By the end of 2027, comparable AI products with materially different brand and interface craft will show a measurable difference in conversion or price realisation attributable to that craft — not to model quality, which will have converged.
Once baseline functionality is table stakes, design quality becomes the differentiator that decides category leaders.
The falsifier: if buyers keep choosing on model benchmark or price alone with brand craft showing no measurable effect, or a no-brand product wins the category outright, this call is wrong.
Why we think it resolves hit: app stores went through an identical phase once baseline functionality became table stakes. AI tooling is entering the same phase faster because the functional gap between competitors is closing within quarters, not years.




