We ran 700 full CRO audits of real landing pages between May and July 2026: desktop and mobile capture, 40+ conversion checkpoints per page. The pattern is brutal and consistent. Pages have mastered looking professional and neglected everything that turns a visitor into a customer.
700 pages by score band. The median is 54, and not one page reached 80. The best result in the entire sample was 75.
We score every page across ten dimensions. Nine measure persuasion: the headline, the proof, the offer, the path to action. One measures how the page looks. Guess which one the average page has mastered.
Average score out of 10 across 700 audited pages.
Every audit names the top problems on the first screen, the issues most likely to cost the visit. Here is how often each theme shows up among those top problems.
Share of audits where the theme appears among the page's top first-screen problems.
Trust and proof (91%). Reviews, guarantees and credibility signals missing exactly where the decision happens. Mobile experience (75%). Most traffic is mobile; most pages are still built and QA'd on desktop. CTA discipline (62%). Competing buttons, vague labels, or the primary action buried below the fold. Pricing and offer (55%). Visitors leave to find the number the page would not show them. Speed and performance (45%). Slow first paint quietly taxes every other fix on the page. Headline clarity (44%). Feature claims and slogans where a concrete outcome should be.
Of the 13.2 leaks on an average page, 10.5 sit in or above the first screen: 80% of the problem lives before a visitor ever scrolls. The headline, the trust signals and the primary action decide the visit in seconds. A team that only fixes its first screen fixes four fifths of its leak.
We measured mobile performance on 682 pages. The median score is 48 out of 100, and the bottom quarter sits at 36 or below. Meanwhile mobile issues appear among the top first-screen problems in 75% of audits. The gap between where the traffic is and where the attention goes has not closed; it has calcified.
Average score out of 100 by industry. Number after each label is pages audited; segments under 12 pages excluded.
Crypto and Web3 pages score worst by a wide margin: heavy on animation and jargon, light on proof and plain answers. Local businesses lead the field, and still average just 59.
The encouraging number in the dataset: the median page has 14 quick wins, changes a team can ship in days without a redesign. Rewrite the headline around the outcome, move proof next to the decision, show the price, collapse competing buttons into one. The average page is not beyond saving. It is unfinished.
This report aggregates 700 unique completed CRO audits run through audit.ninjapromo.io between May and July 2026, combining our local export with the live production store and deduplicating by report id. Each audit captures the live page on desktop and mobile, screenshots it section by section, and evaluates 40+ conversion checkpoints covering the first screen, trust signals, offer clarity, page structure, mobile experience and technical health. Scores are 0 to 100; dimension scores are 0 to 10.
Leak counts use the same counting engine as our production reports, so the numbers here match what any user sees in their own audit. Mobile performance scores come from Lighthouse. All figures are aggregates; no individual site data is disclosed.
Sample caveat: pages in this dataset belong to businesses that requested an audit, which skews toward small and mid-size DTC, services and Web3 companies actively buying traffic. It is not a random sample of the web, and revenue figures inside audits are modeled from benchmarks, not measured from analytics.
The same audit that produced these numbers is free to run. Score, first-screen diagnosis and the leaks we find on your page, in about 15 minutes.