Skip to main content

Scanner guide

The WhyIQ CRO Scanner, explained

The Scanner reads your page, simulates fifty visitors, and tells you why people do not convert. This guide covers every feature: what it means, and how to use it.

On this page: What the Scanner does

What the Scanner does

The WhyIQ CRO Scanner reads your page the way an AI browser agent does, taking a screenshot, the page code, and the accessibility tree. Then fifty simulated visitors read the page and tell you why they would or would not convert. You get four scores and a ranked list of fixes.

What a finished scan looks like
whyiq.ai / results / demo
Scan complete · 1m 47s
Report for launchday.io
14 July 2026 at 09:24 · Page Audited: Landing Page
WhyIQ Score
62/ 100
Typical range 56–68
AI Citability Index
61/ 100
Search Ranking
62/ 100
Accessibility Score
52/ 100
Conversion Performance Summary
Based on 50 simulated visitors testing your page
Why visitors left
Unclear value proposition44%
No pricing visible30%
CTA uncertainty16%
Lack of trust signals10%
FIX #01CRO-01Conversion & Clarity
Hero headline: "Simplify your workflows" does not say what the product is or what it does

Where: Screen 1 on desktop and mobile

High severityEngagementDesktop and mobile38 of 50 visitors
Concrete fix to ship
Name the product and its result in the headline or the line under it, for example "Project tracking that shows what your team ships each week", using only claims you can support.
  1. WhyIQ Score, with its typical range. A 0 to 100 read on how clearly visitors understood the page. The range is where a re-scan of the same page would land, so you can tell a real change from normal variation.
  2. All four pillar scores. Every scan returns a WhyIQ Score, AI Citability Index, Search Ranking, and Accessibility Score out of 100, not one blended number.
  3. Conversion Performance Summary & Why visitors left. Horizontal severity-colored bars showing the top reasons the 50 simulated visitors hesitated or left without acting.
  4. Top ranked finding & fix. Names the exact element, where it sits on the page, how many of the 50 visitors it affected, and the concrete change to make.

How to use it

  1. Go to the WhyIQ homepage and paste the URL you want to scan.

  2. Submit with your email. Your first page scan is free and the report saves to that address.

  3. Wait about two minutes while the scan runs.

  4. Read your scores and fixes on the report.

How to read it

Start with the WhyIQ Score, then read the fixes top to bottom. The items at the top are where you lose the most visitors, so they are where to start.

Free

Your first page scan is free, with no card. You can unlock the full 50-visitor report on your first domain for free by connecting your platform in 1-Click Connect, or see pricing for more scans and multi-page site audits.

Page, site, and HTML scans

There are three ways to scan, depending on what you want to test.

Page scan
Scans one URL. Best for a single landing page or any page you want to improve.
Site scanAgency
Scans a whole site, page by page, in one run. Best for full audits and agencies.
Preview URLFree
Scan a page before it is live by pasting a staging or preview link that opens without a login.

How to use it

  1. For a page scan, paste a single URL on the homepage and submit.

  2. For a site scan, choose the site option and confirm which pages to include.

  3. To test an unpublished page, paste its staging or preview link. It must open without a login.

  4. Optional: open "Add incoming promise or traffic context" under the URL box to paste the ad copy, search query, or AI prompt bringing visitors to the page. You can also test any incoming promise interactively inside the report afterward.

Agency

Multi-page site scans are part of the Agency plan. See pricing for what each plan includes.

Incoming promise & scent match inside the report (sample data)
Incoming promise & scent matchPartial scent match · 52%
Test another ad or AI prompt ↓

Visitors arriving from search or social links for "project tracking for small teams" do not see project tracking or team pricing confirmed until they scroll past the hero.

  1. Incoming promise & alignment badge. Scores how well your hero copy and H1/H2s confirm the exact ad headline, search query, or AI prompt that brought the visitor to the page.
  2. Confirmed vs. unconfirmed terms. Shows which promised terms are confirmed above the fold and which terms are missing, so you can fix post-click scent gaps immediately.
  3. Interactive promise tester. Paste any new ad variation, email subject line, or AI prompt directly on the report to test scent alignment on the spot without re-scanning.

Watching a scan run

After you submit, a progress screen shows the scan working through its steps: it captures the page, reads it, runs the simulated visitors, and scores the result. A page scan takes about two minutes. A site scan takes longer because it repeats this for every page.

How to read it

You can leave the tab open and come back. When the scan finishes you land on the report, and if you have signed up it is saved to your dashboard. You can see the progress screen here.

Anatomy of your report

Your report reads top to bottom in a clear sequence: the four-pillar score strip at the top, Incoming Promise & Scent Match, the Page Overview (proposition understanding, CTA Pull, bounce likelihood, and why visitors left), Conversion Findings & Fixes, the Simulated Visitor Breakdown, the pillar deep dives (AI Citability, Search Ranking, Accessibility, and AI Browser Agent Readiness), and the Build Your AI Fix Prompt & 1-Click Autopilot workbench.

The Page Overview card inside every report (sample data)
Page Overview · Proposition, CTA Pull & exit reasons50 simulated visitors
Conversion Performance Summary
Based on 50 simulated visitors testing your page
Did visitors understand your proposition?
Fully understood51%
Partially understood28%
Not understood21%
Visitor response
CTA Pull56 / 100
Bounce likelihoodModerate
Why visitors left
Unclear value proposition44%
No pricing visible30%
CTA uncertainty16%
Lack of trust signals10%
  1. Did visitors understand your proposition? Breaks the 50 simulated visitors into Fully understood, Partially understood, and Not understood based on whether they could state what you sell and who it is for within five seconds.
  2. CTA Pull ({value} / 100). How strongly your primary call-to-action pulls simulated visitors toward acting, always scored as an integer out of 100.
  3. Bounce likelihood. The overall band (Low, Moderate, High, or Severe) summarizing how many of the 50 simulated visitors left without acting.
  4. Why visitors left. Horizontal severity-colored bars ranking the exact friction reasons that caused visitors to abandon the page.

How to use it

  1. Read your four pillar scores and the Page Overview at the top.

  2. Check your Proposition understanding and CTA Pull (/ 100) to see whether your hero or your offer is the main bottleneck.

  3. Scroll to Conversion Findings & Fixes and start at the #1 ranked fix.

  4. Apply the change directly, copy the AI fix prompt, or ship it via 1-Click Connect & Autopilot.

  5. Re-scan to confirm the score and CTA Pull moved.

How to read it

Each fix names the element, the problem, and the change. You do not need to do all of them. The top three usually move the score the most. You can open a full example report to see the whole thing.

Paid

The full report, including every fix and the 50-visitor breakdown, unlocks via the one-time $9 First Scan promo, a paid plan, or for free on your first domain when you verify ownership in 1-Click Connect.

The four scores

Every scan returns four scores, each out of 100. Here is what each one means.

WhyIQ Score
How clearly real visitors understand your page and know what to do next. Higher means fewer people leave confused.
AI Citability Index
How ready your page is to be cited by AI search engines. It predicts readiness from on-page signals. It does not count real citations.
Search Ranking
Your page's on-page SEO basics: titles, headings, metadata, and structured data.
Accessibility Score
How usable your page is for people with disabilities, checked against common WCAG, ADA, and GDPR signals.

The AI Citability Index is built from eight signals. Each one is a thing AI engines look for when deciding whether to quote a page:

  • FAQ Quality: are common questions answered in a clean question-and-answer form.
  • Answer Clarity: does the page answer the actual question directly and early.
  • Statistical Density: are claims backed by specific numbers.
  • Heading Structure: is the heading hierarchy clean and logical.
  • Content Freshness: does the page show recent dates and updates.
  • Author Attribution: is there a named, credible author.
  • AI Crawler Access: can AI crawlers reach the page.
  • Schema Coverage: is there structured data an engine can read.

Every scan also runs a rendering check and an AI Browser Agent Readiness audit (five checks across server-rendered DOM legibility, accessible action controls in the Accessibility Tree, labelled form inputs, semantic landmarks, and plain-text offer extractability). It compares the HTML your server sends, which is what AI crawlers read because they do not run JavaScript, with the page after it renders in a browser. If the text only appears once JavaScript runs, the AI Citability section opens with an “AI crawlers see an almost empty page” finding, with the fix worded for how your site is built.

AI Browser Agent Readiness: five-check DOM & Accessibility Tree audit
Autonomous AI Browser Agent Telemetry

Can autonomous AI agents (ChatGPT Agent, Claude Computer Use, Perplexity) navigate and act on this page?

1 of 5 checks passed
Agent ReadinessHigh agent friction
48/ 100 · 1/5 passing

Autonomous browser agents face structural blockers in the DOM or Accessibility Tree when trying to read your offer or complete your primary action.

Tested against DOM, Accessibility Tree, and form controls

Server-rendered DOM legibility
Readable

Initial server HTML contains readable headings and body copy, so HTTP-only and headless browser agents can read the page without waiting for client hydration.

Accessible action controls (AX tree)
Unlabelled controls

Link has no accessible name

Fix for AI agents: Give every button and link visible text or an explicit aria-label that names the exact action (for example, 'Start 14-day free trial' instead of an icon-only button).

Form & input machine-readability
Unlabelled inputs

No associated label

Fix for AI agents: Pair every <input>, <select>, and <textarea> with an explicit <label for='...'> element and add standard HTML autocomplete attributes (email, name, organization).

Programmatic focus & semantic landmarks
Minor hierarchy gap

Heading level gap before h4

Fix for AI agents: Fix skipped heading levels and ensure all primary page regions use semantic HTML5 landmarks (<header>, <main>, <nav>, <footer>).

Plain-text offer & pricing extractability
Buried offer / pricing

Pricing is buried or absent above the fold (30% of simulated exits cited missing pricing), forcing comparison agents to hunt or leave.

Fix for AI agents: Surface starting price or free-tier terms near the primary CTA and state a direct definitional sentence in the first paragraph.

  1. Agent readiness score & verdict. A 0 to 100 composite across five deterministic checks measuring whether autonomous AI browser agents (ChatGPT Agent, Claude Computer Use, Perplexity Assistant) can read and act on your page.
  2. Five deterministic checks. Evaluates server-rendered DOM legibility, accessible action controls in the Accessibility Tree, programmatic form labels and autocomplete tokens, semantic landmarks, and plain-text offer extractability.
  3. Paste-ready fixes. Every non-passing check names the exact HTML/ARIA change required and feeds into Fix prompts for Cursor, Claude Code, or Lovable.

Good to know

The AI Citability Index predicts readiness, it does not count live citations. To measure whether ChatGPT, Perplexity, Claude, Gemini, or Google AI actually cite you, use WhyIQ AI Radar.

Reading a recommendation & Autopilot

Every fix is written the same way, so you can act on it without guessing, copy it into your AI builder, or stage it directly through 1-Click Connect & Autopilot.

A structured recommendation card paired with the AI Fix Prompt & Autopilot bar
Ranked finding & fix cardDesktop + mobile verified
FIX #01CRO-01Conversion & Clarity
Hero headline: "Simplify your workflows" does not say what the product is or what it does

Where: Screen 1 on desktop and mobile

High severityEngagementDesktop and mobile38 of 50 visitors
Why it stalls visitors
A first-time visitor cannot tell from the first screen which workflows it simplifies or what kind of product it is. The line under the headline and the navigation labels do not name a category or a result either.
Concrete fix to ship
Name the product and its result in the headline or the line under it, for example "Project tracking that shows what your team ships each week", using only claims you can support.
Build your AI fix prompt · 1-Click Connect & Autopilot
Click Automate These Fixes, pick your tool, and follow the 1-2-3 route
Honesty Gate active
Automate These Fixes →Copy Combined Prompt (2 selected)Route: Pick Tool → 1-2-3
  1. Numbered element & Where location. Names the exact headline, CTA button, or section and where it sits on desktop and mobile.
  2. Severity & visitor evidence chips. Shows impact severity, device scope, and how many of the 50 simulated visitors flagged the issue.
  3. Why & Fix boxes. The amber Why box explains the visitor friction; the green Fix box gives the exact copy or layout change to ship.
  4. Build your AI fix prompt & 1-Click Connect. Copy a tailored prompt for Lovable, Replit, Cursor, or Claude Code, or connect your platform in 1 click to stage or auto-deploy fixes.
Element & Where
The exact part of the page the fix is about (a headline, a button, a form field) and its screen position.
Why (Issue)
What is wrong, why visitors hesitated, and how many of the fifty simulated visitors it affected.
Fix
The specific copy, layout, or technical change to make.
Impact
How much the fix should help, so you know what to do first.
Fix prompt workbench
The same fix written as a paste-ready prompt for the tool you edit the site with: Lovable, Bolt, v0, Replit, Cursor, Claude Code, Webflow, WordPress, or Shopify. Tick multiple fixes to bundle them into one prompt.
1-Click Connect & 4-Mode Autopilot
Connect your GitHub repo, CMS, or IDE via MCP to stage drafts or auto-deploy deterministic fixes with Honesty Gate protection and 1-click SHA-256 rollback.

How to use it

  1. Read the fixes from top to bottom. They are ordered by impact.

  2. Do the high-impact fixes first.

  3. Make the change named in the Fix box, copy a bundled AI Fix Prompt, or open 1-Click Connect & Autopilot to stage or auto-deploy the fix.

  4. Re-scan when you are done to confirm it worked.

How to read it

Impact is a priority signal, not a guarantee of an exact point gain. Work from the top of the list and re-scan to see the real effect. See the 1-Click Connect, Autopilot & MCP guide for how to configure the 4-Mode Switchboard.

Visitor psychology & CTA Pull

Beneath the fixes, the report shows the reactions behind them: how strongly your primary call-to-action pulled visitors, what confused them, which buyer archetypes struggled most, and where attention dropped down the page.

Simulated Visitor Breakdown across the four buyer archetypes (sample data)
Simulated Visitor Breakdown · 4 buyer archetypes
Who left, what they believed or questioned, and what never got seen
Analytical Analyst
12 visitors62% bounced
“Simplify which workflows? I read the headline three times and still could not tell you what this replaces.”
✓ Believed: SOC 2 Type II security badge✕ Questioned: Customer logos
Startup Founder
7 visitors35% bounced
“I get what it does and I would try it. I just want to know if there is a setup fee before I put my team on it.”
✓ Believed: Product screenshot✕ Questioned: "Get Started" button without onboarding scope
Discount Dana
8 visitors71% bounced
“No price above the fold means expensive. Next.”
✓ Believed: "Free plan available" badge✕ Questioned: "Get Started" button without price or trial terms
Stranded below the fold: Customer testimonials (Screen 4, 71% never reached) · Pricing / plans (Screen 5, 86% never reached)
6 visitor types appeared on this page. 58% left without converting.
  1. Visitor archetype, headcount & bounce rate. Groups the 50 simulated visitors by how they buy (for example, Skeptical Evaluator or Busy Decision-Maker) and shows what percentage bounced.
  2. In their own words. A verbatim first-person reaction from that archetype explaining what made them stay or leave.
  3. Believed vs. Questioned chips. Highlights which on-page claims the archetype trusted (✓ Believed) and which claims triggered skepticism (✕ Questioned).
  4. Stranded below the fold. Flags high-value proof or pricing sections buried on lower screens where most bouncing visitors never scrolled.
CTA Pull ({value} / 100)
A 0 to 100 score measuring how strongly your primary call-to-action pulls simulated visitors toward converting, based on offer clarity, friction, and commitment risk. Always formatted as an integer out of 100 (for example, 48 / 100).
Confusion point
Something a simulated visitor misunderstood or hesitated over. Ranked by how many of the fifty flagged it.
Unaddressed objections
Specific buyer questions (such as pricing transparency, implementation effort, or cancellation terms) that visitors looked for on the page and could not find, paired with the exact copy needed to answer them.
Visitor archetype & Win plays
One of four buyer behavioral profiles tested on your page. Each archetype card shows its bounce rate, verbatim quotes, Believed vs. Questioned claims, and tailored Win plays to convert that buyer type.
Visitor-Driven Conversion Plays
High-leverage conversion plays synthesized directly from repeated visitor objections and quotes across archetypes.
Bounce rate
The share of the fifty simulated visitors who left without acting.
Scroll-depth funnel & Stranded below the fold
How far down the page visitors scrolled before leaving, and which proof or offer sections were stranded below where visitors dropped off.

How to read it

Use the confusion points, unaddressed objections, and archetype cards to understand why each fix matters. If many visitors of one archetype left, the Win plays for that archetype matter most for your audience.

Site reports and your dashboard

A site scan returns a site audit: a score for the whole site, a ranked action plan, findings page by page, patterns that repeat across pages, and a plain-language summary written for stakeholders. Your dashboard is where every scan lives, and it is built around the improvement loop: fix, re-scan, and see what cleared.

Your dashboard Overview, where every scanned page lives
  1. One card per page. Every page you scan becomes a card. The page that most needs work is sorted first.
  2. WhyIQ Score. The card leads with the page's current score and its band (Needs attention, Developing, Solid, Strong).
  3. Since your last scan. A movement chip shows whether the score went up or down after your latest change.
  4. Open fixes. How many recommendations are still unresolved, so you always know what is left to do.
Inside a Page Workspace (/dashboard/p/[pageKey]): tracking what moved across scans
Page workspace · /dashboard/p/launchday.io
Where you stand · 3 scans on this page
WhyIQ Score 54 → 62 (+8)Rescan page
What moved since your previous scan
✓ Fixed (2)
Hero H1 names category & audience in 5 seconds
Still open (3)
Primary CTA button (“Book a Strategy Call”)
Changed (1)
Customer logo strip moved from Screen 4 to Screen 2
What's in the way · Checklist & Autopilot
Check off fixes as you ship them, or stage them in 1-Click Autopilot before your next rescan.
Open full report →
  1. Where you stand & score delta. Shows the page's scan count, current WhyIQ Score movement (for example, 54 → 62), and a one-click Rescan page button.
  2. What moved: Fixed, Still open & Changed. Compares your latest scan against your previous scan and buckets every earlier finding into Fixed, Still open, or Changed.
  3. What's in the way checklist & Autopilot. Lets you mark fixes done as you ship them, stage them in 1-Click Autopilot, or open the full report.
Sidebar & quick search (⌘K)
The left sidebar gives you a one-click New scan button, instant page search (⌘K / Ctrl+K), Overview (Clients on Agency), Scans, Billing, Settings, Docs, and the WhyIQ product switcher to jump between CRO Scanner and AI Radar.
Overview
Your pages as cards, ranked so the page that most needs work comes first. On Agency this is the Clients view, grouped by client.
Page workspace
Open a card to see that page's loop: where it stands, what moved since the last scan (every earlier finding marked Fixed, Still open, or Changed), the fixes still open, and your next step.
Scans
Every page and site scan you have run, filterable by type, ready to reopen.
Score trend
How a page's WhyIQ Score has moved across re-scans. A marker shows any date when WhyIQ updated its scoring, so an engine change is never mistaken for your change.
ClientsAgency
Separate workspaces for each client, with their own scans and branding.
Client auditAgency
A white-label document per client, under your agency's name: it leads with what you fixed, then walks every page through what moved and what comes next. Share it as a link or a PDF.

How to use it

  1. Sign in to open your dashboard. The sidebar has Overview (Clients on Agency), Scans, Billing, and Settings.

  2. Start with the card at the top of Overview, the page that most needs work.

  3. Open its Page Workspace to see what moved and which fix to make next.

  4. Re-scan the page after the change to update its trend.

Agency

Site scans, client workspaces, the client audit, and white-label branding are part of the Agency plan. See a sample client audit, or see pricing.

Sharing and exports

You can hand a report to anyone, in whatever format they need.

PDF ReportFree
The full report as a document, ready to hand over.
PDF Executive BriefAgency
A two to three page summary for client handoff.
JSONPaid
The structured scan data, for your own tools or automations.
CSVPaid
The scores, fixes, and visitor data in a spreadsheet.
MarkdownPaid
A human-readable copy of the report as plain text.
Fix promptsAgency
Every fix in the report as prompts for your AI builder or developer.
LLM TextAgency
The report laid out for pasting into an AI assistant's context.
YAMLAgency
The full structured data, for automation pipelines.
Share link
A link anyone can open to view the report, no account required.
White-labelAgency
Reports and PDFs that carry your own logo and branding.

How to use it

  1. Open any report.

  2. Use Export to download a PDF. On a paid scan, JSON, CSV, and Markdown are there too.

  3. Use Share to get a link anyone can open.

  4. On Agency, set your branding once so every report carries your logo.

Re-scan and see what cleared

The point of a scan is the next scan. This is the improvement loop: fix, re-scan, and see what cleared. Re-scan the same URL and every finding from the last report comes back as Fixed, Still open, or Changed, while your dashboard charts the score over time.

A page's score trend, as the dashboard shows it (sample data)
Your progress

Score trend

Showing 2 of 4 scans

Click any point to see what changed

Current score

72

+8 overall

50607080909 Sept14 Sept

Each point is a scan. A diamond marks a date when WhyIQ updated its scoring.

How to use it

  1. Apply the top fixes from your report.

  2. Re-scan the same URL.

  3. Read what moved: which findings are Fixed, which are Still open, and what Changed.

  4. Optional: connect Google Analytics 4 (by adding your unique Viewer service-account email and pasting your numeric Property ID in Dashboard -> Settings) or Plausible to unlock "Did the fix work?" and compare 28-day before-and-after CTA and conversion rates alongside your score trend.

  5. Repeat on your next priority page.

Did the fix work? Before-and-after visitor proof from connected analytics (sample data)

Sample data for this fictional page. Real visits read from your Google Analytics, before and after the element a fix named changed between your scans. These are your actual visitors, not simulated ones.

Hero button

28d before vs 28d after

Primary CTA changed: "Sign up" to "Get Started"

Before: 19 May 2026 to 16 Jun 2026 · After: 14 Jul 2026 to 11 Aug 2026 (days between scans excluded)

  • Clicks on the main button(hero_cta_click)

    Could be noise
    Before change
    2.1%of 1,912 visits
    After change
    2.6%of 1,874 visits

    Clicks on the main button rose from 2.1% to 2.6% of visits. With this traffic the change could be noise.

    The data fits a change between -0.45 and +1.5 points.

Signup form

28d before vs 28d after

Form field count changed: "5 fields" to "3 fields"

Before: 19 May 2026 to 16 Jun 2026 · After: 14 Jul 2026 to 11 Aug 2026 (days between scans excluded)

  • Conversions(sign_up)

    Bigger than the noise
    Before change
    3.0%of 1,912 visits
    After change
    4.9%of 1,874 visits

    Conversions rose from 3.0% to 4.9% of visits. That is bigger than the noise for this much traffic, but a before-and-after reading cannot rule out other causes such as a campaign or the season.

    The data fits a change between +0.69 and +3.2 points.

A before-and-after reading is not a controlled test. Seasons, campaigns or other changes can move these numbers too.

  1. Detected element change. Names the exact element a fix targeted that changed between your earlier scan and your re-scan (such as the primary CTA label or form field count).
  2. 28-day before & after window. Compares real visitor sessions and conversion events from your connected Google Analytics 4 or Plausible account before and after the change.
  3. Noise-aware reading. Flags whether the observed shift is Bigger than the noise or Could be noise, with the 95% confidence interval and the reminder that before-and-after is not a controlled split test.

Want to reopen your first-run setup checklist on the dashboard or replay the interactive report walkthrough? Open Dashboard → Settings and use the Onboarding & product guides card at any time.

Key takeaway

Treat each scan as a checklist. Fix the top items (manually, with AI fix prompts, or via 1-Click Autopilot & MCP), re-scan, and the score tells you if it worked.