Get early accessEarly access
For SaaS marketing sites · early access

When signups drop, know where to look.

RevDetector watches the pages that turn visitors into trials and paying customers. It puts real-user speed, errors, traffic sources, attribution and content checks on one timeline, so you can see which page changed and hand your developer a concrete place to start.

Get early access See how it works
  • One script tag
  • Real visitors, not lab runs
  • Personal data off by default
acme.io · Trial signups Oct 27 – Nov 9 · UTC Sample data
Trial signups · 7d 142 −31.2% vs prior
LCP p75 · mobile 3.9s +1.4 s
Paid sessions · 7d 6,980 +2.1% vs prior
Trial signups per day— this period · - - prior period
30 20 10 RELEASE 4.12 Oct 27 Nov 4 Nov 9
Leak /pricing → /signup · iOS Safari

Trial signup form throws on iOS Safari after release 4.12

TypeError in form-validate.js on 61% of mobile Safari submits. Desktop completion is unchanged.

What it watches

Four signals on one timeline

Most teams check these in four different tools. RevDetector lines them up against the goal you care about, page by page.

Signal 1 of 4

Site health and real-user speed

Core Web Vitals and JavaScript errors from your actual visitors, per page, device and release.

  • LCP, INP, CLS at p75
  • Errors tied to the session
Signal 2 of 4

Traffic

Sessions by source, page and device, with bots filtered out and the prior period beside it.

  • Sources and landing pages
  • Bot-filtered sessions
Signal 3 of 4

Marketing attribution

Which source, campaign and entry page led to each confirmed trial and paid upgrade.

  • UTM and click-ID capture
  • Entry-page attribution
Signal 4 of 4

Content and SEO quality

Indexability, titles, canonicals and broken links on the pages that earn search traffic.

  • noindex and canonical checks
  • Broken links and redirects
How an investigation runs

From a falling number to a ticket your developer can act on

  1. Step 1

    Confirm the drop is real

    Check the goal against the prior period and see how much traffic was observed, so you don’t chase a tracking gap.

    Trial signups142 vs 206
    Sessions observed21,430
    Goal sourceSignup backend
  2. Step 2

    Narrow it to one page

    See which page, device and release the change lines up with, and what else moved there at the same time.

    /signup · iOS SafariLeak
    /pricing · LCP mobileWatch
    /blog/* · organicHealthy
  3. Step 3

    Hand off the evidence

    Share the page, window, error and the test to run, then recheck the goal after the fix ships.

    ReproduceiPhone · Safari 18
    Errorform-validate.js:88
    Recheckafter next release
Walk through a sample investigation
Leak /pricing → /signup · iOS Safari Sample data

Trial signup form throws on iOS Safari after release 4.12

TypeError in form-validate.js on 61% of mobile Safari submits. Desktop completion is unchanged. Start with the validation script shipped Tuesday, then start a test trial from an iPhone and confirm the account is created.

Window
Nov 3 – 9
Sessions hit
1,208
Goal impact
−64 trials
What a finding looks like

Evidence, not a verdict

Each finding says what changed, where and for whom, and what to check first. You decide whether it’s the cause.

Scope
The page path, device and browser where the change shows up.
Window
The dates the evidence covers, compared with the same length before.
Impact
How many goal completions the affected sessions would usually produce.
Next check
A test a person can run to confirm or rule it out.
Where you look today

The questions, and the tabs they usually take

The questionUsually answered inIn RevDetector
Did trial signups really drop, or did tracking break?GA4, the product database, Tag AssistantGoal count from your signup backend beside observed sessions
Which page got slower for real visitors?PageSpeed Insights, CrUX, a performance toolp75 vitals per page and device, marked by release
Is a script error stopping the form?Sentry, the browser console, a QA ticketErrors linked to the sessions that didn’t convert
Did a campaign stop sending the right visitors?GA4 acquisition, the ad platform, a UTM sheetSource and entry page for each confirmed signup
Did a page fall out of search?Search Console, an SEO crawlerIndexability and canonical checks on your key pages
Setup

One script tag, then pick your goal

Add the snippet to your marketing site, mark the event that counts as a trial or a purchase, and choose the pages to watch. Data starts arriving with the next visitor.

  • Personal data stays off until you switch it on per site
  • Works alongside GA4, GTM and your consent banner
  • Bots filtered before anything is counted
<!-- Add before </head> on every page -->
<script src="/sdk/revdetector.min.js"
        data-site-key="pk_live_…" defer></script>

// Mark the outcome that counts
RevDetector.track({ metric_type: "web.funnel", metric_name: "trial_started" });
Guides

Check your site after the next release

Get early access

We’re inviting SaaS growth teams in small groups. Leave your work email and we’ll write once, when your invite is ready.

One email when your invite is ready. No newsletter.