Find what's costing your site its credibility
Visitors decide whether a site looks legitimate before they read a word of it. This lens scores whether yours looks well made or generic, and every finding cites the specific pattern.
What the lens checks
Does this site look credible enough to trust with your money? Every check below scores a piece of that answer, with the evidence quoted.
Type hierarchy
A real hierarchy of sizes and weights, not one font at its defaults. Headings that earn their place instead of shouting uniformly.
Palette discipline
A deliberate palette rather than a stock template look. Fourteen slightly different greys is a finding, not a style.
Template tells
Gradient text, glass cards, stock micro-interactions, and layout clichés. Each one is cited by name when it shows up.
Copy smells
Generic, AI-generated filler. The lens's whole job is spotting templated work, and it quotes the phrase it caught.
How it's scored
- 35 of the 100 points, the same weight as Conversion. Looking credible and persuading are the two biggest levers the rubric has.
- Findings cite the specific pattern: the tell, the cliché, or the quoted phrase. Nothing is scored on taste alone.
- Desktop and mobile are captured separately, so a layout that only breaks on a phone still shows up.
35 / 100
- Conversion35 pts
- Trust & Credibility35 pts
- Page Speed10 pts
- AI & Search Visibility20 pts
What is design taste? Craft, slop, and what a machine can't fake
Taste is the judgment that separates considered design from template and AI slop. What slop looks like, what craft looks like, and how the audit tells them apart.
Frequently asked questions
Can a machine really judge design?
It judges the parts that can be named: template tells, hierarchy, palette discipline, and copy smells. Where the model is involved it has to cite the specific pattern it found, so you can check every finding against your own screen.
What does a finding look like?
The pattern, the severity, the evidence (a quote from your copy or a crop of your page), and the fix. Not 'consider improving visual hierarchy'.
My site is deliberately minimal. Will that score badly?
Minimal isn't the failure mode the lens hunts. It hunts generic: stock gradients, glass cards, filler copy. A restrained site with a real hierarchy and specific copy scores well.
The other lenses
See your own score on this lens
Trust & Credibility is worth 35 of the 100 points. The free audit scores all four lenses and ranks what to fix first.