After finishing a resume, one developer faced the familiar wall of "what should I actually build?" What followed was a practical framework for validating a side project idea — one built to avoid the trap of endlessly preparing to build without ever shipping anything.
The author settled on four filters. First, can it run on free hosting? Spending money before there's any traffic or revenue doesn't make business sense at the earliest stage. Second, keep implementation complexity manageable — high difficulty doesn't make a service good, it just makes maintenance harder. Third, favor low operational overhead and automation — a service doesn't run itself just because it launched; low-effort operations were treated as a non-negotiable design constraint. Fourth, avoid cloning an existing service outright — cloning is fine for practice, but learning product thinking and operations requires something with a genuine point of difference.
The author turned these four criteria into a set of instructions and used them repeatedly with Claude to brainstorm and stress-test ideas — even asking what could be built with public APIs. Good ideas didn't come easily. The breakthrough came from flipping the question: instead of chasing an impressive concept, start from "what personally annoys me?" and then re-apply the same four filters. Notably, the idea that eventually stuck was the very first one that had come to mind — time spent chasing flashier alternatives was, in hindsight, a detour.
These four criteria map directly onto how to plan a marketing campaign or content experiment. Validating with free or low-cost channels before committing budget, and designing for low operational overhead so an experiment can repeat, are exactly the principles behind disciplined growth experimentation.
Using AI as a brainstorming partner is also worth noting: defining clear criteria (cost, complexity, differentiation) up front, then iterating with AI to narrow down campaign concepts or copy ideas, is a repeatable pattern for any team exploring new ideas. To get help structuring experiments and channel strategy, explore Best Partner's services or get in touch.
Whether it can run on free hosting, whether implementation complexity stays manageable, whether operations can be automated, and whether it's genuinely different from existing services. Defining these up front speeds up decision-making.
Instead of chasing an impressive external idea, start from a problem that personally annoys you, then test it against pre-defined criteria like cost, complexity, automation potential, and differentiation.
Turn your validation criteria into instructions and use them repeatedly with an AI assistant to brainstorm and narrow down ideas — a technique that works for new services as well as marketing campaign concepts.
To apply what you just read to your own site, start with a free audit of where things are now.
A strategist replies within 24 hours on business days.
A studio replaced designer, frontend and backend postings with a single product builder role. Dozens of interviews later, nobody was hired. What they changed.
From hotels to F&B to beauty to head of marketing at Sulbing: how volunteering for undefined work, learning the operational basics, and owning a signature widen a marketing career.
Questions are not random. Across different companies and products, the underlying intent is remarkably consistent. Fill in three boxes instead of memorising a list.