The rubric
How we score your product.
This is the Finney Design Rubric. We built it for what AI page builders get wrong, so it catches the tells a human reviewer catches: template hero copy, defaulted layouts, missing trust, broken flows, unfinished mobile, and calls to action that go nowhere.
We judge only the product. Never the idea, never the market, never the business model. Six dimensions, each 0 to 100, averaged into one overall read.
Clarity and Promise
In five seconds, is it obvious what this is, who it is for, and why it is worth their time?
Craft and Design Integrity
Does it look designed on purpose, or defaulted by a generator?
This is where the AI Generator Fingerprint check lives. When a build matches an unmodified Claude, ChatGPT, Lovable, Bolt, v0, or generic-SaaS-template look, we score Craft lower and name the specific tell.
Trust and Legitimacy
Does it look safe and real enough to hand over money or personal data?
Flow and Usability
Can a first-timer actually get through the main task without hitting a wall?
Mobile and Access
Does it work on a phone, and for people who are not using it in ideal conditions?
Conversion and The Ask
Is there one clear next step, low friction to take it, and does that path actually work?
The verdict bands
- 0 to 49Slop
- 50 to 69Rough
- 70 to 84Getting real
- 85 to 100Real product
The overall is the average of the six dimensions, rounded to the nearest integer.
That is the whole method. No opinions on your idea, no benefit of the doubt, no mystery. When you are ready, hand us the URL.