I used to fall for it. New model, new agent framework, new skill everyone's posting about — build first, ask what it's for later.
I stopped chasing that. Not because the tools stopped being interesting. Because "interesting" isn't a reason to build anything.
What actually changed the pattern wasn't discipline. It was having a real problem.
I needed to crack landing pages — run more tests, produce more variations, improve the content I was putting out. More pages, in less time, without the quality dropping. The kind of improvement that shows up as a lower CPA, not just a fuller content calendar.
That's a real need. So I built something for it.
I built a workflow that actually helps. Its purpose: take any external landing page — generic or a competitor's — strip out its psychological structure, and rebuild it according to my template. The workflow is a chain of six skills, each one running right after the last:
- Brief — who's the audience, what's the product, what's the message
- Structure breakdown — Claude reads the landing page and maps the hook, the psychological structure, and the core messages
- Translation — every insight rewritten into my product's language, values, and competitive advantages
- Image prompts — visuals generated to match the new content
- Reassembly — the content rebuilt according to my template
- Final pass — editing and correction
Nothing gets copied. The page is disposable. What I'm extracting is the mechanism underneath it — why the hook works, what psychological move the structure is making, which messages are doing the real persuading.
It didn't just save time. Pages built this way convert better than the ones I wrote from scratch — and the CPA shows it.
The value was never the workflow. It was knowing what I needed before I built anything.
The prompt itself won't transfer — it's built around my product and my template. But you don't need my prompt. Take any landing page. Run it through the same six steps. Swap in your own product at the translation stage. That's it.
A workflow built around a real need is infrastructure. One built around excitement is a hobby that happens to run on Claude.
What's the last thing you built because you had to — not because it was cool?
Originally published in The Growth System newsletter on LinkedIn.