Product

Speed, traffic, attribution and content, measured against the goal that pays

Every chart in RevDetector starts from one question: did this change the number of people who started a trial or became paying customers? Here is what each signal records and how it shows up.

Signal 1 · Site health and real-user speed

See which page got slower, for whom, and since when

RevDetector records Core Web Vitals and JavaScript errors from real visitors. Each page shows p75 values by device, marked with your release dates, beside the goal it feeds.

  • LCP, INP and CLS at p75 per page and device
  • Errors grouped by script and linked to sessions that didn’t convert
  • Release markers from your deploys
Key pages · mobile · p75Sample data
PageLCPINPCLSErrorsState
/pricing 3.9 s 180 ms 0.04 12 Watch
/signup 2.1 s 160 ms 0.02 738 Leak
/ 1.8 s 120 ms 0.01 4 Healthy
/checkout 1.6 s 140 ms 0.03 2 Healthy
Signal 2 · Traffic

Know whether fewer people arrived, or fewer converted

Sessions by source, landing page and device, with known bots removed before counting. Compare any window with the one before it to separate a traffic dip from a conversion problem.

  • Source, medium and landing page per session
  • Bot filtering before anything is counted
  • Like-for-like comparison windows
Sessions by source · last 7 daysSample data
Google Ads
6,980
Organic search
6,240
Direct
4,110
LinkedIn
1,870
Referral
1,230
LinkedIn sessions fell 38% after the campaign’s landing URL changed on Nov 5.
Signal 3 · Marketing attribution

Follow each signup back to where it came from

RevDetector keeps the campaign parameters and click IDs from the first page of a session and attaches them to the goal when it fires. When a redirect drops a UTM, you see it.

  • UTM, gclid and other click IDs captured on entry
  • Same-session entry attribution for every goal
  • Flags parameters lost between pages
Path to a trial signupSample data
  1. google / cpc
  2. /lp/attribution
  3. /pricing
  4. trial_started
  • utm_campaign = q4-attributionKept
  • gclid present on entryKept
  • /lp/attribution → /pricing redirectDrops utm_content
Signal 4 · Content and SEO quality

Catch the release that quietly hid a page from search

RevDetector checks the pages you choose for the mistakes that cost organic signups: a stray noindex, a wrong canonical, a missing title, a link that now returns 404.

  • Indexability, canonical and title checks
  • Broken links and redirect chains
  • Checks rerun after each release
Checks · /blog/ga4-key-eventsSample data
  • Indexable (no noindex)noindex since Nov 6
  • Canonical points to itselfPass
  • Title 30–60 charactersPass
  • Internal links resolve2 return 404

Try it on your own site

Early access teams get a guided setup: pick one goal, add the snippet, and review your key pages together once data arrives.