Check points
Pass earns 1, partial earns 0.5 and fail earns 0. Confidence does not change health points. Missing data never becomes an invented zero.
Methodology
How the customer experience audit is scored
The website check measures how easily customers can find and contact a business using the public pages and sources we could read. Missing evidence stays Not visible. These are the checks and scoring rules used for your result.
scoring/v2instant-cx/v10
Scores and coverage
What the arithmetic means
Area scores
An area score is weighted earned points divided by observed raw check weight. An area needs 50% coverage to show a grade; otherwise it says “Not visible.”
Overall score
The observed-evidence score averages known graded checks within each area, then combines those averages using the published area weights. Known checks count even when their area has too little coverage to show a grade. Unknown, not-applicable and context-only checks earn no points and are excluded from that average. An overall score needs 60% weighted coverage and 4 eligible areas. A website audit can receive an overall score without a matched Google listing. A business-name search still requires a resolved identity. Otherwise the result is a “Limited snapshot.”
Grades
Unrounded scores set the band: A at 90; B at 75; C at 60; D at 45; F at 0. Display rounding never upgrades a grade.
Recommendation strength
Bands: A at 24; B at 16; C at 10; D at 0. Confidence points: High: 2; Medium: 1; Low: 0. Scope multipliers: Page: 1; Site: 1.2; Location: 1.5; Brand: 2.
Identity and website aliases
A listing can match the audited domain, the site’s declared canonical address or og:url, a listing website when the redirect check follows a redirect to the audited domain, or the listing selected in a business-name search. Shared website hosts do not establish identity. A phone-number match supports identity but never establishes it by itself.
Evidence limits
What “Not measured on this snapshot” means
An honest unknown
When an eligible public page was left unread, a fact that could have appeared there is not measured on this snapshot. It is not a pass or a fail: it is excluded from score arithmetic and remains part of coverage.
Evidence-specific, not page-count driven
The decision follows what the crawl left unread, not a minimum page count. A small site that was read completely can still fail a check, while a larger site can leave a fact not measured.
Website checks
8 customer experience areas
CX area
Finding the phone number
Publish a phone number
Whether a phone number is published.
SourcesPublic website crawl (instant/v1)
Put the phone number within one click
Clicks from the home page to the phone number.
SourcesPublic website crawl (instant/v1)
Keep the public site reachable
Whether the public site could be crawled.
SourcesPublic website crawl (instant/v1)
Publish business hours
Whether business hours are published.
SourcesPublic website crawl (instant/v1)
CX area
Mobile contact
Make the phone number tap-to-call
Whether the phone number is tap-to-call on mobile.
SourcesPublic website crawl (instant/v1)
Keep mobile field CLS at 0.1 or less
CrUX field Cumulative Layout Shift when coverage exists.
SourcesChrome UX Report field data returned by PageSpeed Insights v5
Keep mobile layout shift at 0.1 or less
PageSpeed Insights mobile lab Cumulative Layout Shift.
SourcesGoogle PageSpeed Insights v5 mobile lab data
Chrome UX Report field data returned by PageSpeed Insights v5
Keep mobile field INP at 200 ms or less
CrUX field Interaction to Next Paint when coverage exists.
SourcesChrome UX Report field data returned by PageSpeed Insights v5
CX area
Form friction
Offer a public contact form
Whether a public contact form is present; nothing is submitted.
SourcesPublic website crawl (instant/v1)
Put the form within one click
Clicks from the home page to the form.
SourcesPublic website crawl (instant/v1)
Ask for five or fewer required fields
Count of required fields without submitting the form.
SourcesPublic website crawl (instant/v1)
Avoid a blocking CAPTCHA
Whether the form presents a CAPTCHA before submission.
SourcesPublic website crawl (instant/v1)
Label every form field
Whether form fields have accessible labels.
SourcesRendered contact-path pages (axe-core, WCAG A/AA)
CX area
Channels offered
Publish an email option
Whether a public email option is present.
SourcesPublic website crawl (instant/v1)
Offer a text option
Whether a public text option is present.
SourcesPublic website crawl (instant/v1)
Offer a chat option
Whether a chat entry point is present; it is not opened.
SourcesPublic website crawl (instant/v1)
Offer an online booking path
Whether an inspected booking path offers a verified scheduler or time-slot controls.
SourcesPublic website crawl (instant/v1)
Clicks from the home page to the page with the booking link
Clicks from the home page to the page with the booking link
SourcesPublic website crawl (instant/v1)
CX area
Listing consistency
Resolve the business identity
Whether the audited site and listing resolve to one business.
SourcesGoogle Maps listing match facts (DataForSEO; no listing text retained)
Maintain a discoverable Google listing
Whether a matching listing was found.
SourcesGoogle Maps listing match facts (DataForSEO; no listing text retained)
Match the listing phone number to the site
Whether the listing and site phone numbers match.
SourcesGoogle Maps listing match facts (DataForSEO; no listing text retained)
Match the listing website to the site
Whether the listing points to the audited website.
SourcesGoogle Maps listing match facts (DataForSEO; no listing text retained)
Publish hours on the listing
Whether the matching listing includes hours.
SourcesGoogle Maps listing match facts (DataForSEO; no listing text retained)
CX area
Contact-path accessibility
Resolve automated accessibility issues
Count of distinct detected A/AA defects on contact controls.
SourcesRendered contact-path pages (axe-core, WCAG A/AA)
Meet color-contrast checks
Whether contact controls pass automated contrast checks.
SourcesRendered contact-path pages (axe-core, WCAG A/AA)
Give links accessible names
Whether contact links have accessible names.
SourcesRendered contact-path pages (axe-core, WCAG A/AA)
Use accessible tap targets
Whether contact tap targets pass the scan.
SourcesRendered contact-path capture (24x24 CSS px, WCAG 2.2 2.5.8)
CX area
Site speed
Raise mobile lab performance to 90
PageSpeed Insights mobile lab performance score.
SourcesGoogle PageSpeed Insights v5 mobile lab data
Chrome UX Report field data returned by PageSpeed Insights v5
Keep lab LCP at 2.5 seconds or less
PageSpeed Insights mobile lab Largest Contentful Paint.
SourcesGoogle PageSpeed Insights v5 mobile lab data
Chrome UX Report field data returned by PageSpeed Insights v5
Keep lab blocking time at 200 ms or less
PageSpeed Insights mobile lab Total Blocking Time.
SourcesGoogle PageSpeed Insights v5 mobile lab data
Chrome UX Report field data returned by PageSpeed Insights v5
Keep field LCP at 2.5 seconds or less
CrUX field Largest Contentful Paint when coverage exists.
SourcesChrome UX Report field data returned by PageSpeed Insights v5
CX area
Phone experience
Answer the published phone number
Whether the optional phone-test call was answered.
SourcesOptional phone-test call instrumentation (instant-phone/v1)
Show rings before answer
Observed rings before the call was answered.
SourcesOptional phone-test call instrumentation (instant-phone/v1)
Keep the phone tree short
Observed menu levels before a person or terminal route.
SourcesOptional phone-test call instrumentation (instant-phone/v1)
Provide a route to a person
Whether the tested route reached a person.
SourcesOptional phone-test call instrumentation (instant-phone/v1)
Show who answered
Whether a person, AI receptionist, voicemail or unknown answer handled the call.
SourcesOptional phone-test call instrumentation (instant-phone/v1)
Reach a person promptly
Elapsed time from connection to a person.
SourcesOptional phone-test call instrumentation (instant-phone/v1)
Answer the caller's question
Whether the de-identified answer assessment found the question answered.
SourcesOptional phone-test call instrumentation (instant-phone/v1)
Offer a next step
Whether the de-identified answer assessment found a next step.
SourcesOptional phone-test call instrumentation (instant-phone/v1)
Provide a useful voicemail greeting
Whether voicemail named the business, gave hours or a callback promise, and offered another channel.
SourcesOptional phone-test call instrumentation (instant-phone/v1)
Limits and handling
A focused snapshot, not a complete journey audit
Website check
The website check does not submit forms or contact the business. Finding a contact option does not prove it works or that staff respond. An optional recorded AI phone test requires its own agreement.
Priorities stay bounded
The result shows up to 3 fixes, each from a different area, and up to 2 supported strengths from distinct areas. Sparse evidence is never padded.
Reuse keeps its date
A compatible result may be reused only within the 24-hour server-enforced reuse window. The original source and observation dates remain with it.
Automated checks flag common issues; they are not a legal or ADA compliance determination.
Retention schedule
Website evidence: set by configuration. Results and share links: 548 days. Hashed IP addresses and abuse counters: set by configuration. These periods do not apply to the separate optional phone-test recording or request record, which have no automatic deletion date.
Ready to see your score?
Run your free instant CX audit
Start with a website or business name to see what helps or hurts your customer conversion path.