The Consistency Gap

SketchScript had a free tier with almost no paid conversions. That’s a signal worth investigating before building more features.

Three research threads ran in parallel: demand signals, content policy constraints, and conversion patterns — checking whether there was a viable subscription product or just a free tool that would stay free.

Two of the most-requested use cases — headshots and logo creation — turned out to be blocked by content policy. Not edge cases. The first and sixth most demanded categories, both unavailable. That removed two obvious directions and forced a harder look at what was actually buildable.

The conversion research was the useful part.

The assumption going in was that users would pay for better quality or faster output. The research found something different: the feature that converts free users to paid is consistency. Same style, same proportions, same feel across a batch of images. Not quality. Not speed. The ability to trust that image 12 will look like image 1.

This is a solvable problem with current tools. It’s also a different product design than “better image generation.” It’s workflow design — managing style references, seeding parameters, building repeatable presets.

There was also a smaller issue during the session: twice, the system pulled from the wrong aesthetic context and applied private brand guidelines to what was meant to be a generic public demo. The same AI context that knows the internal brand very well will use it by default unless explicitly constrained. Worth watching for. A product aimed at a broad audience needs to stay generic — that constraint has to be stated, not assumed.

The result: a clearer product direction (social graphics, blog headers, sketchnotes for consistency-focused users) and a feature priority list that didn’t exist before. Still early. The research says consistency converts — building for that is the next test.