FiveStrangers sends five AI strangers with real agendas through your site and runs a full technical, SEO and content audit alongside. You get one honest score and the fixes to raise it. Shipping an app instead? The same treatment works on your App Store or Google Play listing.
Audits open in a few days. Tell us your site and we'll email you the moment it's ready.
Each persona has an agenda: skim, compare, distrust, buy. They scroll, click and narrate exactly where your site loses them.
Lighthouse performance, accessibility, broken links, cross-page SEO, security headers. Plus whether ChatGPT and Perplexity can read your site at all.
A versioned rubric turns verdicts and findings into a 0–100 score. Findings without evidence are thrown away, not shown.
Paste-ready rewrites and prioritized fixes. Re-run within 30 days and see exactly what improved.
Arrives from a search. Gives your page ten seconds to explain itself, then leaves.
Browses on a phone, thumb-first. Anything slow, small, or covered by a banner loses.
Reads everything, hunts for pricing, credentials and proof, and notices what's missing.
Has your competitors open in three other tabs. Looks for one reason to pick you, or to rule you out.
Already sold. Counts every click, form field and unclear label between them and done.
Paste an App Store or Google Play link instead of a website and the audit changes shape. Three store visitors judge your listing the way people actually scroll a store, and your title, subtitle, description, icon and screenshots are measured against what each store rewards.
Sees your icon, title and first two screenshots in a results list. Decides in about two seconds whether to tap at all.
Has three apps in this category open. Looks for the one reason to pick yours, and checks whether it still looks maintained.
Reads the description properly, weighs the rating and the update date, and notices when the listing promises more than the screenshots show.
Title and subtitle against their character budgets, the first description lines people see before "more", and keyword use judged the way each store indexes it.
Starting with the first screenshot, the one that shows up in search results, and whether your captions survive thumbnail size.
Your current title, subtitle and opening lines, replaced with versions that fit the limits.
Every persona's verdict, decisive moment and journey, with the exact screenshots they saw.
Each issue quotes the visitor who hit it, names the page it lives on, and shows the screen beside it.
The old line struck through, the new line ready to paste. Written in your site's own language.
Performance, SEO and AI-search readiness, accessibility, code quality, link integrity. Measured, not guessed.
Audit again after your fixes: the next report shows the score delta and every finding marked fixed, still open, or new.
No. FiveStrangers only needs a public URL. The AI visitors browse your site exactly like a stranger from Google would. No scripts, no logins, no tracking snippets.
Usually around 10 minutes. Five browser sessions, a Lighthouse run and the analysis all happen live, and you can watch the progress on the status page.
Yes. The visitors read your site in its own language, and copy rewrites are written in your site's language so you can paste them directly. The report itself is currently written in English.
Each one browses in a real Chrome browser with its own mindset, screen size and patience budget. They scroll, click your buttons, follow links across pages, and narrate their honest reactions. Their verdicts, converted or bounced, anchor your score.
Yes. Paste your App Store or Google Play link instead of a website. Three store visitors decide whether they would install, and your title, subtitle, description, icon and screenshots are audited against what each store rewards. You get the same scored PDF, with rewrites that fit the character limits.
More people ask ChatGPT and Perplexity for recommendations every month. We check whether AI crawlers are allowed to read your site, whether your content is structured so they can cite it, and whether machine-readable data (schema, llms.txt) is in place.
Yes. Scoring follows a versioned rubric, and your re-audit is compared finding by finding against your previous run: fixed, still open, or new.
Your report and email address are stored so we can deliver it and compare your next run. Payment details are handled entirely by Stripe and never touch our servers. Screenshots used during analysis are temporary.
If the audit can't complete (bot walls, logins, broken pages), a paid report is refunded automatically. The failure itself is usually worth knowing about.