ClassLog

I taught group fitness classes, so this was a problem I lived. Every month I did what every freelance instructor does — reconstructed what I'd taught from my calendar and a stack of WhatsApp groups, did the maths by hand, and built different invoices in different formats for different venues. I'm a product manager. I decided to treat my own annoyance as a product.

I started with the problem, not the interface. I wanted to know whether my experience was structural or just my own poor organisation, and the research was unambiguous. A 2024 survey of UK group-exercise instructors found 24% manage up to seven separate WhatsApp groups for their schedule, and more than half avoid taking time off because arranging cover is too stressful. The nearest competitor — Striive, which had positioned itself almost exactly here — had already built, shipped and shut down. That told me two things at once: the niche is real, and the go-to-market is hard. So I wrote it up properly — a PRD with a sharp thesis ("eliminate the freelance friction"), personas, the three problem pillars, and functional requirements down to 15-minute grid snapping and self-bill document generation — before I drew a single screen.

The strategic angle I kept coming back to was Making Tax Digital. It's easy to pitch a tool like this as a convenience — save an instructor a few hours a month — but the more defensible wedge is legislation. MTD for Income Tax is being phased in on falling thresholds, and a full-time instructor teaching across several venues is squarely in scope by 2028. The catch is that these are small business owners who don't understand the rules and won't engage with them until they're penalised for missing them. A tool that quietly keeps their records HMRC-ready — because they were logging their classes anyway — stops being a nice-to-have and becomes the thing that keeps them out of trouble. That reframes the whole product: not "an admin app," but "how you stay compliant without having to think about it."

The decision I'm most deliberate about is terminology, and it sits on the roadmap as a real risk. "Self-bill" is the legally correct term for how most of these venue arrangements work, but instructors don't understand it, and getting it wrong erodes trust in a product that touches their tax. Rather than assume, I scoped it as something to A/B test — "Self-Bill" against plainer language like "Automated Payments" — because in a money product the words are part of the UX. The second call was form factor: this is a tool used in two completely different modes, so I designed it mobile-first for the poolside and the gym floor, and gave it a genuine desktop view for the month-end admin session, rather than stretching one to cover both.

The part I'd point to as the real work is the design system. The first cut leaned on a high-energy electric lime, and it cheapened a product that's fundamentally about someone's money and livelihood — it read like an energy drink, not a tool you'd trust with your tax. So I evolved the palette: kept the Midnight Navy core and the athletic Barlow Condensed typography, but moved the accent to a warm gold, which still reads "Elite Performance" (podium, gold standard) while feeling premium and trustworthy rather than loud. It's a small change that does a lot of work, and it's exactly the kind of judgement that separates a prototype that demos well from a product people would actually pay for.

ClassLog is a concept, and I'm honest about that. It exists as a full PRD, a design system, interactive prototypes, and a pitch deck — Phase 1. The next milestone is an MVP with a cohort of instructors to stress-test the scheduling and billing logic against real, messy schedules. I built it to prove a specific thing: that I can take a genuine problem I understood from the inside and carry it end to end — research, requirements, design system, prototype, story — into something you could put in front of a room and fund.