There's a persistent myth in tech that great products require perfect conditions. A funded team, a dedicated sprint, a whiteboard, a roadmap, three months of runway. The reality is almost the opposite. Some of the most compelling things being built right now are emerging from the margins: the gaps between meetings, the quiet hours after the day job wraps, the train ride where your phone becomes a prototyping tool. These margins aren't a limitation. They're an advantage.
I've spent fifteen years in advertising agencies. I know what full-time creative production looks like. It looks like eight-hour days, weekly status meetings, three rounds of revisions, and maybe one good idea making it to market per quarter. Now I'm building products in sessions that rarely exceed three hours, and I've shipped over twenty of them. The math doesn't add up unless you accept that time was never the real bottleneck.
Why constraints sharpen everything
When you have two hours to build something, you don't spend forty-five minutes debating the database schema. You don't prototype three different navigation patterns to see which one "feels right." You make decisions fast, ship fast, and learn fast. Constraints strip away the performative parts of building and leave only the essential question: does this thing work?
This isn't a new insight. Designers have known for decades that a blank canvas is harder than a brief with parameters. But it's one thing to understand that intellectually and another to experience it at 11pm when you're trying to get a London Pub Guide live before you lose momentum. That kind of time pressure produces a brutal clarity about what actually matters and what's just polish.
The agency world taught me strategy, positioning, audience thinking, brand architecture. All of it valuable. But it also taught me that work expands to fill whatever time is available. Give a team a week and the work takes a week. Give a team a month and it takes a month. The work itself doesn't change. The overhead around it does. Margin-building strips out the overhead entirely.
Single-session products
Some of the best things I've shipped came from single sessions. One sitting. Laptop open to laptop closed. The Forest Quiz (a Nottingham Forest trivia game I built because I love football and wanted something fun to exist) was a single evening. Three hours, give or take. First Out, a tube exit guide for London commuters, started the same way. So did the first version of CultureTerminal.
Single-session products have a particular quality. They're tight. There's no feature creep because there was no time for features to creep. They do one thing, they do it clearly, and they're done. The multi-session products (Modern Retro, Trove) are richer and more complex, but the single-session ones often have a purity that's hard to achieve when you over-think. Speed is a design choice, and margins force you to make it.
The margin builder's toolkit
None of this would be possible without the current generation of AI tools. Claude Code has fundamentally changed what a non-engineer can ship in a short session. I'm not writing code in the traditional sense. I'm describing what I want to exist and collaborating with an AI that can translate strategy into working software. That's a new kind of building, and it's perfectly suited to margin time because it compresses the feedback loop from days to minutes.
But the tools are only half the story. The other half is taste. Fifteen years of evaluating creative work, understanding audiences, and thinking about positioning means I can make product decisions quickly. I know what good looks like. I know when something is overbuilt and when it's undercooked. That intuition, combined with fast tooling, means a two-hour session can genuinely produce something worth using.
What margin-building teaches you about priorities
When time is genuinely scarce, you learn what matters. Not in a motivational poster way. In a practical, immediate way: you have two hours and you need to decide what to spend them on.
You learn that design systems are a luxury. You learn that the perfect database schema is irrelevant if nobody ever sees the product. You learn that a slightly wonky mobile layout is fine if the content is good. You learn that done is better than perfect, not because someone told you that on a podcast, but because the session is ending and you're either deploying this thing now or shelving it until the next window opens.
You also learn what you actually care about. When time is abundant, everything seems worth doing. When it's scarce, only the things that genuinely excite you survive the filter. I've had dozens of product ideas. The ones I actually built are the ones I couldn't stop thinking about. The others died natural deaths, which is exactly what should happen to ideas that aren't strong enough to compete for limited building time.
The case for imperfect conditions
The startup world glorifies the garage. The maker community celebrates the side project. But most advice still assumes you need to carve out dedicated, protected time to do meaningful work. I'd argue the opposite. The lack of protected time is what makes the work meaningful. It forces you to be honest about what's worth building. It keeps you from over-engineering. It rewards decisiveness over deliberation.
A few months ago I was a strategy director between roles, spending evenings scrolling feeds and feeling like I should be doing something more productive. Now I've got twenty-four live products, a portfolio that demonstrates real building ability, and the genuine capacity to turn an idea into a working thing in a single session. That transformation didn't happen because I suddenly found more time. It happened because I started treating the margins as the main event.
The products are the proof. And the fact that they were built in gaps, in short sessions, under real constraints? That's not the weakness of the work. It's the whole point. The margins are where the most honest building happens, because there's no room for anything except the work itself.
