There's a version of this story where I built something cool. Something with a slick landing page and a waitlist and a Product Hunt launch and a thread on X that went viral. Something that sounded impressive at dinner parties. An AI-powered something. A blockchain-enabled something else. The kind of product that gets covered in TechCrunch and forgotten six months later.
Instead, I built a tube exit guide. I built a pub directory. I built a weekend activity finder for parents with a young child. I built a Japanese restaurant guide for London. None of these will ever trend on Hacker News. None of them sound revolutionary when you describe them in a sentence. They're, by any startup standard, deeply unsexy problems.
And they're the best things I've ever made.
The glamour trap
In advertising, we used to call it "award bait" - the kind of work that was designed to impress other advertisers rather than solve a real problem for real people. Beautifully crafted, conceptually clever, strategically irrelevant. It won trophies and gathered dust. The clients loved the awards. The audience never noticed.
The tech world has its own version of this. Products that solve problems nobody has. Beautifully designed apps for workflows that don't exist. AI tools that automate things nobody was doing manually in the first place. They get funding, they get press, they get a standing ovation at demo day. And then they quietly shut down eighteen months later because it turns out solving imaginary problems isn't a sustainable business model.
I fell into this trap early on. When I first started building things with Claude Code, my instinct was to reach for the ambitious idea. The platform. The ecosystem. The thing that would require a deck and a vision statement and a five-year roadmap. But I kept coming back to the same question: would I actually use this? Not "could I imagine someone using this" but "would I, personally, reach for this on a Tuesday morning?"
The First Out origin story
First Out started because I was annoyed. I was standing on a Central line platform at Oxford Circus, crammed into the back carriage, and when the doors opened I was staring at a wall instead of a staircase. I had to walk the entire length of the platform, swimming against the current of people coming the other way, to get to the exit. This happens to millions of Londoners every single day. It's a tiny problem. A thirty-second inconvenience. But it happens over and over and over, and nobody had built a simple, clean tool to solve it.
So I did. Fifty stations. All eleven lines. Front, middle, or back of the train. Which carriage to stand in to be nearest the exit when the doors open. That's the entire product. It does one thing, and it does it well. It works offline because you need it underground where there is no signal. It loads instantly because when you're rushing for a connection you don't have time to wait for a spinner.
First Out will never be a unicorn. It'll never have a Series A. It'll never be the subject of a breathless profile in Wired. But it saves me thirty seconds twice a day, five days a week. That's roughly forty-three hours over a decade. Forty-three hours of my life, reclaimed from the mundane friction of navigating the underground. Multiply that by every person who uses it and you have something genuinely valuable, even if it never sounds impressive in a pitch.
The parenting problem
Little London exists because every Friday evening, my partner and I would have the same conversation: "What shall we do this weekend?" And every Friday evening, we would end up scrolling through the same five websites, none of which quite had what we needed. Too much information, not enough curation. Events that had already sold out. Activities that were listed for a different age group. Suggestions that required a car when we don't have one.
The problem with existing children's activity resources isn't that they lack information. It's that they lack perspective. They try to be everything to everyone, which means they're nothing to anyone. What I wanted was something opinionated. Something that understood that a family with a child under five in Zone 2 has different needs than a family with teenagers in Surrey. Something that felt like getting a recommendation from a friend who actually knows your situation.
That's what Little London became. Not comprehensive - curated. Not for every parent - for parents like me. The ugly problem was "what do we do this weekend?" and the answer was a website that treats that question with the same care and taste that a restaurant guide treats "where should we eat tonight?"
Why ugly problems make better products
There are three reasons ugly problems produce better products than pretty ones.
First, you understand the user because you're the user. When I built First Out, I didn't need user research or persona workshops or empathy mapping exercises. I was the persona. I knew exactly what the product needed to do because I needed it to do that thing. Every decision was filtered through the simplest possible question: does this help me get off the tube faster? If not, it doesn't belong.
Second, ugly problems have clear success metrics. Did I find the right carriage? Did we have a good weekend? Did I find a decent pub? These are binary questions. You either solved the problem or you didn't. Pretty problems tend to have fuzzy metrics - engagement, retention, "delight." Ugly problems have concrete ones. And concrete metrics force you to make better products because there's nowhere to hide behind a beautiful interface when the product simply doesn't work.
Third, ugly problems keep you honest. When you're solving a real problem that you personally experience, you can't fool yourself into thinking the product is good when it isn't. If First Out sends me to the wrong end of the platform, I feel it immediately. If the pub guide recommends somewhere terrible, I'll have a bad evening. The feedback loop is immediate and personal. You eat your own cooking, every single day.
The London Pub Guide and the taste layer
The London Pub Guide might be the best example of what happens when you apply taste to an ugly problem. There are hundreds of pub guides online. Most of them are SEO-optimised listicles with stock photos and descriptions that could apply to any pub in any city. "Cosy atmosphere, great selection of ales, friendly staff." Meaningless. You could write that about ninety percent of pubs in London and be technically correct about all of them.
What makes a pub guide actually useful isn't information - it is opinion. It's someone saying: this one is worth going out of your way for, and this one isn't. This one has the best Guinness south of the river. This one has a garden that catches the last hour of evening sun. This one is perfect for a first date and this one is perfect for getting quietly drunk with an old friend. That's taste applied to a mundane problem, and it transforms the product from a database into something you actually want to use.
The ugly problem was "which pub should I go to tonight?" The answer required not just data but judgment. Not just information but perspective. And that's what makes solving ugly problems so satisfying - they demand the same creative thinking as the pretty ones, but they ground that thinking in reality.
Build what you need
If I could give one piece of advice to anyone thinking about building a product, it would be this: build something you need. Not something you think the market needs. Not something that fills a gap in a competitive landscape analysis. Not something that sounds good in a pitch deck. Build the thing you wish existed when you were standing on the wrong end of a tube platform, or scrolling fruitlessly for something to do with your son on Saturday, or trying to find a decent pub in a neighbourhood you don't know well.
Those problems aren't glamorous. They won't get you on stage at a conference. But they'll get you something better: a product that works. A product you use. A product that solves a problem so clearly that people understand it the moment you explain it. And in a world drowning in over-engineered solutions to non-existent problems, that simplicity is worth more than any amount of hype.
Give me the ugly problem every time.
