Building Without Code

Here is the uncomfortable stat. What used to require a team of 5 and 6 months, one person built in weeks. 20+ live products. Zero lines of code written by hand. Zero budget. No dev team, no sprints, no standups.

Not because the tools got easier. Because the bottleneck was never the code.

The bottleneck was always knowing what to build. Knowing what to cut. Knowing who it was for and why it mattered. The bottleneck was taste. And taste doesn't require a computer science degree.

27Products Live0Lines of Code15+Years in Strategy

The bottleneck was never technical

I spent fifteen years as a Strategy Director in advertising. Good agencies, interesting clients, talented people around me. And every single day, ideas died in decks. Not bad ideas. Good ones. Sharp, clear, well-researched ideas that never became anything because the gap between "what if" and "here it is" required a team, a budget, a timeline, and a chain of approvals that ground the original spark into dust.

That gap defined my career. I could see things clearly. I could articulate them. I could convince a room. But I could never make them real on my own. The strategy was always dependent on someone else's execution.

AI tools like Claude Code collapsed that gap to zero. Not by making coding easy. I still cannot write code. I don't know JavaScript from Python in any meaningful way. What these tools did was make the code irrelevant. The code became the commodity. The scarce resource shifted to the thing it was always supposed to be: the thinking.

The gap between "what if" and "here it is" used to be a team, a budget, and three months. Now it is a weekend and a clear idea.

What to build. Who it is for. What to include and, more importantly, what to leave out. How it should feel. What the name should communicate. Where it sits in the world. These are not engineering decisions. They are strategy decisions. Taste decisions. The same decisions I had been training for across fifteen years in advertising.

I just never had the tools to act on them alone. Now I do.

What I actually built

Let me be specific. Because "I built things with AI" means nothing without evidence.

The Relevance Index scores 1,203 brands on cultural relevance. Not vibes. Real data. It pulls from Wikipedia pageviews, Reddit discussion volume, and three independent AI analysis calls per brand. It scores each brand across five dimensions: Attention, Conversation, Creation, Desire, and Zeitgeist. It includes stock market data for 260 publicly traded brands. It updates automatically every Wednesday at 3am. It has 1,201 individual brand pages, 15 category pages, and AI-generated insights powered by Claude. It runs itself.

The Pattern is a daily AI culture briefing. It consumes over 50 RSS feeds from CultureTerminal, my culture news aggregator. Every morning at 7am, an automated pipeline fetches the latest signals, Claude Haiku synthesises them into sections (Culture Pulse score, The Lead, 5 Signals, The Pattern, One to Watch), and ElevenLabs generates audio using a clone of my voice. It publishes to a website and is available on Spotify. A fully autonomous daily media product, built and run by one person.

Taste OS is a framework for scoring brands on taste. Five dimensions, twenty points each, one hundred total. 111 pre-scored brands. An interactive scorer that lets anyone apply the framework to any brand. It turns the abstract idea of "taste" into something measurable and debatable.

🎯The point: These are not demos. They are not prototypes. They run autonomously. They have real data pipelines. They update themselves. They are products, not projects.

And they are three of twenty-seven. The full list includes a culture news aggregator, a wearable tech newsroom, a parenting activities directory for London, a pub guide, a Japanese restaurant guide, a Japanese language learning app, a tube navigation tool used by parents with buggies, a brand tension index, a Nottingham Forest quiz, a Nottingham Forest card game, a book library, a social link aggregator, a bookmarking tool, and more. Each one live. Each one functional. Each one built by someone who, eighteen months ago, had never opened a code editor.

Taste is the only input

When the tool handles execution, all that remains is your judgement.

This is the part that most people miss when they talk about AI and building. They focus on the tool. They focus on the prompts. They focus on the technical workflow. But the tool is not the differentiator. Everyone has access to the same tools. The differentiator is what you point them at.

What to build. What to name it. What to include on the homepage and what to bury three clicks deep. Whether the typography should feel editorial or utilitarian. Whether the colour palette communicates warmth or authority. Whether the product solves a real problem or just a theoretical one. Whether it has a point of view or is just another feature list.

These are taste decisions. And taste is not a nice-to-have. When everyone can build, taste is the only competitive advantage left.

When everyone can make, what you choose to make is the only thing that matters. The input is taste. The output is product. The code is just the translation layer.

This is strategy applied to products, not decks. The same skill that helped me write a positioning for a brand now helps me position an entire product. The same instinct that told me what a campaign should feel like now tells me what a loading state should feel like. The person who knows what matters now has the tools to prove it.

The uncomfortable slide

I have a conference talk in my head called "The Strategist Who Ships." There is one slide in it that makes people uncomfortable. It is a comparison.

On the left: a typical agency project. 8 to 12 people. A strategist, a creative director, a designer, two developers, a project manager, an account manager, a QA tester. Budget of 50 to 100 thousand pounds. Timeline of 3 to 6 months. Multiple rounds of review. A launch date that slips twice.

On the right: one person with AI. Weekends. Zero budget. Live in days.

The output on the right is not better in every dimension. Agency teams produce polished, tested, enterprise-grade work. I am not arguing they don't.

But the output on the right exists. That is the point. The majority of ideas that go through the left side never ship at all. They die in decks, in budget reviews, in stakeholder alignment sessions, in scope negotiations. The right side ships. Every time. Because the only person who needs to say yes is the person with the idea.

⚡The economics have shifted. Most people in advertising have not noticed yet. The question is not whether agencies are dead. The question is whether the current model of 12 people and 6 months is the only way to make good work. It is not. And pretending otherwise is expensive denial.

I am not arguing agencies are dead. I spent my career in them. I respect the craft. But I am arguing that the economics have shifted in a way that most of the industry has not absorbed. The gap between what one person can do and what a team was required for has narrowed dramatically. And it will keep narrowing.

What this means for brands

Brands should hire people who build, not just people who plan.

The future creative leader is not someone who writes a strategy document and passes it to a team. The future creative leader is someone who goes from insight to product. Who can see a gap in the market on Monday and have a working version live by Friday. Who treats prototyping as thinking, not as a phase that comes after the thinking is done.

Taste, judgement, and shipping speed matter more than technical skill. This is already true and it is becoming more true every month. The person who understands audiences, who has genuine cultural fluency, who knows what good looks like and can articulate why, and who can also ship a working product? That person is worth more than a department.

"Show me what you've built" should replace "walk me through your deck." Decks are promises. Products are proof. The best strategist in the world is the one who can show you a live URL, not just a PowerPoint file.

Decks are promises. Products are proof. The best strategist in the world is the one who can show you a live URL, not a PowerPoint file.

The advertising industry trained me to understand people, to find insights, to shape narratives, to make things resonate. Those skills did not become less valuable when AI arrived. They became the only skills that matter. Because the making is now available to everyone. But the knowing what to make? That still requires fifteen years of watching, listening, and developing your taste.

The pitch

I am not a developer. I cannot write a line of code. I am a strategist who ships.

The 20+ products on this site are my portfolio, my proof, and my pitch. They run autonomously. They serve real audiences. They have real data, real design systems, real points of view. Each one was built with taste as the primary input and AI as the execution layer.

If you are a brand that values culture, if you believe taste is a competitive advantage, if you want someone who can go from insight to live product in days instead of months, let's talk.

The bottleneck was never technical. It was always about knowing what to build. I have spent fifteen years developing that instinct. Now I have the tools to prove it.