Case Lab: how a maths teacher became Udemy's top coding instructor — and the part the headline skips (2026)

The faceless-channel case was fast and fragile; Senja was a slow software compounder. This one is a third shape entirely — sell what you already know how to teach. Rob Percival, a Cambridge maths graduate and former UK high-school maths teacher, posted his first course on Udemy at $199 in June 2014, made one sale in the first 24 hours, and went on to 2M+ students and a reported standing as Udemy’s highest-earning instructor. It’s a clean case to dissect — including the parts the headline number skips.

1. The facts & verification

• Who: Rob Percival — Cambridge maths degree, former UK high-school maths teacher. Before courses he built a web-hosting side business plus freelance web and app development, and it was that income that let him leave teaching.
• What: “The Complete Web Developer Course,” posted on Udemy at $199 in June 2014 — which made just one sale in its first 24 hours, against a stated goal of $5,000/month. It grew into a catalogue under his Codestars brand.
• Numbers: 2M+ students (2.5M+ by some sources); reported as Udemy’s highest-earning instructor; reported revenue from over $2M (earlier coverage) to over $5M (more recent, e.g. a SaaS Club interview framed around “$5 Million With Online Coding Courses”).
• Verification: reasonable — a public Udemy profile (students, ratings, catalogue) plus a recorded podcast interview where he tells the story himself. Better than a screenshot; not audited accounts.

2. The niche & why it works

The niche is the oldest one there is: teach a skill people will pay to learn. Percival’s edge was the overlap of three things most people only have two of — real domain expertise (he could actually build the web/app projects), teaching ability (a trained maths teacher, used to explaining hard things to beginners), and a topic with huge, durable demand (learning to code). Coding sits in a sweet spot: hard enough that people will pay for guidance, common enough that the audience is enormous, evergreen enough that a good course keeps selling. That’s the textbook shape for selling online courses solo.

3. The income model

Paid courses on a marketplace — Udemy. The economics are appealing on paper: build a course once, sell it to thousands, earn while you sleep. But marketplace course income has its own shape worth naming: the platform takes a cut and largely controls pricing and discounting, so volume and catalogue depth matter more than a single hero course. Percival’s answer was a brand and a catalogue (Codestars) rather than one product — the same “portfolio, not one hit” logic we keep seeing. It’s one rung on the broader ladder of how solopreneurs make money, and like most creator income it’s cross-border (students worldwide, paid in different currencies) — the usual declare-it-and-bank-it-properly story.

4. The pattern (stripped of luck)

The repeatable system underneath the headline:

1. Earn the expertise first. He could build real projects and had already gone full-time freelance before teaching them — the credibility came before the course.
2. Pair it with teaching ability. Domain skill alone isn’t enough; explaining it clearly to beginners is the actual product.
3. Ship early, even badly. One sale on day one. He published and improved rather than polishing for a year.
4. Iterate from real feedback. A live course on a marketplace gives you reviews and refund signals fast — he used them.
5. Build a brand and a catalogue, not one course. Codestars turned a single product into a compounding library.

5. The tool stack

Course creation is unusually low-tech, which is part of the appeal. The categories a solo needs:

• A course platform — Udemy (marketplace reach, built-in audience, platform cut) versus a self-hosted platform (you own pricing and the customer). The trade-off is laid out in best course platforms for solopreneurs.
• Recording & editing — screen capture for code-along lessons, a decent mic, simple editing.
• The actual skill being taught — in his case a working dev environment, because the lessons are building real projects.

The point isn’t the specific logos; it’s that the barrier to entry is teaching ability and a recording setup, not capital.

The honest read (gross vs net, and survivorship)

• Gross isn’t take-home. A figure like “$5M” is lifetime/gross revenue, reported, not audited — and before Udemy’s cut, before the ~4 staff salaries, before costs, and before tax. The number that lands with him is a fraction of the headline. Still excellent; not “$5M in his pocket.”
• Survivorship — and this one is loud. Creator income is among the most top-heavy distributions there is. For every top-earning instructor there are thousands of courses that sell a handful of copies. The case proves the ceiling exists; the average course creator earns modestly, and many earn close to nothing.
• Timing was part of it. 2014 Udemy was a different, emptier market than today’s. The playbook still works, but assuming today’s conditions match his is the fast way to disappointment.
• It stopped being solo. Honest caveat: the model scaled into a small team. A true one-person version is possible, but this result included employees.

What we take from it

The transferable asset is the operating system, not the revenue line: expertise + teaching ability → ship early → iterate from feedback → build a brand and catalogue. That’s genuinely repeatable for anyone who actually knows a skill worth learning — and far more durable than chasing a trend. Just hold two truths at once: the system is real, and most course creators earn modestly, so set the expectation at “a solid income stream,” not “Udemy’s top instructor.” If you’re on this path you’re a creator or expert building owned teaching assets; the build is half of it, and the money/legal spine (banking the cross-border income, declaring it) is the unglamorous other half.

The takeaway

• Reasonably verified case: an ex-maths-teacher made one sale in his first 24 hours at $199 (June 2014) and grew to 2M+ students as a reported top Udemy instructor — Udemy profile + a recorded interview as the sources.
• The model: real expertise × teaching ability × ship-and-iterate × a brand/catalogue (Codestars), not a single hero course.
• Gross ≠ net: “$5M” is reported gross/lifetime, before Udemy’s cut, ~4 staff, costs and tax.
• Verified ≠ repeatable: creator income is top-heavy, most course creators earn modestly, and 2014 timing was unusually favourable.
• Honest caveat: it scaled into a small team, not a pure solo — and the transferable bit is the system, not the headline number.

Sources for the case facts: Rob Percival’s public Udemy profile and a SaaS Club podcast interview. Figures are as reported there, not audited accounts; treat them as the shape of the economics.

Part of the guide to building a one-person business.