What this is
A benchmark
A rolling aggregate of completed website scans. It is built from the same six-basic scoring model used in every individual report.
Readiness benchmark // live
A live benchmark for whether business websites are ready.
The Index turns completed Clunky AI scans into a rolling view of how business websites perform across speed, findability, accessibility, trust, action and credible copy.
Current benchmark // average
73/ 100
Good
The latest 90-day benchmark is 73/100 across 283 scans. The median site scores 73, while the top quartile starts around 81.
Window
28 Jul 2026 to 26 Oct 2026
Sample
283 scans
Confidence
Benchmark-ready
What it measures
Readiness
Higher scores mean fewer blockers across speed, findability, accessibility, trust, action and credible copy.
What it is not
Not a scrape
It is not a list of every site on the web, and it does not publish private scan reports. Public numbers are aggregated.
How to read the score
Higher is better. The score is a readiness score, not a shame score. It asks whether the site is ready to do its job across the six basics.
85-100SharpRare. Strong enough to use as a benchmark.
70-84GoodWorking, with a few visible places to tighten.
55-69ClunkyThe normal danger zone for SMB websites.
40-54Very clunkyImportant basics are costing trust or action.
0-39BrokenFix before treating the site as reliable.
How to use it
Read it as context, not a trophy cabinet.
If your site scores below the Index average, the question is not whether the site is embarrassing. The question is which basic is costing attention, trust or action first.
If your site scores above the average but below the top quartile, the opportunity is usually a focused cleanup rather than a full rebuild.
If your site reaches the top quartile, keep monitoring it. New AI-built websites, new browser behaviour and new content patterns move the benchmark over time.
Distribution
The average is useful. The spread is more useful.
Quartiles show where a scanned website sits against the current sample. A score of 72 means more if the median is 58 than if the median is 74.
P25
66
A quarter of sites score at or below this.
Median
73
The middle of the sample. Half score lower, half higher.
P75
81
Top quartile starts here. This is a useful improvement target.
P90
87
The top ten percent of the current sample.
0Website Readiness Score100
Six-basic breakdown
Which basics are dragging sites down?
These are category averages across the current benchmark. They help readers see whether the market is mainly failing on speed, findability, accessibility, trust, action or credible copy.
Does it load fast enough?
Performance & Technical · 20% weight
Mobile speed, Core Web Vitals, HTTPS, SSL and HTTP/2.
70
Average
Will anyone find it?
SEO & Discoverability · 20% weight
Search metadata, headings, structured data, robots, sitemap and crawlable content.
61
Average
Can everyone actually use it?
Accessibility · 15% weight
axe-core violations in the rendered page.
83
Average
Does it feel trustworthy?
Trust + visual polish · 15% weight
Trust markers, policy basics, structured organization data and visual polish.
61
Average
Does it help people take action?
Conversion + layout · 15% weight
Above-fold calls to action, form basics, layout clarity and funnel focus.
76
Average
Does the copy feel credible?
Copy Credibility · 15% weight
Specificity, sentence rhythm, repeated phrasing and generic AI-shaped wording.
89
Average
Builder signals
AI builders are shown as signals, not accusations.
The scanner detects builder fingerprints where it can. We publish builder-level rankings only once there are 30+ recent samples, so early rows should be read as directional diagnostics.
Scores describe websites in the detected sample, not every site made with that builder and not the builder company itself.
Lovable
Framer AI
Bolt