^ PRODUCT · WHAT'S INSIDE
Upload a screenshot. Get a Lift Report scored across eight conversion dimensions, with the rewrites already drafted and the priority fixes ranked.
¶ 01 · THE 8 DIMENSIONS
Each section of your page is scored 0–100 on these eight dimensions. A weak dimension is a leak.
MEASURES Whether a first-time visitor can tell, in five seconds, what you sell and who it's for.
LOW SCORE Abstract, adjective-heavy headlines like "Reimagine your workflow" that could belong to any SaaS.
PAGELIFT Rewrites the headline in the reader's own words, naming the product category and the outcome.
MEASURES Whether the reader feels they receive more than they pay — in price, scope, or specificity.
LOW SCORE Vague promises ("boost your growth"), no numbers, no boundaries, no reason to act now.
PAGELIFT Reframes the offer with concrete deliverables, guarantees, and a plain reason to say yes today.
MEASURES Whether the page shows evidence a stranger should believe — logos, numbers, quotes, screenshots.
LOW SCORE Generic "trusted by thousands" claims, stock testimonials, no named customers, no measurable results.
PAGELIFT Flags empty proof and drafts specific quote/stat/logo slots the page needs to earn belief.
MEASURES Whether the eye is guided from headline → sub → proof → CTA in one clean path.
LOW SCORE Everything at the same weight — same size, same color — so nothing wins the reader's attention.
PAGELIFT Marks sections where type scale, contrast, or spacing collapse the hierarchy and suggests fixes.
MEASURES Whether the CTA is visible, singular, action-verbed, and matches the intent of that section.
LOW SCORE Ghost buttons, "Learn more" labels, three competing CTAs in the fold, or no CTA above the fold at all.
PAGELIFT Rewrites the CTA in outcome language and points out where a single action would replace a crowd.
MEASURES Every extra field, click, or moment of doubt between interest and conversion.
LOW SCORE Long forms up front, unexplained fields, price hidden three clicks deep, no trust cues near the CTA.
PAGELIFT Identifies friction hotspots and suggests removing fields, deferring signup, or clarifying commitment.
MEASURES Whether the page still works when it's a 375-pixel column — hero legible, CTA reachable, proof visible.
LOW SCORE Desktop hero cut off on mobile, CTA below three screens of scroll, forms wider than the viewport.
PAGELIFT Reads the layout at mobile widths and points out where the hero, proof, or CTA breaks down.
MEASURES Whether the first screen carries enough — headline, sub, proof cue, CTA — without becoming noisy.
LOW SCORE Either an empty hero with a headline and nothing else, or a wall of features stacked on entry.
PAGELIFT Scores what earns space above the fold and drafts a tighter first-screen composition.
¶ 02 · HOW THE ANALYSIS WORKS
We do not run a rules engine or a pixel-level linter. Your screenshot is analyzed by a vision model against a fixed CRO checklist.
PNG or JPG, up to 8 MB. The screenshot is sent to our server over HTTPS; the raw image is never stored on the anonymous demo path.
We call the OpenAI Chat Completions API with model gpt-4o, passing the screenshot as an image content block plus a system prompt that acts as a senior CRO auditor.
We use Structured Outputs (strict JSON schema) so every response has the same shape: lift score, per-section scores across all eight dimensions, priority fixes, rewrites, A/B variants.
The prompt is locale-aware. A French landing page gets French rewrites; an English one gets English. The scoring rubric is the same in both.
¶ 03 · WHAT YOU GET
One page. Six blocks. Everything you need to fix your landing without opening a second tab.
A single 0–100 score with the one-sentence verdict. The number you check first, the sentence you send to your team.
Every section of the page (hero, proof, offer, CTA, features, pricing, FAQ, footer) scored across all eight dimensions.
The changes ranked by expected lift, so you know which edit to ship first — not a wall of forty suggestions.
The original section copy alongside a rewritten version, one click to copy. Ready to paste into your CMS.
Two alternative headlines and CTAs per section, ready to drop into your testing tool.
Client-ready Lift Report as a branded PDF. Included on Growth and Agency plans.
¶ 04 · WHAT PAGELIFT DOESN'T DO
A short list is more useful than a long promise. Here is what Pagelift is not.
Pagelift suggests variants. You still need a testing tool (Optimizely, VWO, PostHog, GrowthBook) and real traffic to prove a change lifted conversion.
Pagelift never touches your CMS, your codebase, or your live page. It gives you the diagnosis and the copy; you ship the change.
Pagelift scores what a stranger sees on the page. It has no access to your GA, PostHog, session recordings, or funnel data — and no way to know your true baseline CVR.
Pagelift catches the boring, repeatable leaks. It won't tell you what your positioning should be, who your ICP is, or whether the offer is right for the market.
Scores are directional. A page that follows every suggestion can still underperform if the traffic, the offer, or the market is wrong. No auditor — human or model — can promise otherwise.
— If a tool tells you it will "double your conversion in 24 hours," close the tab. What actually moves conversion is a tight offer, honest copy, and one edit at a time — measured.
^ TRY IT
Three free audits, no card, in seconds. See a full Lift Report before you decide anything.