CELab: The Customer Education Lab

In this episode of CELab, we welcome content creator Kevin Lee to unpack how “scrappy” social-style video can transform Customer Education and scaled Customer Success. Kevin shares his journey from running a Rubik’s Cube meme account and a 200K-subscriber YouTube channel to launching Atlassian Answered, a video series that has generated over 2 million organic impressions and hundreds of thousands of views—without paid promotion. Along the way, he explains why most B2B content fails to land, how to “earn” attention instead of assuming it, and why simplifying corporate jargon is like teaching calculus to someone who’s still in geometry.

Together, we dig into the craft behind effective educational content: from scripting for real humans instead of internal stakeholders, to using video as a scaled “virtual CSM,” to practical production tactics like no-pause editing, intentional cropping, and focusing on audio over fancy visuals. They also explore how employee-generated content (EGC), personal branding, and authentic on-camera presence can amplify product education, drive adoption, and build trust at scale. If you’re in Customer Education, Customer Success, or Product Marketing and you’re serious about using content to move the needle, this conversation will give you a playbook you can start testing this week.

Highlights:

How to turn complex B2B products into simple, relatable stories your customers actually get

Why audio quality, pacing, and presence matter more than cinematic visuals

The “geometry vs. calculus” model for simplifying corporate jargon

Using employee-generated content (EGC) and personal brands to amplify your product education

Practical editing tricks: no-pause audio, crop-in/crop-out, and using text sparingly but strategically

How to repurpose video across the journey: onboarding, in-product guidance, help center, and renewal

Customer education has outgrown the LMS era. In this episode of CELab we down with Zayd Badwan, Chief Revenue & Customer Officer at Uplimit, and Yash Tekriwal, who leads education and cohort programs at Clay, to ask a hard question: if your main metric is still course completion, are you already behind? Together, they unpack how AI, cohort-based learning, and practice-first design are transforming customer education from “check-the-box training” into a true lever for product adoption and revenue.

You’ll hear how Uplimit and Clay are using AI to automate the admin work that slows teams down, generate practice experiences at scale, and actually measure behavior change and skill improvement—not just time spent watching videos. The conversation ranges from cohort-based onboarding for enterprise customers, to AI-powered role plays and rubrics, to tying learning programs directly to product usage and activation metrics. If you care about Customer Education, Customer Success, or Enablement, this episode is a blueprint for moving beyond the hype and building programs your CFO—and your customers—will love.

Key Takeways:

The “sucker’s choice” in learning: scale with async and get low outcomes, or deliver high-touch experiences that don’t scale—and how AI changes that tradeoff.

Why completion rates aren’t enough: shifting from “who finished the course?” to “who can actually do the behavior that drives product value?”

Cohort-based customer onboarding:

Clay’s enterprise cohorts that get customers to a live, deployed workflow before their first call with a strategist.

Using “onboarding complete” as a signal that a customer is ready for strategic work, not just basic setup.

AI-powered practice at scale:

Role plays, brainstorms, and screen-recording uploads graded by AI against a rubric.

GE Healthcare’s example: thousands of SME hours saved by automating role play assessment.

From content consumption to competency:

Designing exercises that assess how well someone understands and applies concepts, not just whether they watched a video.

Moving from binary “activated / not activated” to a spectrum of competency and product proficiency.

Admin and operations automation:

AI grouping learners, tracking participation, and sending smart nudges in Slack/Teams.

AI-assisted course creation from slide decks, product docs, or talks—getting teams 80–90% of the way there.

Behavior change as the real goal:

Measuring skill improvement (e.g., 12% skill uplift) and tying it to time-to-value, product usage, and revenue.

Using product data (activation, feature usage, MAU) together with learning analytics to prove CE/CS impact.

Social and community-based learning:

Learners coaching each other in Slack, not just relying on CSMs or TAs.

Encouraging learners to share their wins publicly (e.g., LinkedIn posts about what they built) to drive advocacy and organic growth.

Personalization with AI “tutors”:

Tailoring scenarios and role plays to learner role, industry, and use case (e.g., government, specific product surface areas).

Cutting down “re-learn what I already know” time and focusing on the delta that actually matters.

Strategic implication for CS leaders:

Free...

In this episode of the Customer Education Lab, Adam Avramescu sits down with his Personio teammate and Senior Learning Experience Designer Roberto Aiello for a deep dive into what modern customer education really looks like in 2026. They unpack how an international company with an English working language and a massive German market grapples with localization, terminology, and tone—including why “just translate it” fails, and how to design for truly native experiences instead. Roberto shares concrete stories from Personio’s Voyager Academy, like localizing character names, handling UI screenshots in multiple languages, and deciding what not to translate.

From there, the conversation shifts into AI-assisted content creation and distribution. Roberto walks through how he used AI to co-create a compensation course—turning scarce SME time into a solid 60–70% draft—and then scaled that approach into reusable assistants for brainstorming, course design, and promotional comms. Adam and Roberto also dig into how they mapped academy content to activation milestones, setup wizards, and implementation email journeys, so customers see the right learning at exactly the right moment. If you care about Customer Education, Digital CS, or scaled CX, this episode is a masterclass in moving from “more content” to measurable customer outcomes.

Episode Highlights:

Why localization ≠ translation, and how cultural nuance (e.g., The Simpsons in Italy) changes tone, jokes, and stereotypes.

The trap of forcing English tone and structure into other languages—and why native reviewers must have the final say.

How Personio decides what terminology to localize vs. leave in English (“Dinglish”), especially in tech and HR products.

Practical workflow: producing content in English, then using vendors + native reviewers + AI to localize efficiently.

Using AI assistants (via langdock) to:

Draft HR skill courses (like compensation) from SME interviews and public domain expertise.

Encode style guides, glossaries, and frameworks so drafts are on-brand and consistent.

Generate promo copy (emails, posts) as part of a repeatable, scalable workflow.

The real benefit of AI: freeing up time for strategic thinking—like distribution, campaigns, and mapping to business outcomes.

Why translation memory and maintenance are harder than translation itself, and why integration across tools matters.

Leveraging tools like Parta for on-brand, collaborative content creation, and Clueso for scalable, localized video.

Personio’s AI Surge Week: workshop design to help non-experts adopt AI for content creation through live examples and iteration.

Rebuilding the essentials learning path by mapping:

Activation milestones 

Academy content 

Setup wizard steps 

Implementation email journeys

Example: surfacing targeted content at the people data upload step to reduce anxiety (GDPR, privacy, data scope).

Why “more content” is not the goal—surfacing the right content at the right time is.

The future: products that “talk” to custom...

Customer education is outgrowing the idea of “just an LMS.” In this episode, Dave Derington sits down with Mia Čomić, Product Marketing Manager at LearnWorlds, to unpack how the platform is evolving from a traditional course host into a true learning operating system for modern SaaS. Mia shares how her background in Comparative Literature, localization, and academy building shaped her perspective on learning as a continuous product experience—not a one‑off content drop.

We explore how AI is reshaping Customer Education: from speeding up content creation to orchestrating just‑in‑time learning, routing the right learner to the right asset at the right moment. They dig into the 2026 State of AI in Customer Education report, the idea of “structured professional balance” between humans and AI, and why completion rates are no longer enough. If you’re leading Customer Education, CS, or scaled enablement and wondering how to make AI practical (not gimmicky), this episode is for you.

Key topics & takeaways

Why LearnWorlds is shifting from “course host” to learning product operating system

The difference between a content library vs. a true learning system

Why Customer ed teams “ask for a library but really need a system”

Using AI for orchestration, not just content generation

What “instructionally intelligent AI” means in practice

The concept of structured professional balance: AI speed + human control

AI as an orchestration layer across content, product signals, support, and lifecycle

Customer education as a stress test for AI‑driven learning

Why completion rates are the wrong north star; outcomes and behavior change matter more

Budget realities for AI in CE and how teams are experimenting with small spends

How the 2026 State of AI in Customer Education Report was built (quant + expert interviews)

In this end-of-year special, Adam Avramescu and Dave Derington look back at 2025 through the lens of the things they love most outside of work: video games and live music. They unpack how titles like Helldivers 2, Peak, and Megabonk model great onboarding, community-based learning, and continuous engagement—and what Customer Education and Customer Success teams can steal from modern game design.

On the music side, they explore Louis Cole, King Gizzard & the Lizard Wizard, Black Country, New Road, St. Vincent, and Tune-Yards. Adam and Dave connect venue, context, and stagecraft to how we design learning environments, build resilient teams, and foster community around our programs. If you’re thinking about onboarding, product adoption, AI-era education, or how to keep your learners coming back, this episode is packed with fresh, unexpected parallels.

Key topic bullets (Customer Education & CS–focused)

Onboarding by design: How Helldivers 2 uses scaffolding, mission flow, and “muscle memory over menus” to teach complex mechanics—mirroring ideal SaaS onboarding.

Context > content: Why the venue (Small venue vs. big noisy hall) transformed Louis Cole vs. King Gizzard orchestral shows—and how that maps to learning environments and attention.

Performance support vs. upfront training: Games as a model for balancing early instruction with in-the-moment guidance.

Community-based learning: Peak as a metaphor for learners pulling each other up the mountain via collaboration and shared discovery.

Resilient teams after setbacks: Black Country, New Road rebuilding after losing their lead singer—parallels to CE teams navigating layoffs and org changes.

Emergent learning & retention: How Megabonk and other games keep players experimenting and returning, similar to driving ongoing product adoption.

Edges and surprise in learning: St. Vincent and Tune-Yards using surprise, constraints, and audience participation to create memorable experiences—vs. bland, predictable training.

AI and emergent workflows: Early thoughts on how AI products shift education from “where to click” to “how to think and prompt.”