The user journey: from first touch to payment
How we grew course-page conversion to payment systematically: we watched the metrics of the whole funnel, tested hypotheses and scaled the ones that worked to the whole lineup
3%+
course-page conversion to payment, against a market average of 1–1.5%
1–3% → ~10%
request-form conversion on the pages where it was lowest
10+
product pages launched and updated
Context and goal
The product
A group of EdTech companies: new courses and cohorts launch every season. Each course has its own page, and that page is the main sales channel: it turns traffic into requests and revenue.
The goal
Grow conversion along the whole path to payment — visit → request → payment — shorten the deal cycle, and scale what works to the whole course lineup.
My role
I ran the whole path together with the designer: I owned the product logic, page structure and page metrics, the hypotheses and user testing; the designer made the layouts, and an editor helped with the copy.
Approach: from funnel metrics to scaling
We worked in a cycle. We watched the metrics of the whole funnel and found the step where people dropped off. We formulated a hypothesis and tested it on one page. The metrics decided whether to scale the change to the other pages or roll it back.
Metrics of the whole funnel
Visit, request, choosing a tier, payment: you can see at which step people drop off. Separately, the deal cycle: how long it takes from the first visit to payment and what share of purchases falls on the last days before the course starts.
Hypothesis and test
A hypothesis went live on one page. New programmes were sometimes shown to loyal alumni before launch: 10–20 people selected by LTV and activity answered within 1–2 days whether they wanted to sign up, why, and what they would change.
Decision and scaling
If the metrics grew, the change was rolled out to the other pages; if not, it was rolled back. Alumni feedback went into the layout before the build.
This works for any product sold through a page: SaaS, a subscription service, a marketplace.



Hypothesis 1: a first screen with the course result will bring more people to the pricing tiers
The first screen is the first drop-off point in a page funnel: this is where a visitor decides whether to keep reading.
Why we thought so. A page that describes a course in general terms doesn’t show what people will get, and visitors leave before reaching the pricing tiers.
What we did. Every page starts with the course result. On Figma Motion Ninja the first screen mirrors working in Figma: designers’ cursors move the labels. On Vibe Coding Master it’s a code editor and a programme laid out as the steps of building a service, with the author’s experience backed by companies and launched products.
What came of it. The Figma Motion Ninja group filled up completely. The approach was rolled out to every new page — each one starts with what a person will be able to do after the course — and groups consistently filled to 90% and more.

› See the whole Figma Motion Ninja page scroll
› See the whole Vibe Coding Master page scroll
Hypothesis 2: removing the extras from requests and payment will bring more people to a purchase
The request and the payment are the last funnel steps: this is where it’s easy to lose a purchase that has almost happened.
What got in the way. On the pages where conversion was lowest, request forms converted at only 1–3%. At the payment step some payments failed and the buyer couldn’t see what went wrong; the promo-code field was hidden, and extra instalment options confused people.
What we did. We reworked the request forms. At the payment step we added payment retries and a clear error message, moved the promo-code field into view and kept instalments only where they help. For companies — payment by invoice and a separate page with a request form.
What came of it. Request-form conversion on those pages grew from 1–3% to about 10%. The payment step is shared by all courses, so the changes worked across the whole lineup at once.
Hypothesis 3: a real deadline will shorten the deal cycle better than a discount
An expensive course has a long deal cycle and a short sales window: people put the purchase off until the last moment.
What got in the way. Visitors put the decision off, and part of the demand doesn’t turn into payment before the course starts.
What we did. The price rose in steps, each with its own date on the page. When enrolment is closed, the page switches to a waitlist and keeps demand until the next cohort. And at the moment of payment we offered a second course 20% cheaper and course bundles to raise revenue per buyer.
What came of it. Most sales of a short course came just before the start date: a real deadline worked better than the price.
The result. Course pages convert to payment at 3% and above, against a market average of 1–1.5%. I launched and updated 10+ pages on this cycle.
Which hypotheses worked
- 01
If the first screen shows the course result, more people reach the purchase. The Figma Motion Ninja group filled up completely, and once the approach was rolled out to every page, groups consistently filled to 90% and more.
- 02
If you remove the extras from requests and payment, more people reach the purchase. Request-form conversion on the pages where it was lowest grew from 1–3% to about 10%.
- 03
If a purchase has a real deadline, people decide faster than with a discount. Most sales came in the last days before the course started.
- 04
If you offer a second course right at payment, people buy more. A 20% discount on a second course and course bundles raised revenue per buyer.
«He treated the website as an acquisition product and controlled the funnel stages and payment scenarios that matter most to the user, which directly affected how much revenue ultimately reached the P&L.»