of malice, usually. Out of habit — the quiet, compounding habit of expecting the builder to look a certain way, and funding accordingly.

We want to be precise about what that habit costs, because you deserve the real numbers, not encouragement dressed up as data.

The Arithmetic You’re Already Living

In 2025, U.S. startups founded entirely by women raised 1.1% of all venture capital dollars — down from an already-dismal 2.1% the year before, in a year that was otherwise a record for the industry (PitchBook, 2025). Put the two ends of that distribution side by side and the picture sharpens into something almost hard to believe: all-female teams raised a total of $3.2 billion across 794 deals, while all-male teams raised $191.1 billion across 10,048 deals (PitchBook, 2025).

Sit with that ratio. For roughly every dollar that reached an all-women founding team last year, all-male teams received about sixty. And the gap widened rather than closed: all-male founding teams raised 21% more capital than the previous year, while all-female founding teams raised 22% less (PitchBook, 2025).

It gets more pointed at the top of the market. In 2025, 124 new unicorns — privately held startups valued at $1 billion or more — emerged in the U.S. Of those, just 20 had at least one female founder, and not a single one was founded by an all-female team (PitchBook, 2025). And the earliest door, the one that’s supposed to be the most open, is closing too: the share of female founders receiving their first venture check peaked at 27.7% in 2021 and fell to 21.2% by 2025, a sign that fewer new women-led companies are getting in at all (PitchBook, 2025).

None of this is a referendum on the companies. That’s the part the numbers make unarguable.

The Returns Tell the Opposite Story

Here is what makes the funding gap not just unfair but irrational. Women-founded companies generate 78 cents of revenue for every dollar invested; male-founded companies generate 31 cents (Boston Consulting Group). The founders receiving the smallest share of capital are the ones deploying it most efficiently.

The bias doesn’t usually announce itself. It hides in the questions. Investors consistently ask men “promotion-focused” questions about growth and opportunity, while women are asked “prevention-focused” questions about risk (Harvard Business School). One founder is invited to sell a vision; the other is asked to defend against loss — and founders asked promotion questions go on to raise six times more capital (Harvard Business School). Same company, same numbers, different script. The scrutiny compounds: the market increasingly rewards founders with a “proven track record,” and because that track record has historically been male, the bar quietly rises for everyone who isn’t. So women are made to prove more with less, and then are funded less for having less — a loop that has held, by PitchBook’s count, more or less unbroken since 2008.

If you have felt that asymmetry in a room and wondered whether you imagined it, you did not. It has a citation.

What We Decided to Do About It

It would be easy to write a letter that admires the problem. This is not that letter, because admiration doesn’t move capital, and capital is the thing in short supply.

Half of our portfolio is founded by women. Not as a quota, and not as charity — as a straightforward consequence of underwriting companies on their merits in a market that systematically misprices them. When 78 cents comes back on the dollar and the check sizes stay small, the mispricing is the opportunity. We didn’t build a women-founder portfolio to make a point. We built it because the discipline of looking closely — Butler’s discipline, at the thing itself rather than the story about it — kept leading us to the same founders the rest of the market walked past.

We say the number out loud for two reasons. The first is accountability: a claim of support that can’t be measured is just sentiment, and you’ve had enough of that. The second is that we want it to be ordinary. The goal was never to be remarkable for backing women. The goal is a market where fifty percent doesn’t merit a headline because it’s simply what the returns dictate.

To You, Specifically

So this is our notice, plainly: we see the arithmetic you’re working against, and we see you working anyway.

We see the founder who walked into the room already braced for the risk questions and answered them without letting the ambition drain out of her voice. We see the one self-funding on personal savings and credit cards because the checks didn’t come, building revenue the market told her to go get before it would believe her. We see the one whose company throws off more revenue per dollar than the funds passing on her will admit, and who will build it regardless of who finally notices.

The queen was always at the center of the hive. It took someone willing to look to say so. We are proud to be among the ones looking — and prouder still of what you are building while the rest of the industry catches up to the math.

The narrative is changing, slowly, and not because it was handed to you. It’s changing because you keep making it impossible to tell any other story. We’re honored to help fund the next chapter of it.

With respect, and with conviction,

High Street Equity Partners

Every vintage year tells a story, but 2022’s is unusually clean. It sits on a fault line. The funds that closed that year were raised in the euphoria of 2021 and deployed into the reckoning of 2023 and 2024. They caught the top of the market on the way in and the bottom of it on the way out.

We can say that with some authority, because High Street Equity Partners’ inaugural fund is a 2022 vintage. We began investing in late 2022—right as the correction started—and our own results now sit inside the same cohort we’re about to map. Per Carta’s 2025 VC Fund Performance Report, which aggregates data across more than 2,500 emerging funds, our first fund ranks in roughly the top decile of its vintage, with a TVPI of 2.24x and an annualized IRR of 22.7%. For context, Carta’s data puts the 90th-percentile 2022-vintage fund at about 1.29x TVPI and 17.6% net IRR. We’re writing about this cohort as a member of it, not an observer above it. That’s the lens for everything that follows.

We mapped the 2022 pre-seed and seed cohort—the firms writing the first institutional checks into companies that, in most cases, still don’t have revenue. Roughly forty funds, spanning generalist platforms like 500 and blockchain specialists like Ripple and 15th Rock, geographically-anchored vehicles like Invest Nebraska and idea fund, and mission-specific players like SoGal and Ajim Capital. What the map makes visible is not a trend. It’s a filter—and we’re inside it.

Why the 2022 Vintage Is the One to Watch

Vintage year matters more at the earliest stages than almost anywhere else in venture, and it matters most when the macro environment shifts violently mid-cycle. A 2022 seed fund made its first bets when seed valuations were at record highs, then had to support those companies through the sharpest tightening in growth capital since 2008. The Series A that looked automatic in 2021 became a cliff. Bridge rounds and down rounds replaced markups. The companies that survived did so on fundamentals, not momentum.

This is precisely why the vintage is diagnostic. In a rising market, everyone looks like a good investor—entry discipline is invisible when the next round always clears. It’s the down-cycle vintages that separate underwriting from luck. And the Carta data shows the 2022 cohort sorting hard: across 564 funds in the vintage, the spread between the median and the top decile is stark, with the 50th-percentile fund sitting near flat on paper while the top decile is already marking meaningfully above cost. Dispersion this early isn’t noise. It tells you the cohort is being sorted by discipline rather than by a broad tailwind lifting everyone.

Our own position in that spread is the argument we care about most. We didn’t outperform because we were clever about the market—we outperformed because we deployed slowly into a market that punished everyone who didn’t. As Carta itself notes, deal activity since late 2022 normalized to roughly half its prior pace as capital shifted from rapid deployment to disciplined company-building. We started investing into exactly that reset. The constraint was the education.

What the Map Actually Shows

Three features of the 2022 pre-seed and seed cohort deserve attention—and in each, we can point to our own book as evidence rather than theory.

The first is the persistence of specialization. This is not a cohort of undifferentiated “we invest in great founders” generalists. It’s populated by funds with a defined edge—sector, geography, or founder network. Ripple Ventures and the blockchain fund carry explicit technical theses. Invest Nebraska, idea fund, and 1843 anchor to regions the coastal firms structurally underweight. SoGal, Ajim, and Kepple Africa Ventures build around founder populations and markets the standard Sand Hill funnel misses. We built the same way—concentrating narrowly in Future of Work, Future of Care, and emerging technology rather than diversifying broadly. In a scarce-capital environment, a differentiated deal funnel isn’t a marketing story. It’s the difference between proprietary access and paying up in a crowded auction.

The second is geographic dispersion. The map is not a San Francisco org chart. It reflects a structural reality we’ve written about before: outsized companies increasingly emerge outside the major coastal hubs, anchored by deep technical talent pools and regional sector specialization. A 2022 seed fund positioned in an emerging hub had a real advantage during the correction—less valuation competition on the way in, and a founder base whose burn discipline was forced rather than optional. Cheaper entry into a market that punishes waste is a structural tailwind, not a consolation prize. Our own diligence framework was built to source founders precisely in those overlooked markets, and it’s a meaningful part of why we’ve avoided the down rounds and write-offs that have hit many of our vintage peers.

The third, and least discussed, is the pre-seed-to-seed compression this cohort has had to navigate. When a fund writes the very first institutional check, its return depends entirely on the company clearing successive financing gates it does not control. For the 2022 vintage, the first of those gates—the seed-to-Series-A graduation—slammed shut for a large share of the market. How a manager behaved when a promising portfolio company couldn’t raise its A tells you more about the firm than any winner in the book. We paced capital deliberately and reserved to defend our highest-conviction companies through exactly that squeeze. The markups that followed weren’t luck; they were the reserve strategy working as designed.

The Best Practices This Vintage Rewards

If 2022 is a filter, it’s worth naming what it filters for. The patterns that produce durable early-stage performance are not new, but this cohort makes them legible—and they map almost one-to-one onto what we’ve learned running a 2022 fund in real time.

Conviction over coverage. The funds that outperform in this vintage are not the ones that made the most bets. They’re the ones that made the fewest, held them longest, and reserved enough to defend them when the market stopped cooperating. Concentration is uncomfortable in a bull market and vindicated in a bear one. We concentrated capital in our highest-conviction companies and stayed there.

Entry discipline as the whole game. At the seed stage, price paid is the single most controllable determinant of return. A 2022 fund that held its valuation discipline against 2021 comps bought itself a margin of safety no amount of later-stage brilliance can manufacture. You cannot underwrite your way out of an entry price that was wrong. We priced conservatively while peers chased inflated rounds—and the absence of down rounds in our book is the receipt.

Patience as strategy, not temperament. The historical data on breakout early-stage outcomes points to long duration—the gap from first check to real liquidity routinely runs the better part of a decade. A 2022 seed fund being judged today by a nervous LP is being judged prematurely. The managers who understand this—and who set LP expectations accordingly—will still be standing when the marks catch up to the fundamentals. Our numbers are early marks, not verdicts, and we say so plainly. But early marks in the top decile of a brutal vintage are worth paying attention to.

The Takeaway

The 2022 vintage will produce a smaller number of celebrated funds than its 2020 and 2021 predecessors, and that is exactly why it will produce better information. Boom vintages reward participation. Correction vintages reward judgment. This cohort of pre-seed and seed managers was handed the least forgiving conditions in recent memory, and the ones who navigate them well will have proven something the froth years never could: that they can pick, price, and hold under pressure.

We’re not writing this from the outside. High Street was built for a moment like this one—when discipline, conviction, and partnership matter most—and our own vintage handed us the test on day one. The firms worth watching from 2022 aren’t the ones with the flashiest logos on the deck. They’re the ones whose discipline was forged precisely when discipline was most expensive—and most rare. We intend to be one of them, and the early data says we’re on our way.