Prompt to
Product
A designer's field test — building the same news app across 4 AI builders to find out who actually delivers on the brief
ROLE
UX Researcher
BUILDERS
Lovable, v0, Bolt, Emergent
YEAR
2026
01 - OVERVIEW
A designer's field test
This project started as a question with no clean answer — with so many AI app builders launching at once, which one is actually worth using? Rather than read about them, I built the same app in four of them using the same single prompt and evaluated each on output quality, design decision, iteration responsiveness, and free tier usability.
"What started as a tool comparison became something more useful: a lesson in why the design process still matters even when AI is doing the building."
02 - DISCOVERY
Jumping in blind
This project began without a plan — no design process, no PRD, no defined feature list.This project began without a plan — no design process, no PRD, no defined feature list. Just an idea and a builder. Replit was the first tool used, not as part of the evaluation, but as a starting point to see what would come out of building with no direction.
What emerged was rough but immediately useful — each iteration revealed what was missing, what felt off, and what the app actually needed to work. Features weren't decided upfront, they were discovered through building.
Back to basics
After the discovery phase with Replit, it was tiem for formalize what had been learned. The features that surfaced through iteration, the interactions that felt missing, and the design decisions that needed revisiting all pointed to the same thing — a clearer picture of what the app needed to be. That clarity became the PRD. Rather than a spec written in advance, it was built from experience — a documented standard every builder in this evaluation is measured against. Every builder starts from the same place and gets the same opportunity to get there.
Navigation gaps, missing states, and interactions the brief never specified — all of which became requirements.
The discovery phase produced the PRD. Every builder in this evaluation starts form the same place and gets the same opportunity.
03 - THE BUILDERS
Candidates for the job
Before any builder was tested, the app was built first in Replit with nothing more than a problem statement. No wireframes, no feature list, no design direction. What came out of that first build wasn't polished — but it was enough to react to. Seeing the app exist in any form made it immediately clear what was messing, what felt wrong, and what interactions the experience actually needed. Every friction point, every missing feature, every UI decision that needed revisiting became a requirement. Replit wasn't part of the evaluation — it was the discovery phase that made the evaluation possible, and the standard every builder is measured against.
04 - ITERATION WALKTHROUGHS
See what each builder made
05 - BUG SPOTLIGHT
When the builder doesn't fix it
Not every issue gets resolved across iterations. This is one that persisted — the comments background bug that revealed how builders handle state management differently.
06 - SCORECARD AND THOUGHTFUL ADDITIONS
How they stacked up
Not every issue gets resolved across iterations. This is one that persisted — the comments background bug that revealed how builders handle state management differently.
07 - TAKEAWAYS
Lessons to take onto the next project
This project taught me more about the importance of the design process than it did about AI tools.
01
Going into a builder without a design direction produced something — but reacting to what came out wasn't the same as designing with intention.
02
All AI builders are excellent at executing a clear brief and much weaker at filling in what wasn't specified.
03
The builders that produced the strongest outputs weren't necessarily the most technically capable — they made the best design decisions in the absence of explicit instructions.
04
That's a design judgment call, not an engineering one — and it's where the gap between a good builder and a great one actually lives.
Piqued your interest?