From sales call to self-serve student

Product School runs one of the largest product management communities in the world — courses, certifications, events. When I joined, enrolling in a certification meant scheduling a call with sales. The course pages were thin. There was no pricing, no schedule, no way to commit without talking to a human first.

I led the end-to-end design across the full enrollment ecosystem: the cohort component where students make their first real decision, the self-serve checkout that removed sales from the critical path, and the LMS that brought everything a student needed into one place.

When I joined, enrolling in a Product School certification meant one thing: scheduling a call with sales. The course pages were thin — no pricing, no schedule, no way to commit. If a prospective student wanted to know what they were signing up for, they had to talk to a human first.

This wasn't entirely wrong. High-ticket ed-tech products often need that conversation. But it created a real bottleneck. The sales team was spending time on students who weren't ready. Interested students were dropping off before ever reaching them. And we had no visibility into where or why.

That last part is what drove the approach from the start: before redesigning anything, we needed to be able to measure it.

The cohort component

This is the component a prospective student sees before anything else in the checkout process. It looks simple: a list of upcoming cohorts. Pick one, enroll. But it sat at the intersection of almost every tension in the product — how much information is too much, whose name goes first, and for a long time, which action to prioritise: speak to sales, or enroll directly.

When I joined, the component showed almost nothing. Dates, availability, a button. No instructor, no price, no context.

Every change after that started with a hypothesis. We'd form one — showing the instructor increases trust and CTR — build a version, measure it against click-through and conversion, and run qualitative interviews to understand the why behind the numbers. Then do it again. Layout changed. Information density changed. The instructor went from a text line to a photo to a full credential. Filters appeared, disappeared, came back differently. The CTA itself was contested for a long time — whether to lead with "schedule a call" or "enroll now," and what it meant for the business each way.

What the data kept confirming: the instructor is the main selling point. Not the dates, not the format. Who is teaching this, and where did they work? That's what moved people forward.

The current version — course name, level badge, dates, instructor photo and title, spots left, enroll CTA — is the result of years of that loop. Not a redesign. A continuous optimisation.

The checkout

This was the 0-to-1 moment. The decision to let students enroll directly — no sales call — changed the business model. And it didn't land without friction internally.

Sales was sceptical. Marketing had concerns. There was a real question about whether self-serve would cannibalise the pipeline or complement it. We shipped it anyway, carefully, and let the data make the argument.

The checkout itself went through more iterations than any other surface. The goal was always the same: reduce friction without reducing confidence. A student committing $3,000–5,000 needs to feel sure, not rushed. So every change was measured carefully — number of steps, how information was grouped, how interactive components behaved, what the order summary showed and when.

The hardest moment in the flow has always been payment plan selection. Upfront, installments, financing, bundles — multiple certifications at different price points, multiple ways to pay, real money on the line. We never made drastic changes here. We'd form a hypothesis, build, measure, listen to sales calls, collect qualitative feedback, then move — slowly and deliberately.

What changed over time wasn't just the design. It was the relationship with the feature. Early on, the sales team treated self-serve as a competitor. Gradually — as revenue came through and the data became undeniable — it became an asset. They started using the checkout to capture abandoned cart leads. We built and iterated on attribution rules to make sure everyone's contribution was visible. What started as internal scepticism turned into shared ownership.

The LMS

For a long time, being a Product School student meant navigating a patchwork of tools. The certification was here. The materials were there. Calendar invites came from somewhere else. It worked, but only just — and students felt the seams.

It started small. A magic link to share decks with an enterprise client. Then it grew. A subscription plan launched, and suddenly there were cohort enrollments to manage, perks to surface, content to gate. The portal existed, but it kept handing students off elsewhere — to download something, to pay a bill, to join a call. You'd log in to Product School and immediately leave Product School.

The work was to close that gap, piece by piece. Each section of the student journey — cohort access, course materials, membership perks, community — brought into one place, with consistent navigation and a clear sense of where you are and what's next.

The design challenge wasn't just visual. It was about trust. A student who's paid several thousand dollars and keeps getting redirected starts to question whether the product is as premium as the price suggests. Seamless access is a product value, not just a UX preference.

What started as a certification page with a "book a call" button is now a self-serve enrollment and learning ecosystem.

Analytics and experimentation weren't a phase of this project — they were the method. We didn't redesign the cohort component because it felt off. We redesigned it because we could see where students were leaving. We didn't ship the payment plan UI once — we iterated on it. The data kept us honest about what was actually working and what we just hoped was working.