Fireshift / Study / Marketing agency
Marketing agency sites and Google Ads conversion tracking: 41.9% count a click, page load or login
TL;DR. 54 of 129 marketing agency sites (41.9%, 95% interval 33.7 to 50.5%) send Google Ads a conversion on a click, a page load or a login. That is below the all-site 50.8%, and the interval stays clear of it. Four more gaps show: 17.1% convert on a page load, against 26.1% of all sites; 9.3% have a duplicate conversion, against 20.5% of all sites; 0.0% count a login, against 7.9% of all sites; 86.8% send no conversion value, against 68.8% of all sites. We saw the container, not the Ads account: a conversion we flag may be secondary and harmless to bidding.
What did the 129 marketing agency sites send?
Share of marketing agency sites with each pattern, next to all 886 sites in the study. A site can show several patterns, so rows overlap.
| Pattern | marketing agency (n=129) | 95% interval | All sites (n=886) |
|---|---|---|---|
| Any noise trigger | 41.9% 54 of 129 | 33.7 to 50.5% | 50.8% 450 of 886 |
| Button or link click | 29.5% 38 of 129 | 22.3 to 37.8% | 31.6% 280 of 886 |
| Page load (every page or a plain page view) | 17.1% 22 of 129 | 11.5 to 24.5% | 26.1% 231 of 886 |
| Duplicate conversions | 9.3% 12 of 129 | 5.4 to 15.6% | 20.5% 182 of 886 |
| Login counted | 0.0% 0 of 129 | 0 to 2.9% | 7.9% 70 of 886 |
| No conversion value on any action | 86.8% 112 of 129 | 79.9 to 91.6% | 68.8% 610 of 886 |
Intervals are Wilson 95%. Where the marketing agency interval holds the all-site share, we can't say marketing agency sites differ.
What does each pattern mean?
- Click. The conversion fires when someone presses a button or link, not when a form is accepted. It counts people who clicked and then failed to submit. Fix: button click conversions.
- Page load. The conversion fires on a page view of a page that is not a clear success page, or on every page. Fix: conversions on every page.
- Duplicate. The same label sits on two tags, or one trigger fires two labels. Fix: duplicate conversions in GTM.
- Login. A login or account page is a conversion. Fix: login and trial conversions.
- No value. No conversion sends a value, so value-based bidding has nothing to learn from. Fix: conversion value for lead gen.
What can't we see?
- Primary vs secondary lives in the Ads account. A click conversion set to secondary does not drive bids.
- Which plugin or tag manager created the tag. We read the public container. We make no claim about plugins.
- Server-side tags are not in the public container, so real rates are likely higher.
- Cause. Sites in one segment often share tools and agencies. These are shares of what we saw, not proof that the segment causes them. The sample is not random.
Method
Static HTTP read of the homepage, /pricing, /signup and /demo (robots.txt respected) and the public googletagmanager.com containers of 886 sites with a readable Google Ads conversion. A site counts as marketing agency when the study's sample list files it under that segment; each site is in one segment. No forms, logins or bot evasion. Full rules: method. All-site results: the study.
Check your own marketing agency site.
The free scan runs the same checks in seconds. Public data only.
Scan my siteRelated: the full 1,727-site study · primary vs secondary · all guides