POST-CLICK · PRE-FLIGHT

Nobody watches the page after the click.

Every paid click lands on one page. PlayHero reads that page the way a first-time buyer does, scores the handful of things that decide whether it converts, and hands you the fix before the spend starts — then keeps watching after launch.

Free · no login · reads the published URL · about 3 minutes

LANDING PAGE · PRE-FLIGHT

LAUNCH WITH FIXES
64
Human UX 65
Agent readiness 56
  • Destination reachablePASS
  • CTA Clarity45 · benchmark 76FIX
  • LoadFIX
  • Form frictionFIX

playhero.ai/lumen · scanned 2026-09-16 · our own page, unfixed on purpose

THE GAP

The ad gets reviewed. The page gets ignored.

The ad platform approves the creative. Uptime says 200 OK. Analytics reports the conversion rate after the budget is gone. Nobody's job is to read the page the click lands on — so the offer slips below the fold, the form grows a field, the hero CTA loses to a nav link, and the spend keeps running against it.

Ad platform

Reviews the ad against policy. Approves or rejects it. Never scores the destination.

Ad only

Uptime monitor

Pings the URL. 200 OK. Can't tell a working page from a broken one that loads.

Status only

Analytics

Reports conversion rate after the fact. Tells you the number moved, not which element did it.

After the spend

PlayHero

Reads the published page as a visitor. Scores 6 human-UX pillars plus agent readiness, names the fix, re-checks daily.

THE PAGE ITSELF

WHAT DECIDES THE CLICK

The handful of things that decide whether a page converts.

Most landing pages lose for the same few reasons. The scan checks each one against the published page and a benchmark from the PlayHero scan corpus — and every check below names the exact signal it reads, so the score is never a matter of taste.

01Signal Hierarchy · CTA Clarity

The first screen does most of the work.

A buyer decides in seconds whether this page is for them. If the headline doesn't land, the subhead doesn't explain, or the action isn't obvious, the scroll never happens — and the click you paid for bounces.

  • Hero block detected, with one primary headline and a supporting subhead
  • Heading order reads top-down without nesting errors · visible headings only
  • Primary CTA found in main content, not borrowed from the nav
02CTA Clarity

One obvious next step, not five.

"Get started", "Book a demo", "See pricing", "Talk to sales" and "Watch video" on one page is not choice — it's hesitation. The scan counts every CTA-like button and link in the main content and scores the page down when the primary action has competition.

  • Competing CTAs counted · more than one costs points, more than three flags it
  • Self-serve path matched to the GTM motion · sign-up vs. demo-gated
  • Critical GTM mismatch on the page lowers the CTA score
03Content Clarity · Messaging Clarity

Copy that says who it's for and why you.

Clear beats clever. A page that never names its audience or what makes it different reads like every other page in the category. The scan looks for the sentences a skeptical buyer needs and marks their absence, not just their presence.

  • Audience named on the page · "for RevOps teams", "for Shopify stores"
  • Differentiation stated · missing both audience and differentiation costs 10 points
  • Value propositions counted; GTM motion and positioning themes resolved
  • Nav complexity penalised · more than five nav items dilutes the message
04Conversion Friction

The form is where the click dies quietly.

A form with too many fields, no labels, or a demo gate on a self-serve product loses submissions without ever throwing an error. The scan reads the primary form in the main content and checks it against what the motion can bear.

  • Field count vs. the threshold for the page's motion · hidden, checkbox and radio inputs excluded
  • Label coverage · fewer labels than fields costs points
  • Demo-gated or no self-serve path on a PLG motion flagged
  • High-impact accessibility blockers subtracted; a conversational qualifier earns credit
05Load Intelligence

Slow pages bounce before they get to convert.

The heaviest hero image in the deck wins the review and loses the click. Load is measured, never assumed: real Core Web Vitals against Google's published thresholds, and if the measurement fails the pillar reports 0 with the reason instead of a guess.

  • Largest Contentful Paint · over 2.5 s caps the pillar at 75, over 4 s caps it at 50
  • Cumulative Layout Shift · over 0.1 costs 8, over 0.25 costs 15
  • Third-party scripts enumerated by origin and weight
06Trust signals · evidence

Proof where the decision happens.

Buyers trust other buyers more than they trust you. The scan detects the proof elements on the page and records them as evidence beside the finding, so the finding for a weak CTA can say "no testimonial within reach of the form" instead of "add social proof."

  • Testimonial block detected · quote + attribution
  • Customer logo row detected
  • Security or compliance badge detected
  • Pricing visible; free trial or free plan present

Hollow dots: detected and shown as evidence, not yet weighted in the score.

How the number is built

score = human UX × 0.88 + agent readiness × 0.12 · human UX is the mean of the 6 pillars above; benchmarks come from the PlayHero scan corpus, never your CRM.

THE FINDING

Every finding comes with the fix.

One scan returns a verdict and up to three findings. Each one names the problem, shows the score against the benchmark, quotes the evidence from your page, and tells the developer what to change. Copy it into Jira or Linear as-is.

READY

No high-severity finding. Ship it.

LAUNCH WITH FIXES

A high-severity finding needs an owner before the spend starts.

HOLD

The destination can't be captured or is blocked. Nothing to score yet.

Rallycard · CTA Clarityplayhero.ai/lumen

Primary CTA visibility issue

45/ benchmark 76high
Finding

The CTA scores 45/100, significantly below the benchmark of 76, and lacks clear text, which can confuse users about the next action to take.

Fix

Define a clear, action-oriented text for the primary CTA and ensure it is visually distinct.

ApproveDeferDismiss→ owner · re-check daily

AFTER LAUNCH

From alert to verified fix.

Daily, change-gated re-checks on every monitored page. Alerts by email and Slack. Nothing changes without a person approving it, and nothing is marked fixed until the next scan proves it.

EXAMPLE INCIDENT · UNMONITORED

The fix went live. The next scan verified it.

A deploy moved the hero CTA below the first mobile viewport while the campaign kept spending. PlayHero caught the drop on the daily re-check, minted a Rallycard, a human approved the fix, and the next scan closed it.

Change detected
daily re-check
Rallycard created
CTA · high · evidence attached
Human approved
owner assigned · Slack
Next scan verified
score restored · closed

NOTHING TO INSTALL

Reads the page you published. Not your stack.

Traffic

Meta, Google, LinkedIn, TikTok, Reddit, Microsoft — anywhere the click comes from.

Pages

Webflow, Framer, Unbounce, Instapage, Leadpages, Shopify, WordPress, HubSpot, Vercel — the builder does not matter.

Tests

PlayHero writes the test plan. Copy it into Jira or Linear, then run the split in your builder or experimentation tool.

Agents

Rallycards are readable by any agent you authorise over MCP. Agent readiness is scored on every page alongside human UX.

PROOF · OUR OWN PAGE

The specimen on this page is ours.

Follow its evolution →

The scan at the top is playhero.ai/lumen — a live PlayHero page, scanned 2026-09-16, not a demo account. It came back LAUNCH WITH FIXES: overall 64, Primary CTA visibility issue at 45 against a benchmark of 76. That finding is real, and so is the score, unfixed. When the fix ships, the next scan will say so here.

urlplayhero.ai/lumen
scanned2026-09-16
overall64
human ux65
agent readiness56
verdictLAUNCH WITH FIXES
top findingCTA Clarity 45 / benchmark 76

PRICING

What is free. What stays monitored.

THE PROOF

Free

$0

Enough to prove the instrument is real.

  • 1 site · 3 scans a month
  • Human UX and Agent Readiness scores
  • Full seven-pillar breakdown
  • A sample of 3 buyer queries
  • Up to three findings, in full
  • Remaining findings shown as a count
  • Public share link
  • No weekly queries · no history · no alerts
Start free

No credit card. No trial clock.

YOUR OWN SITE

Brand

$24/ site / month
One domain, up to 5 pages, per site
$240 / site / year · two months free

You own one property and you need to be told the moment it changes.

  • Monitoring
  • Daily scans
  • Every finding, not a top three
  • Alerts by email and Slack
  • The record
  • Score history and visual diffs
  • Heroboard: approve, dismiss, defer — permanent audit trail
  • Resolution Briefs included
  • Unlimited teammates
  • Lumen
  • 10 buyer-intent queries per site across 5 AI platforms, every week.
Start 14-day trial

Card not charged until day 14 · cancel any time

YOUR CLIENTS' SITES

Agency

$24/ site / month
One domain, up to 5 pages, per site
$240 / site / year · two months free

Your clients' sites, your name on the report.

  • Everything in Brand, plus
  • Portfolio
  • Portfolio view — clients ranked by what needs attention
  • Separate client workspaces and access
  • Client-facing
  • Resale rights — bill your clients at your own price
  • Billing
  • Pay annually and get two months free
  • 10 buyer-intent queries per site across 5 AI platforms, every week.
Start 14-day trial

Card not charged until day 14 · cancel any time

Above 250 sites, talk to us →

Pay per site, not per seat.

FAQ

Questions before the spend starts.

Do I need to install anything?

No. PlayHero reads the published URL. The builder and host do not matter.

Isn't this what my uptime monitor does?

No. An uptime monitor confirms the URL responds. PlayHero reads the page and scores what a visitor sees — layout, copy, load, and the form — and names what to fix.

Does PlayHero run the A/B test?

No. It scores the page and writes the test plan. Run the split in your builder or experimentation tool.

Does a ready verdict mean an ad platform will approve the ad?

No. A verdict is PlayHero's page-health read, not a policy ruling or an approval from Meta, Google, or another platform.

How often is a monitored page checked?

Daily, change-gated: a lightweight check every day, a full scan when the page changes. Brand and Agency plans include it.

Can PlayHero pause my ads when a page breaks?

No. PlayHero detects page problems and alerts the team. It does not pause campaigns on its own.

Is the score subjective?

No. Every pillar names the signals it reads and the benchmark it is scored against, and interpretive findings are labelled as such. Load is measured or reported as unavailable — never guessed.

POST-CLICK · PRE-FLIGHT

Score the page before the next campaign.

Read the published destination before launch. Start free, no login or installation.

Free · no login · reads the published URL · about 3 minutes