Because of that, we’re compiling an AI playbook for folks working in the Customer Education space as a free resource to help empower the community. We want to crowdsource your ideas and real-world solutions to create a practical playbook!

What we’re looking for

Pick whatever fits your experience, you don’t need to submit all of these! No idea is too big or small for the playbook.

• A quick tip. Even one sentence counts. Something useful you picked up along the way.

• A skill you’ve built or customized. What it does and how you made it your own.

• A workflow or automation. Anything that’s saved you real time, manual or automated.

Or, anything else that you’ve found helpful! Be creative :)

How to submit

Learn more about the info we’re hoping to gather and submit your ideas at this form.

We’re taking submissions through September 1, and will share out the playbook September 10 (I’ll be sure to include it in a newsletter!)

Looking forward to seeing what you have to share-

Christie & Erika

The information access isn’t new. Finding information hasn’t been the hard part for decades now.

“AI is going to change learning”

I keep seeing this claim, or some version of it, popping up on my education-skewed LinkedIn feed. The claims are always big: AI is going to revolutionize education. AI is going to transform how people learn. Perhaps you could argue I’m being a little pedantic here, but words are important: people are confusing access to information with learning.

I won’t argue against the fact that AI has changed how we search for and retrieve information. Even if you’re not turning to ChatGPT or Claude, you haven’t had much choice in the matter: Google shows you an AI overview with every search. And plenty of people are going straight to ChatGPT or Claude to ask a chatbot to explain a concept in plain language, get a summary of an article, or generate a list of steps for a new task.

All of that can be true, but finding information and learning something are not the same activity. I talked about this with regard to MOOCs in my post about cohort learning. So why do I keep yammering about this distinction? Because I think when we confuse information access with learning, we start making some questionable decisions related to learning design.

So what is learning, then?

Learning is what happens when new information gets integrated into your existing mental models and you can apply it when needed. It involves practice, struggle, feedback, connection to prior knowledge, and often other humans. How do we know when someone has learned something? Usually, we should look for behavior change in learners: when a learner can actually do something with what they’ve learned that they couldn’t do before.

Think about it this way: I can easily look up where Dayton, Ohio is. That does not mean that one month later I’ll be able to locate it on a blank map. It doesn’t mean I could tell somewhere else where it is accurately, or that I could drive there. The gap between “I found the information” and “I have learned information” is in its stickiness and its application. Learning is built through practice, through failure (maybe lots, if the topic is complex), through feedback and reflection, and sometimes through watching someone else do it and asking questions.

We know from cognitive load theory that learners can only hold so much in working memory at once. We know from the Community of Inquiry framework that social presence matters for deep learning. We know from decades of research on retrieval practice that effortfully recalling information is what cements it, not just reading it again (or having a chatbot re-explain it).

AI doesn’t change any of that. The cognitive architecture of the human brain did not update when ChatGPT launched.

Okay, so what does change?

I’m not saying AI doesn’t affect instructional design at all. It does, just in different places (some of which I’ve talked about before).

Here’s what I think is actually helpful for instructional designers to consider:

How we get information in front of learners. The first places people turn to for information aren’t the same anymore. Someone can ask a chatbot a question and get a decent answer instantly (or what they might think is a decent answer). Even if they turn to Google, they’re going to be presented with a chatbot answer first. Especially in customer education, it means our knowledge bases and courses are competing with a new channel. We need to think harder about when and why a learner would come to our designed experience instead vs. relying on an LLM, in addition to ensuring that LLMs can find the correct information about the products we’re supporting.

How we think about assessment. This is something that I think we’re still figuring out the right answer to. AI is definitely pushing us to better assessment design, but sometimes the basics that are worthwhile to assess learners on can be easily produced by AI, and I’m not sure there’s a foolproof solution for that yet.

How we support just-in-time performance. AI handles this pretty well: I’ve fed it screenshots of products I was struggling with and found the answers I needed much more easily than I would have trying to dig through support docs. For instructional designers, this is where defining the actual knowledge gap is useful. If the root problem is simple information retrieval, it’s worth asking whether some things we’ve been training people on would be better served by a well-designed job aid or AI assistant instead.

The real question

The core work of instructional design is not figuring out how to present information. The real work is in understanding your learner, defining what they need to be able to do, and designing experiences that get them there.

How do we change behavior? How do we design learning that persists beyond a learner exiting a course? I think these are more interesting questions (albeit less trendy) than “how will AI change learning?” They’re also not new questions at all.

If your learning experience is just content delivery with a quiz at the end, maybe AI will disrupt it.

If you’re designing for practice, for application, for connection, for the messy human process of getting better at something? I think there’s still value there that AI isn’t currently positioned to uproot.

Do you think this a worthwhile distinction? Let me know your reaction in the comments!

Learning reads and resources

Before I go, there are a couple of learning-related reads I’ve found interesting lately that I wanted to share.

• This LinkedIn post by Dr. Philippa Hardman about designing learning experiences to avoid learner cognitive offloading to AI (she has a Substack article about it, too, if you’d like a deeper dive; but I found her carousel well-organized for quick consumption)

• Hooked on Learning: Why Borrowing from Addiction Science Won’t Make Better Learning Design by Dr. Heidi Kirby in the IDTX Learning & Performance Magazine (skip to p21 to read her article): really enjoying this magazine as a whole, but Heidi’s piece around why influencer/content creator tricks won’t work for learning design is especially good.

• LearnWorlds AI & Instructional Design Mini-Summit: I’m not affiliated with it and I’m not sure how it will turn out, but wanted to share this event happening September 10 that piqued my interest. It also has a self-paced certification available for participants.

Thanks for reading, and see you next time. 👋🏻

A recent Stanford piece, in reviewing how humans learn social conventions, discusses the role of social interactions in how humans learn. How we learn is fundamentally different from how a machine “learns,” and the difference here can change how you look at 1) the actual impact of information access on learning and 2) how you think about designing instruction overall.

Humans don’t just rely on raw information to learn

The premise behind large language models’ “learning process” is basically brute force: gather as much data as possible (legally or not lol), run it through probability models, and calculate an optimal answer. When people tout greater information access as the answer to better education, there’s an assumption tucked underneath that this is more or less how our brains work, too.1

Douglas Guilbeault, a professor at Stanford’s Graduate School of Business, argues our brains don’t operate anything like this. He says that when you watch how people actually behave in the world, most of the time we’re making decisions with very little data and no real way to get more. We’re distracted, we’re limited, we can only hold so much in our heads at once.

The economist Herbert Simon called this“satisficing.” Instead of processing everything, we gather just enough information to make an intuitive leap to a good-enough answer, and then we act. Some of the ways we gather that information might be through reading or research. But another way we do that is through watching others, observing what works and doesn’t work, and copying what they do— something that requires the presence of other humans.

Think about the very first things humans learn when they’re brand new

Where this is the most obvious to me, and what the Stanford article uses as an example, is when you think about how most brand-new humans learn. A toddler can’t consult a dictionary or Wikipedia. They can’t sit down and process a textbook on grammar. And yet, just by listening to the people around them, they slowly absorb the rules of language and move from baby gibberish to making sentences.

What Guilbeault and his colleagues found is that adults learn social norms in a similar way. In their experiments, groups of people reached consensus and figured out unspoken rules with very little information, mostly by observing each other and making inferences about what everyone else was probably thinking. We find common ground across an information gap, and we do it with a few simple rules rather than an exhaustive calculation.

That leap, from “I have no idea what’s going on” to “oh, I get it now,” is something Guilbeault points out is still mysterious, even to researchers. And it’s the thing he says is largely missing from how we talk about AI. Machines process large amounts of information to identify and recognize patterns. Humans watch others, satisfice, infer, and leap.

What does this have to do with social learning?

If a large part of how we learn is by making inferences about other people, then learning with other people isn’t something we can just ignore and expect positive results.

I’ve written before that learning is relational: our backgrounds, our experiences, and who we learn alongside all shape what and how we learn. The Community of Inquiry framework (common in higher ed pedagogy) says essentially the same thing: social presence supports deeper cognitive thinking.

The Stanford research adds more context for why. We are, at a cognitive level, wired to learn by reading the room, to coordinate meaning with the people around us. Take away the other people, and you’ve removed the conditions in which our brains have evolved to learn.

Simply chatting with AI doesn’t structurally can’t offer the same thing. It can hand you an optimal answer. It can’t be the other mind you’re making inferences about, the peer whose confusion mirrors your own, the person whose offhand comment suddenly makes an abstract idea click. While AI chats can be somewhat adjacent to this, they still rely on our individual inputs too much to have the same impact as learning from other people’s interactions with each other and the content itself.

What does this mean for instructional design?

If you’re designing good instruction, whether that’s in customer education or in higher ed, you’re not just teaching facts or technical steps, but demonstrating the meaning and value the information has in the greater context of the world.

That meaning and value—how the information fits into how you think about things day to day, how it shows up in real-world problems, and how you approach the solutions—is the kind of knowledge the research says we acquire by observing and coordinating with others, not by absorbing it alone.

It’s one reason cohort-based learning tends to have better retention and success rates than self-paced, MOOC-style courses. It offers people more than simply information. They have opportunities to grapple with. new ideas together, make meaning out of each other’s questions and experiences in a way that’s difficult to replicate otherwise. Understanding becomes fuller and more enriched from multiple inputs, even when learning about topics in which we already have confidence.

I saw this just yesterday after I held an instructional design-themed lunch and learn focused on needs analyses with my team. I asked one of my teammates (Erika Hickey, who has good LinkedIn content you should follow) for feedback, and she said, “I thought it was helpful! I feel pretty confident in needs assessments BUT I found the real examples the most helpful and hearing how others were thinking about it too!”

Designing for how brains actually work

As access to information through AI continues to grow, we should be cautious not to conflate the speed and availability of information with the potential for humans to learn more quickly and easily. Although AI may have introduced different ways that we can access and manipulate information, it hasn’t changed the fundamentals how we learn. If a large part of how humans learn is by inference, coordination, and intuitive leaps made in the presence of other people, then our job as designers is to build for that, not to optimize it away.

Thanks, as always, for reading! I hope to be back in your inboxes soon with my learning reads for the summer.

That’s me and using Zoom Meetings for workshops that require pre-registration. In this article, I’m going to tell you why this has been such an issue for me, and how I used Claude to help me solve it.

Here’s to hoping this can help someone directly, but if not, I hope perhaps it gives you some inspiration for how you may use AI to automate what does not bring you joy.

Defining the Zoom Meetings problem

Hokay,1 I’m going to try to explain this in as few words as possible.

There are two primary tools that Zoom offers for online gatherings: Zoom Meetings, and Zoom Webinars.

Zoom Meetings allows all attendees to participate on camera, via mic, chat, etc, all very easily. You can set up a registration page for a Zoom meeting, but:

• The confirmation emails aren’t as nice as the ones sent out by Zoom Webinars. They don’t include a real calendar invite- you have to physically add the event to your calendar (an option that is smaller and harder to find) instead of accepting a regular calendar invite. This is unnecessary friction that can prevent people from joining a session simply because they forget to add it to their calendar.

• On the back end, registration info collected doesn’t play well with other tools (so the data from that registration is hard to get out of there into CRMs and such).

Zoom Webinars has better options regarding registration and calendar invites, so there’s less friction in getting people to a session and tracking the data afterwards. However, the attendees are largely stuck behind a screen: they can chat if you allow it, but coming on camera or off mute isn’t very easy. It almost promotes disengagement and lecture-style sessions.

My personal struggle with Zoom Meetings

I want to create workshops where people are actively engaged and asking questions, taking work/practice breaks, maybe screen sharing, and are not stuck to the chat. So I turn to Zoom Meetings.

The primary workshop session I’ve been working on this year is a form of scaled (or group) onboarding. The entire scaled onboarding process, in this example, consists of two live 1-hour sessions and completing an on-demand certification.

When I started all this, I had a separate Zoom registration page for each session. It was difficult to communicate the full onboarding journey to participants with this set up; a lot of people registered for the first session without knowing about the second. Add this to the “no calendar invite” in the registration emails problem and I had two main issues: a conversion issue (attendees registering but not showing up) and retention issue (attendees coming to the first session, but not the second).

My original, painfully manual solution

I had a hunch that the lack of calendar invite was causing some folks to no-show to workshops. So, I created calendar invites on my own calendar (with the guest list hidden), checked the registration logs excessively, and added anyone who registered to the calendar invite.

I cannot emphasize enough how super manual and annoying this was. If I forgot to check the registration page for a bit, or if someone registered for a session far into the future and I didn’t notice, the calendar invite could be really delayed. To the invitee, it probably seemed super random, and may have been less effective.

Doing this helped me solve the conversion problem, but not the retention problem. I had a hunch that the retention problem lived in the whole “separate registration pages for each session” set up- or people not realizing that full onboarding consists of two sessions.

The solution Claude helped me create

I turned to Claude with my problem. Here’s literally what I asked (I’m even leaving my typos in there for your reading pleasure). Note this was in a Claude Project where I had attached documents for a bit more context.

Note also that I had literally just started using Claude and I didn’t really understand the Claude terms for stuff at all.

I’m exploring options to help me automate registration and calendar invites for a scaled onboarding cohort. the way this works is I host a new scaled onboarding cohort each month. The onboarding cohort attends 2 live 1-hour zoom sessions. I need to track new registrants all from the same form and add them to the appropriate calendar invites (the month cohort that they choose on the form).

assume registrants are stored in a google sheet. ist here an artifact or cowork flow that could help me with this ? what would you recommend?

Welcome to my brain, y’all. Here’s the solution that I created with Claude (no we did not need an artifact or cowork lol, an artifact is especially wrong here wow):2

1. One Form to Rule Them All: I now have one Google Form that acts as registration for both sessions. Participants select which month they’d like to join the scaled onboarding (specific dates are listed in the form).

2. Google Apps Script runs the show: Claude wrote a Google Apps script that I’ve connected to the form response spreadsheet. It runs every 30 minutes. It finds new registrants.

3. Everyone gets a calendar invite for both sessions right away: The Google Apps Script then finds the calendar events on my Google Calendar that the registrant has signed up for, and adds them to those calendar events (thus sending the registrant calendar invites).

4. Everyone also gets a confirmation email: The Apps Script triggers a confirmation email to registrant from me.

5. I get a heads up if something fails: If for some reason this doesn’t work, the script sends me an email detailing the error so I can manually add the registrant to the right calendar events/follow up.

To be clear here: the Google Apps Script Claude wrote is managing everything here. Claude is not involved in any of the automation described. It just helped me build the solution.

What I need to maintain manually with this solution:

1. Consistently named calendar events. I have recurring calendar events in my Google Calendar and recurring Zoom Meetings linked to these, so this is not a large manual hurdle. I have marked the events so that guests cannot see others on the list.

2. Updating the form: I need to update the form monthly to get rid of old cohort options and add new months. Could probably automate this but don’t really care enough for now because it’s nbd.

3. Slight script tweaks: Depending on some specific stuff around when the workshops fall, I might need to adjust a line in the Google Apps Script monthly after making that form update.

The TL;DR here is that I went from checking and manually doing stuff DAILY to now only having to do stuff monthly, and my attendees have a clearer process (the Google Form is much more customizable for registration than the Zoom registration and I can explain the whole two-session cohort process there).

Do you want to read more about how all of this works? I’ve summarized it in a Google Doc linked below (with all specific company details stripped out).

Check out the details.

So does it work?

I tested it a bunch with test email addresses - I was able to get some successful tests and also saw how it handled errors.

It’s now live for one set of onboarding sessions that I run, and it’s worked for everyone who has registered so far!

It’s only been live for about a week, so the jury’s still out on whether or not this is going to meaningfully contribute to solving my conversion or retention problems.

BUT, I have taken a repetitive task off of my plate. Working on back-end data stuff next (tracking who attended/didn’t attend and communicating that to the necessary systems).

Did this help you at all?

Even if it was just in how you approach thinking about what AI can help you with?

Let me know in the comments! Or, tell me about how you solved similar problems (aka tackling tasks that do not bring you joy). Another alternative: ask me questions.

I often find a lot of similarities when reflecting on my teaching in higher ed and designing learning for a B2B SaaS company, and here we are again. In thinking about the past few months teaching college and scaling onboarding workshops, I’m remembering that so much of education has nothing to do with the content itself at all.

But what about the learning objectives

This isn’t a new revelation or something. In fact, I wrote about it (indirectly) on my teaching blog back when I was teaching high school over a decade ago.

I’m handing a student a quiz about the French Revolution, but she complains to me that she hasn’t eaten in the past 24 hours due to conflict with her family.

I see it again now:

In my online college class, a student disappears for a while, eventually letting me know they’ve been clocking 50 hour weeks on top of their full-time course load.

Customers miss one of their two onboarding sessions because they are juggling multiple commitments and just…forget that they signed up for a thing.

While all of these circumstances carry differing levels of urgency and intervention on my part, they all reflect a similar issue: we can’t teach anything if our learner can’t show up, for whatever reason.

You should be designing and evaluating learner nudges as part of your curriculum design

Ok, so, accounting for learner environment is technically part of the earlier phases of ADDIE. So this is not Brand New Information. But I’ve rarely seen ADDIE (or SAM, or your learning design model du jour) really consider this in practice, in an iterative and ongoing fashion.

I think part of the reason for this is because a lot of what people think of as “instructional design” centers around designing compliance training, which assumes some kind of built-in incentive for a learner to complete your course. I also think it’s because a lot of instructional designers have no control over what instructors or facilitators might do with a course once it’s released into the wild (see: higher ed).

There are also probably a lot of other reasons I’m not thinking of, feel free to add a comment and let me know. ⤵️

Leave a comment

Anyway: I think it’s worthwhile for instructional designers to design, test, and iterate learner nudges as part of your curriculum design. Let’s dig into some examples here of what I mean (and maybe you’ll see why I think some IDs find this difficult and abandon ship).

I’m going to provide an example from customer education in SaaS and an example from teaching higher education. Read them both or skip to the one that’s more interesting/applicable to you, it’s okay, I won’t know.

What does designing learner nudges look like? A customer ed example.

I’ve been working on a scaled onboarding workshop in my current role (read: lots of customers are coming to a group onboarding session vs. attending a 1:1 onboarding).

The full training consists of two live sessions in Zoom Meetings, which I’ll refer to together as one cohort:

• One targeted at account admin, focused on account set up.

• One targeted at admin + end users, focused on implementation how-to and strategy.

Note that I specified Zoom Meetings, not Zoom Webinar. I made this choice intentionally because it allows me to interact with attendees more in-session, but it changes the registration experience and is a little clunkier because of it.

How I built in iteration: I started running one cohort per month in January. At the end of each month, I’d look at the cohort data to determine: how many scaled onboarding-eligible customers joined? Are these customers showing up to one or both sessions?

Here’s a graphic I shared on LinkedIn on what the timeline has looked like:

Problems that revealed themselves across time:

1. Registrant conversion: Converting workshop registrants to workshop attendees (which is a perennial issue for anyone in this space).

2. Workshop cadence: I held onboarding session #1 in week 1 of the month, we broke for two weeks, and then met again in week 3 of the month. This was too much time in between workshops.

3. Registration woes: People were registering for the 1st session, but not the 2nd session (so a cohort retention issue). Session #1 and #2 are currently two separate registration forms.

How I adjusted/iterated to nudge learners into attendance:

1. The conversion problem: Registrants were sent a confirmation email from Zoom that includes a link to add it to their calendar, but did not actually add it to their calendar (this is a limitation of using Zoom Meetings). I had a hunch that registrants were not adding it to their calendars, so I created calendar invites on my own and added registrants to the invite early on. I had already been sending reminder emails to registrants the day before and continued doing this. Conversion increased to 100% attendance in April and May.

2. The cadence problem: I noticed that we had a large drop-off in attendance from session #1 to #2 with two weeks in between sessions, in addition to participants voicing concerns over the gap. I reduced this to 1 week in between sessions almost right away.

3. The retention problem: I added a Zoom poll to the end of session #1 asking folks if they wanted me to register them for session #2. I haven’t run the numbers on increased retention yet, but anecdotally noticed a larger increase in May.

This process could still be a lot smoother and less manual! I’m working on a larger fix for that so that folks register into cohorts instead of individual sessions and hoping to add some automation in for the calendar invites.

What does designing learner nudges look like? A higher ed example.

The TL;DR (too long; didn’t read) version of this is: student disappears for a while, so you ask them how they’re doing and if they need help.

That’s pretty much it. I’ll also send out a course-wide “pulse check” survey to see how stressed students are feeling if I notice large amounts of late or missing work.

The longer version:

Sometimes this can be difficult to notice when you have approximately 5 million other things happening, so I’ll oftentimes build in checkpoints for myself to do this. For instance, before the students complete the Unit 1 project in my course (approximately 4-5 weeks into it), I’ll send the same standard email outreach to most of them; while the email is the same, I send them individually. Something like this:

If there are students who have gone missing or are missing major assignments by that point, I’ll revise the email to mention the missing assignments and remind them of they can do (in my class, they can submit late without penalty).

A lot of times, it only takes one nudge to get a silent student back into the course. Showing concern and offering help can go a long way.

Build those checkpoints into your syllabus, around large projects or difficult topics, so it becomes impossible to forget. Keep your “standard email” copy somewhere so it’s easy to copy and paste when needed.

My problems were different, so my solutions were different

My solutions for college students ghosting me and my solutions for getting tech customers to show up for onboarding look different because my problems were different. My learners were different.

But each time, I built in time for myself to check in and figure out who was dropping off and how I could nudge learners back on course.

Nudges can look different! They can be reminder emails, calendar invites, or perhaps even a tweak to course navigation to nudge learners into the desired behavior. If you’re designing learning, don’t forget to think beyond your content. Consider what you can change about your learner environment and the facilitation of learning to ensure that they’re being nudged toward completion.

But Christie, you have access to way more parts of course facilitation/implementation than I do 😭

I do recognize this might be true for some! But I still want you to consider: What data do you have access to? At what points might it make sense for you to “check in” on a course and re-evaluate where learners may have hit a snag or dropped off?

Maybe that’s after a full semester of running it (in higher ed). Maybe that’s 90 days after you launch an on-demand course (in customer ed). I’m not sure what might work best for you! But: schedule a check in for yourself to evaluate where you might need some nudges.

Yes, course completion is a vanity metric. But we need to get learners there first before we can start to evaluate anything meaningful ;)

I’d love to hear what nudges or environmental changes/learner nudges have worked for you! Share them in the comments.

For those of you who may have been 1) not in the education scene in 2012 or 2) are much younger than me (I was teaching high school history at that time), MOOCs, or massive open online courses, were all the rage. Universities were doomed. MOOC platforms like edX, Coursera, and Udemy were bound to take over. Anyone could take any course they’d like, at any time!

Over a decade later, the MOOC takeover…hasn’t happened. But why? And why are most customer education academies still offering MOOC-style courses (that is, self-paced, on demand)?1

This article isn’t really about MOOCs themselves. Really, I’d like to focus on what MOOCs are missing and how we can design better asynchronous learning opportunities.

What if we could take the scalability and accessibility (to an extent) of MOOCs and make them better?

Why MOOCs suck

You could think that’s my opinion, sure. But research indicates that most course completion rates for MOOCs range between 6%-14%. That pretty objectively seems not great. Could you imagine if a school was like, hell yeah, we graduate 10% of our students?

But why are completion rates so bad? Let’s look at some characteristics common in MOOCs.

MOOCs are mostly passive content consumption

Most “learning” consists of watching videos or completing readings. If active work is involved, it’s difficult to impossible to receive clarification or help. There’s no real incentive to complete the course except out of sheer desire, but that can only take you so far if a course has like 200 videos to watch.

MOOC timelines and accountability nudges are impersonal and not truly time-bound

The platform will give you a recommended timeline and “nudge” you if you’re behind. But they’re automated nudges, not concerned outreach from a human. I spoke to someone once taking a MOOC who mentioned that the nudges made them feel stressed out- that they were trying to take their time with the content, not keeping with the “recommended” (but not required) timeframe, and the platform kept reminding them that they were behind.

MOOCs lack support systems

MOOCs may contain community forums, but the participants posting there may not be completing the course at the same time as you; it can become just another resource to read to find information. The “community” just becomes a sea of names you’ll never see again.

Additionally, oftentimes MOOCs contain small errors or inconsistencies - which can be problematic when there’s no real course point of contact, and can lead to folks being frustrated and giving up.

If you haven’t figured it out from reading my newsletter in the past, I really like learning! But I don’t think I’ve ever actually completed a MOOC (even one I paid for!). I’ve gotten either bored watching endless videos, or frustrated because I couldn’t figure out how to get around a snag in an assignment, and given up. I can’t really blame anyone who doesn’t love learning as much as I do for doing so as well.

Why cohort learning can be better

First, let’s talk about what cohort learning is, loosely. Cohort learning typically refers to a group of people going through some kind of course together. This could be anything from a traditional college classroom course that meets in person, to a 2-week bootcamp style certification course that’s largely asynchronous (but is completed in a defined two week period of time).

The biggest difference between cohort learning and traditional MOOCs are that there’s a group of students progressing through the learning together, at the same time. Typically, it also involves some kind of instructor or facilitator involvement, although the level of that involvement could be varied.

Cohorts provide structure and accountability

Cohort-based learning usually provides an actual time constraint, not just a suggested one (like a MOOC might). The course might last two weeks, six weeks, or some other time period- but it’s constrained, with checkpoints.

Participants move through the content checkpoints as a group, and thus can tackle challenges and potential confusion together, even if it’s not synchronously. They can share what they’ve learned and the experiences they bring to the table, enriching the already-existing course content. If the cohort course has a facilitator, they can check in with encouragement and group feedback, giving nudges to those who haven’t participated.

Check ins and accountability from other humans provides a real advantage over automated/machine-based check ins- if you haven’t yet, read my last post about AI feedback.

Cohorts provide community and belonging

At least in higher education settings, social presence leads to higher persistence (not dropping out) and satisfaction with courses. A sense of belonging in courses also contributes to greater success- learners need to feel like they’re valued and part of a larger network to which they can contribute.

Cohort-based learning allows participants to feel not only like they’re learning themselves, but that they have something to offer and contribute back to others.

Passive content consumption is supplemented by community engagement

The method by which content is served in cohort-based learning is often the same as MOOCs: video, readings, perhaps some demos. But supplemented with shared experiences: discussion/Slack forums, open office hours, or shared work hours can make passive content more dynamic and engaging. Participants can use these opportunities to discuss questions or share their own experiences.

Designing better online learning

While for some SaaS customer education teams, cohort learning isn’t new (I’m looking at you, HubSpot Academy bootcamps), it seems like the shift to cohort learning is growing. Clay University and Maven are just a couple of rising stars that offer cohort-based courses.

Customer ed teams (as with any learning design) should be cautious in proceeding with cohort-based learning when it comes to timelines and their audience. Cohort learning doesn’t automatically mean better learning. Even in college settings, online courses often have higher attrition rates than in-person courses.

Designing effective and scalable cohort learning experiences requires detailed planning about course timelines, balancing practical amounts of content consumption and community interaction, an involved facilitator, and practical application activities.

So how do we design effective cohort-based learning?

In my next article, I’ll share how I designed the cohort-based academy I mentioned earlier, which saw nearly 70% retention rates and an NPS score of 78.

In the meantime, let me know in the comments: have you ever successfully made it through a MOOC? How has an instructor/facilitator, or course community, helped you get to the finish line?

Thanks for reading and see you then!

That’s the question I posed in my last post, after I spoke with with James Meaden on his podcast The Alive Conversations (you can listen to our conversation here).

Before I get to the “how,” though, I think it’s important to discuss why human connection is important in learning.

The social aspect of learning

The importance of a community of learners

Learning is relational. Our backgrounds, our experiences, and who we learn with all impact what and how we learn. Let’s take a hypothetical example from the college course that I teach, Gender and Technology. I invented this particular example, but something like it occurs nearly every week:

Students are collaboratively annotating an article about predictive AI technologies. A criminal justice major mentions that they’ve learned about how these predictive AI systems can be used to try and determine which criminals are likely to reoffend. Another student replies to them, mentioning that they didn’t fully understand how predictive systems might be harmful until this example was presented.

This specific moment of learning happened because these 2 students were learning together. The Community of Inquiry framework, which is often cited in higher education pedagogy, emphasizes the importance of social learning in supporting deep cognitive thinking. Having conversations with others around complex topics can often allow for this kind of deep cognitive thinking to occur.

When learning with others, learners have access to a greater number of perspectives and schemas from which to draw. According to a 2024 meta-analysis, social presence is linked to stronger persistence (likelihood of completing the course) and learning satisfaction.

The importance of an involved instructor

In addition to the role of other learners in learning, an actively involved instructor can influence learner confidence and persistence as well. In this 2018 conference proceedings paper, I wrote how accessibility, approachability, and respect were three especially important qualities for instructors to demonstrate in their classes- especially online. These instructor qualities can help learners persist and stay motivated through difficult learning experiences.

I found evidence of this in my own mid-course evaluation survey this semester (my class is asynchronous and online, so I never meet my students in a live session). The responses were just released earlier this week.

When asked, “What do you like most about this course?” many of the responses focused on my qualities as an instructor, not the course content or activities (although those also came up- here are the full results, to be transparent):

“The instructor gives prompt and detailed feedback on assignments.”

“The interactiveness of the class.”

“I like how engaged the professor is and I enjoy learning about the connection between gender and technology.”

“I like the easy access I have to not just the reading, but also to the professor.”

“I like how approachable and accomodating [sic] she has been all year. She really helps clarify any questions that I had in an awesome way.”

I recognize that these comments only demonstrate what the students like or appreciate about the class and don’t prove that they’re learning any better than in a course where the instructor is less engaged. But engagement and feeling like you have an involved instructor does have an impact on motivation, confidence, and self-efficacy, and thus does have an impact on learning.

Learning is more than just an individual making mistakes and correcting them

A lot of the conversations around AI that I’ve seen lately have talked about using AI as tutors or to free up teacher time by providing feedback. These conversations are essentially boiling learning down to a matter of an individual’s interaction with content: engaging with content, perhaps making mistakes, learning from them. While of course learning sometimes does happen this way, there are plenty of cases when is not quite that simple. In this model of learning, the relational component is really stripped down to almost nothing, and that matters.

A recent MIT study1 looked at source of feedback in relation to student motivation and future engagement. It found that students invested more time and effort into the next steps of a coding assignment when they believed they were receiving feedback from a human vs. when they believed they were receiving AI-generated feedback. (emphasis my own in the below quote)

“Notably, high ratings in the AI condition did not translate to greater effort: believing feedback is helpful is not the same as being motivated to engage deeply with it…This dissociation aligns with social presence theory, which suggests that even minimal cues of human attention can shape motivation and effort in online environments. Learners may engage differently when they perceive a human presence attending to their work, even a stranger, due to social motivations that AI does not elicit.” [from p6 of the MIT study]

So if we know that learning is relational, shaped by peers, instructors, and the sense that another human is actually paying attention, then designing learning experiences can’t just be about delivering content.

Learning design needs to continue to think about and intentionally incorporate presence. That might mean:

• Creating space for learners to learn from one another

• Building moments where the instructor shows up in visible and human ways

• Structuring activities that allow different perspectives to surface

These choices acknowledge something important: learning doesn’t happen in isolation, only as interaction with content, even when it’s asynchronous.

In the future iterations of this series, I’ll explore some practical ways to intentionally design for human connection so that digital learning feels a little less isolated and a little more relational.

Let me know how other people have shaped your learning in the comments. Thanks, as always, for being here!

The heart of learning

If learning were truly only about content and access to content, wouldn’t we all be geniuses by now?

But there’s so much else to it: motivation, self-efficacy, external support and accountability. I really hope you’ll check out our conversation here:

The Alive Letter

09 | Alive with Christie DeCarolis | The Heart of Learning

In this conversation, Christie DeCarolis (Learning Designer, Adjunct Faculty at Rutgers, Learning Consultant) and I explore the multifaceted nature of learning, emphasizing the importance of emotional engagement, psychological safety, and the evolving role of AI in education. We discuss how learning is not just about content delivery but about creating …

Listen now

7 months ago · 1 like · James at Alive Institute

And with that, an upcoming series:
Connection by design

If you follow me on LinkedIn, you may have been noticing me, ahem, harping lately on the importance of humans in learning. I teach an online course part-time for Rutgers that started a few weeks back, and I’ve been launching new live workshops at my day job lately, so it’s been quite top of mind for me: how do we design space for human connection in learning so that the learning itself is more effective?

This isn’t a new obsession for me, actually: back in 2018 I gave a conference presentation entitled “Building Rapport Online” at the Distance Teaching and Learning Conference in Wisconsin. Here’s the proceedings paper I wrote for it, if you’re so inclined to read it- it won an award and I was invited back the next year to deliver the talk again.

This has always been an important topic- but somehow it’s becoming more important as we try to replace human…everything with AI.

My next [insert yet to be determined number here] posts will focus on this- specifically, different strategies I’ve used to take advantage of human connection to help people learn better, in places from college classes to product trainings.

In these increasingly digital times, human connection often to be more intentionally built. Thus, the series title: connection by design.

I hope you’ll join me! But first, listen to my conversation with James- let us know what you think.

New Year’s resolutions are really A Thing, huh? People always seem to pick big goals: they want to completely turn some aspect of themselves around.

It doesn’t quite sit right with me. I never liked new year’s resolutions.

I do have one resolution for you at this moment though: I promise this post is connected to learning design.

My beef with new year’s resolutions

This isn’t universally true, but I feel like most people’s new year’s resolutions suck the joy out of life. I will resolve to never eat a potato chip again. I will meal prep every weekend until my knuckles bleed. I will wake up at 4am every day to do 545 push ups. Ok, I’m exaggerating here a little, but have you heard of the 75 Hard challenge? Because based on that, I’m not even exaggerating that much (please don’t do 75 Hard).

This all sounds really punishing, huh? Do we hate ourselves? Are our resolutions rooted in shame? Is it really reasonable for you to go from doing 0 push ups to 545? Do you want to have any fun in 2026, or do you want to walk around with useless noodle arms most of the time?

If you have set a resolution for yourself: think of it in your head right now. If you haven’t, but you have an idea for something you want to do that seems scary, think of that too. Keep that in your mind for the rest of this article.

Considering resolutions

We like to set bold goals. That’s fair. Setting small goals seems boring. But big goals, with no plan—and which may assume you are someone you simply aren’t—is a fool’s errand. Perhaps before proceeding, we should just sit with our ideas for a moment.

• Have you left yourself any room for failure?

• Are you making plans that work with how you know you operate? (you probably know yourself pretty good, you spend a lot of time with you and all)

• Are you okay with being bad at something, maybe even for a long time?

If the answer is “no” to any of these, I’d encourage you to apply some learning design strategies to your resolutions.

Let’s backward design the shit out of this

1 Enter: learning design. I’d like to try approaching new year’s resolutions as I would teaching learners something new. I don’t want to introduce shame, or militaristic discipline techniques, or to leave no room for mistakes.

We are not assuming you are starting the new year as a bad or inferior version of yourself. Maybe you just want to tweak some stuff, or learn something. We need to meet learners (or, in this case, you) where they are, and design from there.

Instead of a new year’s resolution as a reinvention, let’s think of a new year’s resolution as a scaffold to a better version of an already-good you.

So how would we backward design this?

First: what’s your goal? Let’s pretend you want to start working out more regularly. In this hypothetical scenario, you don’t exercise at all right now. Your resolution is to start going to the gym 4x a week.

A regular, non-learning design approach to this resolution is to do just that: start going to the gym 4x a week. Maybe you do that for a week or two. Perhaps it involves aimless wandering because you’re not sure what one does at the gym. Then, inevitably, one week you get stuck at work late, or you have a bad head cold and you only get to the gym twice and so you just stop going completely and your resolution is dead.

In this fake story, your first steps are also the end goal. But starting there doesn’t make sense! Using backward design assumes that’s where you end up, yes, but going straight there requires a large leap in behavior that would be easier if we inserted some steps in between.

Scaffolding your resolution

In learning design, we use scaffolding to expand on learner skills and knowledge:

…instructional scaffolds are temporary support structures [instructors] put in place to assist students in accomplishing new tasks and concepts they could not typically achieve on their own. Once students are able to complete or master the task, the scaffolding is gradually removed or fades away—the responsibility of learning shifts from the instructor to the student.2

For someone who doesn’t exercise at all, going to the gym and exercising is a new task. It’s a pretty safe assumption that some sort of scaffolding would be helpful here even though we’re not talking about a formal learning environment.

Perhaps some scaffolded steps might look like this:

• Go to the gym once and meet with a personal trainer to plan a simple full body routine suitable for a beginner.

• Go to the gym once a week and follow the routine you planned.

• Do that for 3 more weeks. Leave some flexibility so that if you “miss” your gym day, you can go another day.

• Add a second gym day to each week and continue to follow your routine.

• Do that for 3 more weeks. Leave some flexibility so that if you “miss” your gym day, you can go another day. Or, just miss a day and go 2x a week for another week.

• Add a third gym day to each week and continue to follow your routine.

Do you see where this is going? Instead of diving in head first, we’ve taken your end goal and outlined some steps that should help you reach that goal with support and flexibility. I wouldn’t even change or add to the routine until you’ve consistently been going to the gym 4x a week, and fulfilled your initial goal.

Don’t underestimate the power of small changes

So what does all get me, in the end? New year, same me. Except now I’m going to go to the gym once a week (and eventually, 4x a week). That seems a whole lot easier than being a whole new person, doesn’t it?

Design backward from your end goal using small, scaffolded steps.

Add in a vegetable a couple times a week. Go to bed ten minutes earlier. Those things don’t require a personal reinvention.

New year. Same you.

While the first is super recent (and unsurprising given the dumpster fire of a job market this year), the others come from earlier this year, especially considering I was pretty absent in January & February 🫣.

Subscribe now

A quick reintroduction

I think I have 50% more subscribers than I did 2 months ago, so I wanted to take a minute to re-introduce myself to new readers!

I’m Christie, an instructional designer, weightlifting enthusiast, beginner runner, cello student, and bean lover. I started my career as a high school history teacher before earning a Master’s in Instructional Design & Technology. I worked as an instructional designer at Rutgers for 8 years, where I helped faculty improve their classroom pedagogy, designed in-person, blended, and online courses, and developed and delivered training on educational technologies.

More recently, I’ve been working in Customer Education at SaaS (software as a service) tech companies, where I help design and deliver live product training to scale customer learning and product use. I still teach an online course for Rutgers each spring.

Anyway, this is where Teaching with Tech in Mind came from! My brain, thinking about teaching people how to use tech and/or teaching with tech. If you’re new, I’m so happy you made it here.

And finally, thank you!

Thanks for choosing to read my little corner of the Internet this year. I’m excited to keep learning and growing with you all in 2026!

Anyway, I hope my 2023 thoughts still resonate with you here in the future. Happy holidays ❄️

It’s the holiday season, y’all (or perhaps it has been since October, according to retail stores in America). Before we cross over into the Time of New Year’s Resolutions, we encounter what I’m referring to as the ✨Peak Reflection Times✨. On LinkedIn, at least, it seems that everyone is sharing everything they’ve accomplished in their career over the course of 2023. [Note from 2025 Christie: you also have no idea what kind of not-so-great stuff is going on with these people that they are choosing NOT to post on ye olde social media.]

And I’m here to remind you, as you scroll past these posts and perhaps experience some FOMO: you are more than your job.

And what a year 2023 was. In the tech sector, it’s been a very bumpy one. It seems like the news articles about layoffs have been coming at us at a steady pace. Maybe you’ve been impacted by these layoffs, or maybe you’ve been feeling stuck because of the tight job market. [2025 note: oh 2023 Christie, YOU DON’T EVEN KNOW]

We are not what we have produced.

I had the opportunity to do a lot of content creation this year for my job. I could join the Peak Reflection Times and tell you about the educational opportunities I have created at my company. If you’re reading this from somewhere in the United States, perhaps you feel compelled to do the same: the US really loves equating productivity and being busy with success!

But that isn’t what I’ve actually been reflecting on the most. In this final month of 2023, I’ve spent the most time thinking of what I’ve done outside of work and how I can keep that person growing in 2024.

So here’s what I’ve done outside of work this year:

• Read 43 books. Maybe I will try to get to 50 before the year runs out. They were all fiction!

• Completed around 275 workouts. This is an estimate because I didn’t actually count, but I usually exercise 5-6x a week, so I split the difference between 250 (5x a week) and 300 (6x a week), counting 50 weeks so far this year.

• Taken 42 cello lessons. Plus lots of practicing outside of lessons!

Within all that time, I PRed some lifts. I learned how to reach basic sheet music even though I had no musical background. But I didn’t actually produce anything tangible (except maybe some muscles? Maybe?) I didn’t even track any of that as it was going–I just had to do some calendar math to figure out how many cello lessons I’ve taken and my Kindle has reported to me how many books I’ve read.

Reflecting on this stuff is important to me because I (for the most part) have control over the hours outside of my job. If I lost my job, how might I fill my time? Who could I keep being?

Did I meet my personal goals?

You may have noticed that I reported on what I’ve done outside of work this year without some kind of standard measurement metric that I set myself against. Like, I was I trying to get to 50 books? Or 300 workouts?

I don’t know! I kind of hate New Year’s Resolutions (more on that later) and setting personal goals for myself (more on that later too). So I didn’t start this year with any particular goals in mind. I just thought, “hey, I want to do this stuff. Let’s see how much I can do and if I like it.”

And I did like it. So I’m going to say yes, I did meet my goals.

In considering today’s push for AI (and what that means for human workers), Kentz writes about the shift from human telephone operators to automatic telephone switching and the unexpected impacts:

Yet something else disappeared alongside the operators. In small towns, operators had served as informal emergency dispatchers. When someone called about a fire, the operator didn’t just connect to the fire department—she rang every household on the affected street. During medical emergencies, operators tracked down doctors at dinner parties or movie theaters. They knew which families had cars and could drive people to hospitals. They remembered that the Miller family was out of town and their pipes might freeze.

This infrastructure wasn’t recorded in any job description. It emerged from relationships operators had with their communities. When automation arrived, there was no line item for “emergency network coordinator” or “community information hub” because those functions had never been formalized. They simply evaporated.

This really makes me think about how we are incorporating AI into learning, and the unexpected ripple effect that it may have like the one Kentz describes. Especially in traditional schooling (K12 and higher ed), students’ relationships with teachers frequently have a huge impact on their lives- and often not due to the content or skills they’re being taught.

Since I’m working in customer education currently, I want to examine it from that lens as well. We should consider: what happens when we allow people to rely on AI for something like product education? What happens when we remove human touch points and relationships from a customer learning experience? On-demand options can be more efficient for customers- but does it also make it easier to drop a tool when there’s no face connected to it?

Mike Taylor’s “The Right Kind of Difficulty: Why Worked Examples Beat Practice for Novices”

Taylor explains how to approach adding “desirable difficulties” to learning when you’re designing for novices vs. experts. Problem solving and practice are important to strengthen memory and retrieval- but, that doesn’t mean learning designers should present learners with problems right away. He walks through how to effectively incorporate problem-solving with novice learners, who might become too cognitively overwhelmed if they’re thrown into problem-solving right away.

Research shows that jumping straight to problem-solving can actually impair learning by overwhelming working memory and blocking schema construction (Sweller & Cooper, 1985). Novices don’t need more practice problems. They need better examples.

In this article, we’ll unpack the Worked Example Effect: a robust finding from Cognitive Load Theory showing that novices learn complex skills faster and retain them longer by studying expert solutions than by solving problems independently.

I love considering this from a customer education standpoint because a lack of worked examples is often a huge roadblock for customers getting started. They feel unsure and not quite ready to dive in, even if they have the technical know how; we have to bridge the gap for them. I think this is the issue I was poking at in my article from earlier this year, “Closing the loop for learners.”

In customer ed, we often lack the business-relevant, full worked examples to provide (someone in customer success may have easier access to a full worked example- in my experience, CE usually has access to marketing-style customer stories that are often lacking important details for application).

I think this is where LLMs could prove useful for us in CE: could a chatbot provide us with a full worked example to use as a foundation and revise to provide to our learners? How can we better incorporate these into our training content?

Dr. Christie Vanorsdale’s “Stop Building Content Libraries. Start Building Learning Ecosystems.”

Dr. Vanorsdale discusses an issue common to many organizations: training isn’t supported by a broader learning infrastructure- where professionals can practice what they’ve learned, and where peer learning can take place. This is important because it’s those experiences where you’ll find the most impactful learning: through experience and collaboration.

The shift L&D leaders need to make:

OLD: “We need more courses”
NEW: “We need systems that support how people actually learn”

OLD: “Let’s build a better training catalog”
NEW: “Let’s architect an ecosystem where formal training, peer learning, and experiential practice are connected”

OLD: “Our LMS should house all learning content”
NEW: “Our LMS should be one node in a larger infrastructure”

Vanorsdale outlines a 3-step process to outline your current ecosystem, identify integration gaps, and design for ecosystem, not catalog. She provides a list of questions that are really useful for learning practitioners to reflect upon.

How can we make our learning experiences part of an ecosystem, and not just different pieces of content that we are pointing or mass emailing to learners? We’ve been doing a lot of thinking around learning journeys on my current team, and I really appreciate having this article to reference.

Hello! Magazine’s “Why we miss structured education in our 20s and 30s and what it says about our brain health”

This piece looks at the “personal curriculum” trend that’s cycling through various social media platforms lately (I’m not personally on TikTok so I can’t attest to its presence there, but I’ve seen it on YouTube).

Most iterations of the “personal curriculum” that I’ve seen focus on arranging content for yourself to consume over a set amount of time, so Kelly Heurer’s (VP of Learning from the Project Management Institute) contributions to the article stood out to me:

Kelly advised: “Don’t recreate school; recreate what actually works. Most adult learning doesn’t come from classrooms. A well-known model from the Centre for Creative Leadership - based on studies of how leaders actually develop - suggests that the majority of professional growth comes from learning on the job (about 70 per cent) and from mentors and peers (about 20 per cent), with formal courses or books making up a much smaller slice.

“That means the best ‘curriculum’ for your 20s and 30s is often integrating small learning loops into the work you already do, and cultivating relationships with people who can give you feedback and perspective.”

I have been feeling the impact of real-life practice & collaboration on my learning lately, both personally and professionally. Here are just a few ways:

• Talking to people on LinkedIn: I’ve been connecting with and speaking to more learning designers on LinkedIn lately, and the conversations often make me consider learning design in different ways that I would not have considered had I stuck to consuming content and simply keeping it in my own brain.

• Writing about learning design: Both in this newsletter and on LinkedIn, I’m forced to reflect upon best practices and how I can improve. Even exercises like writing this article help me really reinforce what I’ve read in the past. Why even write a newsletter when you barely have any subscribers? I’ve thought in the past- but it’s really because the process of reflection helps me be better at what I do.

• Practicing stuff: Both tinkering and deliberate practice is always super helpful to me. Professionally, I love tinkering (like, making small non-serious things, that I maybe could eventually turn into a project, you know?) to try things out because it’s way more low stakes that way. And personally, learning cello has really taught me that deliberate practice is so important. I can play the same piece over and over again a million times and not get better at it (really) unless I focus on practicing specific skills, specific measures, slowing down. Your approach does matter!

Reflection is the underrated hero here.

I’m emphasizing this because I’ve found myself saying it over and over again in various places lately. So: how are you incorporating reflection opportunities in your learning design?

I hope this newsletter has encouraged YOU to reflect a little bit! Can we make this a learning ecosystem? Let me know your thoughts (remember, peer collaboration- also a great way to learn 🤓).

In November, I co-hosted an instructional design-centered career event with Clea Mahoney, manager of Customer Education at Rewst. I’ve known Clea for [insert unknown number of years here], after virtually meeting her and re-encountering her in various instructional design-related communities (see: our section on networking and its importance).

What sparked the creation of this event? This year, I was on the job market, and y’all… it’s wild out there. For quick background, I’ve interviewed and gotten offers for jobs every year since 2020 (note: that doesn’t mean that I have actually changed jobs that frequently). This year is just different. There’s a lot of competition, especially for remote roles.

Despite all that, this past spring, I was able to secure interviews with 15 organizations. I posted on LinkedIn wondering if my experiences might be helpful to anyone–and Clea, who landed her current role two years ago and has also acted as hiring manager in the current market–offered to collaborate. It made sense! We have similar career paths, recently moving from instructional design in higher education to SaaS customer education.

After hosting our session, we wanted to provide a takeaway resource for those who couldn’t attend! Here we are.

Introducing Clea

I learned a lot from participants’ questions in our live event, and went on a few tangents related to my experience – what worked and what didn’t as I looked for opportunities in customer education.

I’d tried all sorts of training and instructional design work over the years, in my very varied background. I dabbled in HVAC distributor marketing, higher education administration and faculty training, and internal learning and development at a tech company and at a large accounting firm. Helping adults learn in non-traditional ways made my heart sing, and when I learned about customer education — helping busy adults learn how to use products and develop skills that (usually) doesn’t come in the form of “compliance training” – I had a hunch that working in the field would be a better fit for my skill set (technical training, instructional design), passions and values (creativity, fun, play), and bank account (yes, I get paid more now than I have in non-customer education work although that is not necessarily the norm).

Finding roles

So, how can you stand out in the current job market? Here are our top tips for you to apply in your job search, from learning about roles in the field of technical customer education, to how you show up in interviews (which we hope you get!)

Clea says…

Do: Broaden your search around job titles. I really wanted “instructional designer” but noticed that those roles all kinda looked different. Some wanted a team lead, others wanted straight-up eLearning development (which is not my bag). Check out Christie’s page for some titles you can search for. I applied for “Training and Community Enablement Manager,” got it, and am now Learning Experience Design Manager -- a title I pitched to my boss at the start-up where I work at the time of writing this!

Don’t: Assume you have to meet every criteria spelled out in the job description. These are written aspirationally. Women often hold back from applying for roles. If you think you’d be a good fit, just apply!

Christie says…

Do: Look outside of LinkedIn! It’s expensive for companies to post open roles on LinkedIn, and digging through the listings/searching by specific job titles is often frustrating. The most helpful job boards I’ve found are Skip, Teamed, Ed Tech Jobs, and edtech.com.

Don’t: Automatically disqualify yourself for roles based on lack of experience with a specific technology. Of course, if a job is largely about building Storyline modules and it’s the first thing listed under “Required Experience,” that should give you pause. But overall, specific tools can be learned. Have you had experience with similar tools? Have those examples available in your portfolio. Also consider: do you want a job that requires you to spend most of your time building in a specific tool? (Maybe the answer is yes! That’s fine! But often it’s not).

Networking

Clea says…

Do: Reach out to folks you admire in the field to ask about their experience. It’ll help you get a feel for what a [insert whatever title] in Customer Education does, and I find that folks in our world love sharing tips to help others get into the work that we do. Personalize your outreach.

Don’t: Ask for chats with hiring managers of jobs you’re applying to. It’s just cringey. If you think you’d be a good fit, apply...don’t ask for a chat with someone who’s swamped with applications already. Do let them know you’re applying (and I know Christie has some great tips for that!) but don’t ask for a “chat” or a “call.”

Christie says…

Do: Join communities and take part in their networking events! Two great ones are CustomerEducation.org and FOHE.org (Future of Higher Education). Both offer opportunities for virtual networking with other professionals and advertise other events as well. And for those with whom you network–follow up! Let them know how your search is going and how they’ve helped you. Building a network requires a bit of maintenance along the way.

Don’t: Expect to hear back from everyone you message, and don’t expect every message to go somewhere. Yes, a lot of your messages will go unanswered–whether you’re just requesting a networking chat on LinkedIn or in a community, or if you’re messaging a hiring manager about a role. Don’t take it personally. The receiver is probably busy, or perhaps not as chronically online. You will hear back from some people. Some of those replies may not lead anywhere. But remember, in the end, you only need one!

Framing your experience

Clea says…

Do: Tailor every application (cover letter and resume) to the job you’re applying for. Hiring managers want to see your passion for their company and team, and a vague application that looks like it was blasted to 25 different jobs will get passed over. What’s a priority for the role, and what does the company seem to care about? Do your research and speak to your relevant experience on those points.

Don’t: Lie about your past titles/experience. If you’re a K-12 teacher, share that, and show us how it translates to customer education. “Creating content that engages highly distractible students” is valuable! “Applied ADDIE in working with stakeholders and clients” when you really did “lesson planning for 5th grade math” is not authentic (and personally, I just find it a little weird). Yes, it’s true -- some employers might overlook teachers. But the good ones won’t. And you deserve to work for a good employer, not just any employer.

Christie says…

Do: Lean into your niche. Even if you have lots of generalized experience, you probably have a small area of expertise that you’ve really honed. For me, it’s facilitating live training. While I did apply to jobs outside of this particular focus, I leaned hard into the applications that highlighted a live training component. I made sure to highlight those experiences in my resume,

Don’t: Focus on listing every job responsibility you’ve ever had on your resume. The name of the game in the current job market is getting noticed. Frame your resume bullet points around your impact instead of your day-to-day work. Here’s one example: instead of saying “Led 1:1 consultations with faculty on integrating tech in their teaching,” try “Increased edtech adoption through developing trusted relationships with faculty in 1:1 consultations.” You can see more examples on my job search tips sheet. I also included a description of how I tried to make tailoring my resume for every job quicker and more efficient.

Here are the links compiled nicely in one spot, plus some other resources:

• Christie’s 2025 job hunt tips

• CELab podcast featuring Clea’s story about launching a customer education LMS

• Christie’s portfolio (as an example of a portfolio)

• A participant shared this LinkedIn post with good LinkedIn L&D follows (Clea recommends Heidi Kirby and Tim Slade particularly)

• Clea’s recommendation for career coaching: Work with Santana Inniss or learn more about her work with the GLOWI initiative, which offers affordable coaching for women.

Best of luck, and connect on LinkedIn if we aren’t already!

Christie & Clea

I don’t love the default to video in instructional design these days.

I think there’s a pull to make short video clips due to the “attention economy” (people have shorter attention spans! people don’t want to read! let’s grab their attention through video! etc), but sometimes learning isn’t best served through video.

First thing: I am not a video hater! Some people commented as if I was saying video is evil and we should never use it. No- I hate video as the default.

So how do I decide when and how to use video? Here are some best practices I use to guide intentional decision-making around content modality.

Most of this is informed by Mayer’s principles of multimedia, which are research-backed best practices for designing multimedia instruction. They’re based on three assumptions:

• Dual-channel assumption: People process auditory and visual information through two separate channels.

• Limited-capacity assumption: This assumes that people can only absorb so much information at one time; an overload of information will result in some of that information not being absorbed.

• Active-processing assumption: People should be actively engaged in the learning process, not passive consumers.

When should you use video?

Here are the questions I tend to ask myself when I’m designing learning experiences:

• Are you explaining graphics or data-based visuals that would benefit from on-screen highlights as they are explained?

• Are you demoing a process (whether it’s software or in physical space)?

• Is demonstrating emotion/authenticity important?

If your answer is yes to one or more of these, video might be a good option (although it doesn’t have to be the only or best option, still).

Explaining graphics or data-based visuals

Mayer’s signaling principle indicates that it might be helpful to incorporate into a video if you’re creating instruction about complicated graphs or charts. Use cues like arrows, bold text, highlighting that sync with narration to guide the learner’s attention. This can potentially reduce confusion from too much information on the screen.

Demoing a process

Whether you’re showing someone how to format a document using styles in Microsoft Word, or demonstrating how to perform a deadlift, Mayer’s multimedia principle recommends using a combination of words + images. “Words” and “images” can take multiple formats: “words” can refer to text or audio narration; “images” can refer to static graphics, animations, or video.

In the case where I’m making a video, I’d likely choose spoken narration to accompany the video.

Demonstrating emotion/authenticity

This is a little outside the scope of this post, but I think sometimes we ignore the emotional component of learning when we talk about designing training.

Including short learner success stories as video can be helpful to demonstrate authenticity and sincerity. Including short instructor intro videos might help learners feel more connected to the process and content. These are usually presented as “talking head” videos (where you’re presented with a portrait-style video of a person talking at the camera).

In this case especially, I think the information should be readily available outside of video as well, but having a video option can be helpful.

In fact, Mayer’s research (specifically the image principle) suggests that talking head videos are not good for learning. The Digital Learning Institute recommends:

“Consider using talking head videos initially to develop connections and build trust only.”

That pretty much hits the nail on the head with what I’m trying to suggest here. Building connections with learners can be helpful for the overall learning experience, but talking head videos may not be the most effective way to transmit information.

How can you make your use of video more effective?

We can again use Mayer and also Universal Design for Learning (UDL) principles to guide us here!

Provide context for your video

Personally, I hate when I reach a video in a course and I have no information regarding what I’m about to watch. It also makes it difficult for me to find the information again later on, as there’s no text to quickly skim and remind me where information might be within a particular video.

But beyond my personal preferences: Mayer’s pre-training principle recommends that people learn better from multimedia when they know the basics going in. So when you’re presenting learners with a video, it could be helpful to lay out the key terms or concepts in text beforehand. I like to include these things, plus a short (1 sentence) description of the video and what the main learning goal is.

Additionally, this helps activate prior knowledge, priming learners so they’re ready to build upon the schema they already have in place.

Provide multiple means of representation

I’m tipping my hat to UDL for this one:

“Learners differ in the ways they perceive and make meaning of information...Learning, and transfer of learning, occurs when multiple representations and perspectives are used… there is not one means of representation that will be optimal for every learner; providing options for representation is essential.”

In practice, this means that learners should have choices and options from which to learn.

• If concepts are explained through video, the key points might be available in text as bulleted summaries.

• If there’s a demo video, perhaps screenshots + text is also provided.

Learners can use what is best for them in that moment, or even use all modalities and reinforce learning through different processing channels.

For instance, someone in my comments mentioned their preference for recipes as video. Personally, I hate recipes delivered through video! I’d much rather learn how to make something following written instructions. I have come across recipes where I can only find a video version, and I decide to skip it because of that. But really, it’s best when both written + video instructions are provided.

Don’t overload your video with on-screen text

I’m going to be honest and say I’ve done this before. It’s hard to not do this, because key stakeholders in learning design (SMEs, leadership, edu-adjacent teammates) will push for lots of on-screen text. Sometimes learners will even push for more on-screen text (people don’t always ask for what is actually best for them!) It’s difficult to pry on-screen text from people’s hands. 😬

But what does Mayer’s research show? There are a few principles that bring this up specifically. The redundancy and modality principles suggest that we learn best from the combination of spoken words + images specifically. Adding text onto the screen in addition to narration can result in cognitive overload.

If you are using on-screen text: The audio narration should read the text as it is presented on-screen. This might be helpful when doing something like providing instructions. Graphics shouldn’t be used during this time, ensuring cognitive channels aren’t overwhelmed.

My final video recommendations

• Make what they’re about to watch clear: Give the video a clear title, a brief description, and explain what they’re about to learn/why it’s important.

• Summarize the key points of the video in text: Below the video, include a couple of key bullet points as a summary.

• Chunk your videos wisely: Make it easier for learners to find (and re-find, if needed) what they need by keeping your videos chunked by topic. This doesn’t necessarily mean they have to be short, but they probably will end up being on the shorter end if you’re segmenting your videos properly.

• When in doubt, defer to Mayer.

Thanks for reading, and for those of you in the US: I hope you have a restful Thanksgiving next week!

Those of us that are chronically online see a lot of discussion about sharing your Storyline modules, developed slides- lots of finished products. Perhaps you might get a peek at someone who’s mapping learning objectives to outcomes or at some storyboards in development.

But what else goes into designing effective learning? There aren’t a lot of discussions around soft skills required in instructional design, which only paints half the picture of what the day-to-day job looks like.

The soft skills I shared

Building relationships

Figuring out who knows what is essential, especially if you’re serving as an in-house instructional designer. Who are your subject matter experts? How can you foster ongoing relationships with them?

Asking questions

A lot of times people think they need training about one thing, but the true need might be buried by something else (technical issues? process challenges? missing resources?) It’s our job as IDs to figure out what the knowledge gaps are and where they’re rooted.

Listening

This seems pretty self explanatory. If you are going to ask people a bunch of questions, you need to be pretty good at listening to their answers and asking follow up questions. And do a lot of “What I’m hearing is...” in response to make sure you’re both hearing the same things.

Investigating

Asking questions, good listening skills, and relationship-building are all part of your investigative process as an ID. Trying to figure out *why* specific learning is needed, what’s been done before, and what’s worked/has not worked is important. What resources already exist about a topic? What’s already been tried? If it’s a brand new initiative: what has worked well for other concepts or topics?

The soft skills others contributed

From Evan Hall:

Coaching - yes you ask questions and listen, but it’s also your job to lead to a positive outcome. Empathy - beyond investigating the “why” for facts, digging into motivation as well.

From Dana Kocalis:

Negotiation skills: Whether you are discussing pay rate, timelines and due dates, or creating and managing expectations - negotiation skills can help create outcomes that work for both parties.

Educating others - Many SMEs don’t know our world - the world of eLearning. Helping them understand how they can help you help them is critical. Educating SMEs by giving them a brief overview of the project and your process, along with how often you will need their participation, review cycles schedules etc.

From Cory Smith:

A big one for me is adaptability. Much of what we do is around facilitating and embedding some kind of behavioural or performance change. The ability to adapt and quickly relate to different environments and situations provides a head start in approaching the why/what/how from an appropriate perspective (whilst at the same time, not being assumptive and being comfortable with contradiction - haha) 🙂

From Doris Harris:

Ability to synthesize info: having the “outsider” perspective can be helpful when working with SMEs who often are too close/married to the material. Being able to synthesize sometimes disparate info will help you come up with innovative solutions.

From Vince De Freitas:

Design itself. Generally, I’ve noticed the field to be a bit overdeveloped in Instruction part of the title.

UX in learning has been lagging for years, and you don’t need to be flush with budget or tooling to design thoughtful, user-centric, clean and beautiful experiences, you just need to upskill on design as a broader craft and find ways to apply the principles to ID work.

And I loved this comment from Heather Porterfield:

We are detectives of a most scrupulous nature! We have to see everything, break it apart and put it back together again. It more than just making things pretty.

What do you think?

I had a lot of fun seeing other’s contributions to this list, and I’d love to hear from you! What do you think are the most under-discussed or valuable soft skills required for instructional design? Please share here, or go to the original LinkedIn post to share your thoughts!

This was the problem I was presented with recently. I had data (in the form of poll responses) about what product features attendees were most eager to start implementing, and I wanted to follow up with them the next week to ensure they had the tools (and a nudge!) to get started.

Here, I share an outline of my process using ChatGPT to create a solution. I’m intending for it to give you some ideas to run with; it’s not written in such a way that I expect you could replicate exactly what I did. I hope it inspires you for the future!

Problem: Here’s where I started

• A spreadsheet of webinar poll responses and related respondent data (their name/email/basic account information)

• Six different email templates that I wanted to send out; the specific template sent to an individual would be based on poll response and their account level

• Further need to customize email (with things like attendee name, appropriate company contact name, etc)

• A basic knowledge of HTML (but really no real programming knowledge beyond some basics of around syntax)

My goal

Streamline sending appropriate follow-up information to webinar attendees based on their poll response, including email personalization. I also wanted to ensure accuracy of the emails, as 6 templates + further personalization introduces a lot of room for human error.

How I used ChatGPT

First, I asked ChatGPT the following. Note that I spent approximately zero time crafting this prompt…you are witnessing my brain vomiting into ChatGPT. I kind of just wanted to see what it would offer.

I need to send a slightly different version of an email to approx. [#] of people based on specific criteria. There are a couple of pieces of text (names) that need to be changed in the email as well. Can you help me create something to make this easier?

It gave me three options:

1. Mail Merge using Excel + Gmail

2. Use Python to automate email sending

3. Use a CRM or marketing tool

Options 1 and 3 wouldn’t work for me based on the data I was working with and the specifics of what I was trying to do. Thus began my journey with ~vibe coding~.

Solution: ChatGPT and I vibe code a Python script together

I asked ChatGPT to help me with number 2, and we had a conversation during which it wrote various Python scripts for me until I had what I needed.

Essentially, here’s how I ended up implementing the Python script ChatGPT drafted:

1. I placed each of my 6 template emails into the script, assigning each a variable.

2. I entered the personalization variables within those emails, corresponding to the correct columns in my spreadsheet (e.g., {FirstName}, {CSM}).

3. In my spreadsheet, I added a column named “Version” and assigned each attendee to the appropriate variable of the template email they should receive.

4. I exported the spreadsheet as a CSV.

5. When I ran the Python script, it parsed the CSV and used its data to auto-generate an HTML file for each row (attendee), using the assigned email template and adding the appropriate personalization variables based on column data.

6. I could then copy the HTML into gmail to send (it took me <20 minutes to send a lot of personalized emails, also double-checking some details to make sure my script was working properly 😅).

Some lessons along the way

ChatGPT wasn’t a super great teacher, but I didn’t really give it any context about how much Python I did or didn’t know. Here are some things I stumbled into while implementing this solution:

• It provided too-basic instructions: The instructions ChatGPT gave for running the script required lots of questioning from me. It provided super basic steps for installing Python on my computer using the terminal. That step may have freaked some users out.

• …again: When it gave me instructions for how to run the script on my computer (using the terminal) it did not tell me I had to point the terminal to my specific folder path where the script and CSV lived, so I just got an error at first.

• You need to provide lots of details to generate the right solution: After trying script v1, I realized that my script was generating .txt files. I wanted to email hyperlinks in my email, which wouldn’t be included in a .txt file. It took a lot of back and forth with ChatGPT to figure out how to edit the script so that I could add embedded links to my email (vs. just copying and pasting a raw url). We landed on creating HTML files, which required that I add HTML to my email templates within the script. I know HTML so that wasn’t a big deal, but might have added too much complexity for some.

• You have to know enough to ask the right questions: I am not sure I would have eventually landed on the solution I did without knowing to ask certain questions. ChatGPT would give me solutions that sounded like they would work the way I wanted them to, but when I asked further, it would clarify that the solution would work slightly differently.

• …LLMs make stuff up, so keep an eye out: When I tried to use ChatGPT to generate the HTML for my email templates to make it faster, it started hallucinating urls.

• You can do a lot without learning anything: I’m kind of impressed (horrified?) at how much Python I did not learn in doing this. I definitely see the limitations in trying to create something larger-scale without error.

What I’m excited about moving forward

Did creating this possibly take longer than just personalizing the emails manually? Yes. Do I think it was worth it? Also yes.

First, I do think it reduced the possibility for user error in trying to copy and paste the correct email and make the correct personalization changes for every individual email. There were just too many spots where I could have messed this up.

Additionally, I think this will be useful to me in the future. Now that I have the base script + process down, I can continue to format my spreadsheets properly and substitute in new template emails for post-webinar, personalized outreach based on poll responses or chat interactions. I’ve actually already used the script again, within 10 days of creating it.

I think this will help me most to personalize post-webinar outreach in order to:

• Answer questions for struggling users

• Provide helpful nudges for eager/excited users

• Reduce information overload by providing only resources the user has expressed interest in

Wrapping it up

Was my process perfect? Definitely not. But it worked, and it showed me what’s possible when you mix curiosity and an assist from AI.

I hope this sparks an idea for how you can take what you already know about your learners and turn it into something actionable!

Carl Hendrick’s “The Paradox of Memory: Why Forgetting Makes Learning Possible”

Hendricks discusses the role of retrieval vs. storage in memory and how our brain decides what to put into long-term storage (instead of the more easily-retrieved working memory). He discusses the importance of “adaptive forgetting” and how it creates necessary space for abstraction and connection in learning.

“Consider two students preparing for an exam. One crams the night before, achieving high retrieval strength but little storage strength. The other spaces their practice over weeks, encountering repeated moments of forgetting followed by effortful recall. On test day, the crammer may outperform in the short term, but months later, it is the student who embraced forgetting and retrieval who still remembers.”

What this is making me think about:

A lot of learning design I’ve encountered doesn’t offer a lot of time for spaced practice: learners are assessed or apply knowledge right away, but in many cases (I’ve encountered them both in a college classroom and in SaaS customer education) we don’t know if the learner is continuing to put that knowledge to use–which would encourage its storage in long-term memory.

What can we do to further encourage this by making it a part of our learning design?

Dr. Christie Vanorsdale’s “#16 Templates as Cognitive Architecture”

Dr. Vanorsdale discusses specifically how templates help learners manage cognitive load while developing expertise.

“Templates work because they align with how the brain actually builds expertise. Expert performance relies on pattern recognition—the ability to quickly identify familiar configurations and apply appropriate responses.

Templates help novices:

• Recognize effective patterns before they have enough experience to discover them independently

• Build mental libraries of proven approaches they can adapt

• Develop pattern matching skills by seeing variations on effective themes

• Reduce cognitive load during application so they can focus on customization and creativity”

What this is making me think about:

Dr. Vanorsdale’s article addresses creating templates for beginners vs. intermediate learners vs. advanced learners.

I’m a big believer in providing templates to learners to help ease the transition from learning to applying new knowledge. I haven’t played with the idea of scaffolded templates, however.

It’s something I think I want to be more aware of: how can I explicitly design templates for beginners vs. more advanced learners to help them better apply what they have learned? Templates for advanced learners should allow for more flexibility, and creating templates designed specifically for different levels will help learners better transfer the skills in their specific contexts.

Am I missing any clear spaces where I can provide some kind of templated learning, whether that’s detailed guidance for beginners or loose frameworks for experienced learners?

Mike Kentz’s “Stop Skimming Chats, Start Annotating Them”

Kentz provides a really useful idea for teaching students to think critically about how they’re using AI: by having them annotate their own chatbot transcripts. While this strategy is very classroom-centric, I think the discussion of metacognition and technology use is an important takeaway which can be applied more broadly.

“The prevailing approach to AI use - especially among young people – is scattershot, untracked, and passive in nature. By stepping back and annotating oneself on the page, we wake ourselves up. We develop, essentially, and entirely new section of brain development devoted to a specific metacognitive act associated with our technology use.”

What this is making me think about:

Outside of formal classrooms (workplace learning circumstances: whether it’s internal L&D/enablement or external, customer-facing training), we often drop the opportunity for learners to reflect because we laser focus on application and transfer.

But what if encouraging metacognition through reflection makes application and transfer more likely? What if community reflective opportunities allow learners to learn from each other? How can we incorporate reflection effectively into various workplace learning strategies–in a way that provides a clear benefit to the learner and doesn’t just feel like busy work?

A LinkedIn post from Michael Forder about course design decisions

I was about to publish this newsletter issue when I came across this LinkedIn post from Forder. I appreciate the visual he’s provided, which sets each course design decision on a spectrum between “student supporting” and “teacher supporting.”

It is easy enough to say that courses are designed with specific students in mind, but when viewed critically, how many course decisions put the actual (not assumed) student experience on top?

What this is making me think about:

I think oftentimes learning designers get caught up in designing specific details of learning experiences (whether it’s an asynchronous online course or a live training) that has no measurable impact on the learner experience. How can we better center our design decisions to improve learning? When might it be appropriate to lean more toward the instructor-supporting side of the spectrum?

Wrapping up

That’s it for today, folks. Lots of questions, no answers! I guess my overall goal for this post is similar to what I discussed in reviewing the Kentz article: to provoke some self-reflection from you, reader. I hope this inspires you to think about how you’re approaching designing learning experiences.

Thanks for reading!

In this newsletter, I’m going to share how we can better incorporate agentic engagement into adult online learning experiences and why it’s important.

Agentic AI has been a hot topic lately, but it has nothing to do with this post. Womp womp.

What is agentic engagement?

Agentic engagement is the extent to which learners influence their own motivation and learning environment–in other words, how are they using their agency to influence their learning? How do they contribute to the construction/flow of their learning activities?

It assumes a level of learner participation that goes beyond simply clicking through a course. Instead, learners are active participants in shaping their learning experiences.

A lot of characteristics of agentic engagement are easily recognizable as good practice to increase engagement overall. Examining how your learning experiences incorporate agentic engagement specifically can be helpful to increase learner interest and opportunities for knowledge co-construction (or less jargon-y: more social learning experiences).

Some characteristics of agentic engagement include when learners:

• Share expectations and interests

• Ask questions

• Express frustrations/understandings

• Take initiative to seek alternative learning options

• Express opinions and preferences

To close the loop on social learning here: when learners talk about their learning out loud, they learn from each other.

Why is agentic engagement, specifically, important?

A learner can be fully engaged without being agentically engaged. A learner can complete all of the necessary tasks required of a course, but never share their own interests, ask questions, or seek alternative solutions.1

Most of the research (some of which I’ve included in the footnotes below) around agentic engagement is centered around students in K12 or higher education learning environments. Based on Knowles’ theory of andragogy, however, I think it’s just as important to consider for adult learners in most learning settings.

In my April post about social learning, I focused on Knowles’ principles readiness to learn and role of experience.

However, there are two different principles that strike me as evidence that agentic engagement is important to explicitly incorporate into adult learning environments:

• Self Concept: Circumstances that allow adult learners to self-direct.

• Intrinsic Motivation: “Adult learners are responsive to some external motivators (better jobs, promotions, higher salaries), but the most potent motivators are internal pressures (the desire for increased job satisfaction, self-esteem, or quality of life).” (directly quoted from the Knowles link above)

The wrong way to encourage agentic engagement

I don’t think these two principles are completely ignored in adult learning design. I just think that most adult learning opportunities target them with strategies that aren’t quite effective. These strategies assume that adult learners come to us ready to agentically engage in order to learn.

Self concept

Adults are often given broad agency over self-directed learning. Sometimes, too much choice leads to overwhelm, which can lead to no decision being made at all. Some examples:

• Companies provide professional development stipends for the year, but no meaningful conversations happen with employees around how the stipends might be used or tied to on-the-job goals. Oftentimes, these stipends are underutilized.

• On-demand learning options are chunked into microlearning to accommodate adults’ busy schedules. Which often turns into lots of small course options! This can be an overwhelming amount of choice regarding learning content. It might create difficulty for a learner trying to quickly determine what’s most relevant or helpful given their individual needs/goals.

Intrinsic motivation

We often assume adults have intrinsic motivation, or they don’t. They’re excited to learn, or they aren’t. So we provide them the necessary options to self-direct (as discussed above), but assume there’s not much we can do about increasing intrinsic motivation if it doesn’t already exist.

What if we stopped assuming adults will engage agentically from the start? What if, instead, we provided nudges encouraging them to do so?

How can you meaningfully encourage agentic engagement?

Too much agency might result in inaction, but finding the “goldilocks” amount of agency for learners is possible, and could help spark a dying ember of intrinsic motivation.

Instructional designers designing for adult learners should strive to find the “goldilocks amount” of agency in learning. This will help learners take the first step toward action and initiative within their learning journey.

Providing explicit (but bounded) opportunities for self-direction, and multiple and varied opportunities for expression, is key.

Here are some concrete ideas to implement in your L&D, training, post-secondary, and/or customer education settings.

Surveys as a mechanism for self-direction or expression

Before either live or self-paced training, learners may require some kind of nudge toward finding or expressing what content would be the most helpful for them.

• Live learning: Surveying participants before live learning provides them a direct opportunity to share their expectations and interests for the session. Be wary of making even the question too open-ended, though: instead of “What do you hope to learn during this session?” you might try “We’re hoping to walk you through X process using Y best practices during this session. What can you share about yourself or your experience that would make this most helpful for you?”

• Self-paced learning: Recommending specific learning content to self-paced learners based on a survey or personalized “quiz” asking about their role or needs can be helpful in narrowing down options for self-direction to 3-4 (instead of 10-15+).

Mid-learning check ins

Asynchronous, self-paced courses often strip away the opportunity for learners to express frustration or confusion as they progress through course material. This is somewhat more available in live sessions if a chat or Q&A area is available, but still relies on learners to take initiative if confused.

• Self-paced learning: In self-paced courses, it might be helpful to stop learners periodically and ask them to rate how confidently they feel they could apply what they’ve learned so far (1-5 Likert scale style). Personally, I would be more specific here than just say “how confident are you in applying what you’ve learned so far?” - swap that language out for something about what they should be able to achieve from the learning objectives. I’m remaining broad here for the sake of the post 🙂

  • If they rate the scale 3 or below, provide automatic feedback which points the learner to specific resources to further explore, or instructs them on how to find more live support (a chatbot on the support site, a dedicated point of contact, or the instructor, depending on your setting).

  • Aggregated data could also provide you with great feedback on where learners are getting stuck in the course, as an added bonus!

• Live learning: In live webinars, instead of relying on participants taking initiative to post questions in the chat, it could be helpful to give participants a chance to participate in a poll to express how they are feeling about the content. Sometimes I’ll just give them a plain Likert scale, sometimes I’ll use emoji along side descriptions to help them express their emotions: 😰 😵‍💫 😎 🥱

  • This can provide just-in-time information so that you can increase/decrease the pace, pause and review, or pivot topics if needed.

Less self-direction in what to learn, more self-direction in how to learn it

As already mentioned in this post, I’ve seen a trend of training and L&D teams shifting more toward microlearning and self-paced content, and lots of choice in what to learn in one modality.

What if, instead, we offered fewer learning paths, but multiple ways to achieve them?

• Offer key topics as both live training and self-paced courses.

• Self-paced courses are often incredibly linear. They lack opportunity for adult learners to self-direct their learning path through a course. Where possible, offer branching options to encourage learners to participate agentically within the learning content. Simply asking a learner “Would you prefer to watch a video or read a blog post about X topic?” and then providing the appropriate content based on their response is a straightforward way to start implementing more agency within learning.

Incorporating learner choice into continuing learning

Once learners have completed a live or self-paced training, offer them recommendations for the next steps based on their specific use case/needs or desired format.

For example, after completing an introductory product course, a learner might be presented with options to either:

• Check out a more technical deep dive in a live webinar or in relevant help articles

• Learn implementation best practices in a live webinar or in a video recording

Wrapping up

YO, this got LONG (for me) 😅. I’m just seeing a lot of clicking-as-engagement out there, folks, and I think we can do a better job of examining what types of engagement we are encouraging in our learning and how we might better encourage all types of engagement.

Engaged learning is more effective. If we can be more intentional and well-rounded about how we’re engaging learners, and even connect it to adult learning theory, we’ll all be better served. 🫡

Signing off for today, thanks for reading till the end!

During a recent meeting for one of my current projects, we were discussing the importance of having a safe space for failure in learning. It reminded me of the newsletter I wrote back in December 2023 about the idea of programming failure in our learning design. There are a lot of new faces around here since then, so I figured it might be a good time to share this again. Thanks for reading.

I think about learning when I’m weightlifting a lot. Is that weird? Over the summer, I tried to hit a new squat PR, and I failed. I had to completely drop the barbell because, well, my legs did not want to stand me back up again. As of a few weeks ago, I was able to squat that weight for four reps successfully [Summer 2025 addition: this is now a warm up weight for me]. So what happened in between, and how does this connect to failure being important to learning?

When you’re lifting weights, you can do a lot of the movements without really knowing the movement you’re doing. Here’s an Instagram reel that talks about doing a movement vs. using a muscle to perform the movement:

Oh, you didn’t know you were signing up for a newsletter about exercising, you say?

What I learned from failing a squat

I really had to reassess what I knew about squatting and what made my squats successful. In practicing squatting at lower weights, I made note of:

• The points at which each foot drove into the ground (not just from the heel)

• Actively pushing my knees out to engage the glute medius

• The feeling of driving the floor away vs. using my back to lift the weight up/stand up

• My focus on keeping my core tight and my back tight

Simply put, failing my squat forced me to examine how I was squatting and where I could improve. I took those improvements and put them into practice. I learned. I could probably have kept on squatting at lower weights doing what I was doing–the failure pushed me to better understand what I was doing.

When do we encourage failure in our learners?

Most of our first experiences with formal learning occur in school, where success is incentivized (through grades) and failure is penalized (also through grades). This doesn’t leave a lot of room for failure, and I think, leaves the opportunity for a deeper learning on the table. Being wrong about something, failing something, offers us a better opportunity to understand when we are right and why. When learning skills like learning a new tech tool, I believe failure can give us the confidence we need to explore and solve problems, especially when we encounter a behavior that is unexpected.

I’ve worked with so many people that are afraid to try things in technology because they’re afraid they’ll break something. I promise, you (usually) won’t!1 Sometimes, people are embarrassed to walk me through what they are doing when I’m helping them with a tech tool “because they might mess up.” That’s okay! Messing up means we can learn!

Generally, in formal learning, I don’t think we encourage failure enough.

Subscribe now

How can we encourage failure in our learners?

Battling the mindset that failure should be avoided is pretty tough. The formal education system really grinds the idea into people’s minds that being wrong is bad. So how can we encourage failure, and learning from that failure, in our asynchronous and synchronous training? Here are just some ideas to start:

Synchronous training

• Share your mistakes: Be candid with your learners! Share an example where you did something wrong and explain how you corrected it. This can help normalize failure for your learners.

• Ask learners to share their mistakes: Then, ask your learners to share something similar! This can be helpful if you have power-user attending training especially. It demonstrates that even successful power users had to start somewhere.

• Share a “reset point” with learners: This would vary from product to product, but if a user messes something up, where can they go back so that they don’t need to start all the way over but they can wipe out any mistakes they’ve made?

• Explicitly share resources and where they can help if learners encounter an issue.

Asynchronous training

• Allow quiz retakes: If learners are taking some kind of formative assessment like a quiz, allow many re-takes. This can force learners to review their responses and search for the correct responses–hopefully, in the end, better understanding what makes the right answer correct.

• Encourage reflections: Offer opportunities for reflection. Ask the learners to submit how trying a new tool felt for the first time: did they succeed? Would they do something differently next time?

• Discuss mistakes as a community: Create a community space specifically to discuss mistakes. Learners can share where they struggled, and tips for others on how to avoid the same errors.

The last thing we want our learners to do when they’re trying a new tool or strategy is to give up. Programming failure can help people learn what the best next steps are if they hit a road bump. What else can they try? Where can they get help? Success is not a linear journey, but learners need to know where they can turn should something unexpected happen that prevents them from reaching their goal.

If you’re not familiar with Bloom’s Taxonomy or haven’t come across it in a while, here’s a quick refresher. It’s defined by Wikipedia as a “framework for organizing educational goals.” Unfortunately, the framework of organizing these goals is often represented using a pyramid, which suggests a hierarchy of sorts:

Most educators will consult Bloom’s taxonomy when writing learning objectives for their lessons, looking for “action verbs” from which they can create measurable assessments. For instance, I can’t measure if a student “understands solving for slope,” (what does “understanding” look like?) but I can measure if a student can “solve the equation for slope.”

The problem with Bloom’s

The debate has been had time and time again: representing learning in this way suggests a hierarchy of sorts. According to Bloom’s critics, the verbs under the “remember” category, the so-called “lower level” thinking skills, are viewed as less important, while the verbs that fall under the “create” category are seen as the end-all, be-all goal. Educators are often pushed to include “higher level” Bloom’s verbs in their lessons to ensure intellectual rigor in learning.

It also suggests you’ll always learn in a specific order: first you remember something, then you apply, then analyze, etc. That’s not always the case, though; a more discovery-based learning activity might have a learner try creating something, analyzing that creation, and then gaining a deeper understanding of what the primary components are. Or, they may be learning something that requires them to hop all over Bloom’s categories.

Let me be clear: I agree with this critique of Bloom’s. I don’t thinking representing learning in a linear or hierarchical fashion is helpful. It’s not useful to do a bunch of remembering (“lower-level Bloom’s”) and then do nothing with that knowledge; at the same time, it’s not helpful to create a bunch of stuff (“higher-level Bloom’s”) without a deeper understanding of the what, why, and how behind the creation.

However, I do think Bloom’s can be helpful. In learning design, it can help us more clearly define goals and understand the type of thinking required to meet them.

The problem here isn’t Bloom’s, then—it’s the linear fashion we’re using to interpret the taxonomy. I think, as humans, we tend to default to seeing things as linear. Thus, learning is linear: we have chapters that move forward in textbooks. We work our way up from Calc I, to Calc II, to Calc III and Linear Algebra. We set up our modules in our Learning Management Systems numbered 1, 2, and 3.

We’ve put Bloom’s into this bucket as well: first we remember, then we apply, then we analyze and evaluate. Neat steps, moving forward.

Breaking out of linear education

For anyone who didn’t subscribe to this newsletter when I first released it, it might be news to you that I’ve been taking cello lessons. I don’t have any musical background, really, except playing flute for a little in 4th grade (I have no actual memory of this).

More recently, I started reading the book Uncommon Measure: A Journey Through Music, Performance, and the Science of Time by Natalie Hodges recently. One quote really struck me, in thinking about not only learning cello but about learning in general:

Rondos have the mesmerizing effect of moving you linearly and circularly through time, at the same time: By progressing from theme to theme through the development sections, the rondo consistently brings you around and back again to the music with which you began.

Learning cello is, quite obviously to me as a learner, both linear and circular. I progress through learning new skills and positions while coming back to and re-learning those same concepts in a new way: the D-major scale in a new octave. Counting triplets vs. counting quarter notes or eighth notes (I’m still pretty beginner here people).

And then I got this anonymous feedback from my end of course eval, emphasis my own (another reminder, I teach an asynchronous course part-time for a local university):

I think I learned the most from the Unit Projects because they made me think about all the modules in conjunction with each other….but all of them still taught me to think about how each topic flowed into each other. Especially in online classes, it's very easy to forget what the previous module was about, so the unit projects eliminated that by always having us think about the unit as a whole.  

In rondos, the listener and the player are aware on some level that they’re progressing through the circle. You’re moving linearly toward the end of the song, but you have a sense of its spiraling, repeating nature. I think this spiral is what gets lost in education.

How can we make the circle of learning more visible?

Revealing the spiral

I STILL PRESENT EVERYTHING SO LINEARLY IN MY OWN COURSE. GAH!

Here’s one example, at the beginning of the course, when I’m giving students a run down of what they’re about to learn for the term:

and I typically include as the first slide in each brief lecture video a quick run down of how all of the immediately recent/forthcoming content fits together:

I’m clearly trying to make the relationship between different topics stronger and more connected, but I fear they’re still disjointed being shown only in a linear fashion (layer this on by putting all course content and assignments into the extremely-linear LMS modules).

What is the rondo for learning?

I get that some students (based on that anonymous comment) are seeing how our course topics fit together as we loop back around to connected concepts. But as I’ve thought about Bloom’s, the Uncommon Measure quote, and my student’s comment, I’ve really been thinking about how I can practice this in my learning design more intentionally.

How can I make it almost expected or anticipated: bringing learners back to the familiar, but each time, possessing a larger network of knowledge with which they can make connections?

From linear to spiral

I don’t really have an answer for this yet, but it’s something I want to continue thinking about. What are anchoring themes I could write into curriculum for students to revisit with a regular cadence—that they can build upon, tweak, and manipulate?

I’m reminded of something Justin Cerenzia wrote over at The Academic DJ recently about a structured constructive retrieval and elaboration intervention. I recommend you check it out in full, but he discusses having students revisit portions of a predetermined list of course concepts every eight days, in the same structured activity.

To me, the idea of the learning spiral provides not only more opportunities for reinforcement, but can exist an anchoring schema which learners can use to more explicitly build and grow their conceptual connections.

Concluding questions

As I often do, I’m going to end with questions to consider the next time you’re designing learning, whether it’s a one-hour workshop, a full semester course, or a series of training videos.

• How can I create a learning spiral with anchoring themes?

• What content signals can I use to help indicate this reinforcement?

• What activities can I incorporate to reinforce the spiral more intentionally?

• What can I incorporate into my existing summative activities to more intentionally include the spiral?

Have you tried to incorporate a more circular approach to learning in your teaching or training? Share your strategies with me in the comments!

Thanks for reading and learning with me 👋🏻