$3 Million in Sales That Never Happened
Toolab was spending $40,000+ a month on Google Ads, and 1 in 5 of the sales it paid to chase were never real. Three hidden GA4 sources were counting the same orders up to 32 times, inflating reported revenue by $3M+ over two years and quietly draining the budget. We found every source of duplication and rebuilt their Tracking to count each order exactly once.
Every figure is measured from the store's own analytics at the transaction level: duplicate counts, phantom revenue, and the post-fix result verified order by order. Client named with permission.
Three GA4 implementations running at once. None visible from the admin.
Toolab came in with what looked like a straightforward tracking fix. Underneath, three separate Google Analytics implementations were running simultaneously on the same property: the platform's native GA4 app, a hidden channel-level GA4 field, and a GTM configuration. None were aware of the others, and all were invisible from the standard store admin.
Together they inflated GA4 revenue reporting by 39% vs actual store sales and fed duplicated, inflated conversion values into Google Ads Smart Bidding. Consent Mode was broken, resetting on every page and creating GDPR exposure. Microsoft Ads was sending malformed revenue with Enhanced Conversions switched off.
The worst of it was outright duplication. The same orders were counted up to 32 times, so 1 in 5 reported sales never happened at all. In 2024, the ten most-duplicated orders were counted 190 times between them for just 10 real purchases. Reported to the ad platforms across two years, that added up to $3M+ in phantom revenue the store never made.
The ad platforms optimize on what you report. Toolab's dashboard showed 45,314 conversions, but only 35,722 were real customers. The platform was paying to chase nearly 9,600 buyers who never existed, so Toolab was effectively paying about 27% more for every real customer than its dashboard showed. At $40,000+ a month in spend, roughly a fifth of the budget, around $8,000 a month, close to $100,000 a year, was pushed toward customers who did not exist. Fixing the data recovers that spend without adding a single dollar to the budget.
Trace every source. Remove the duplicates. Rebuild what was broken.
We traced all three overlapping GA4 sources down to a single GTM pipeline, including one hardcoded source found only through direct code-level inspection of live pages. Consent Mode V2 was rebuilt from scratch. Microsoft Ads purchase tracking was rebuilt with clean revenue, correct UET configuration, and Enhanced Conversions switched on.
The standard approach of checking the admin panel would not have found the hidden channel-level GA4 field. This required code-level investigation of the live site to surface every active tracking source.
What we fixed, source by source.
Purchase data that reconciles with real sales. One source of truth.
Purchase data now reconciles with the store's real sales, verified order by order at the transaction level with zero duplication. Google Ads and Microsoft Ads bid on real, consent-compliant revenue instead of inflated, duplicated numbers.
Not sure your revenue numbers are real?
If GA4 and your store sales do not match, your ad platforms are bidding on fiction. We find out why.
Get Your Free Audit →The Problem That Surfaced Three Weeks Later.
About three weeks after delivery, the client noticed Google Ads spend and conversions had dropped. A full independent diagnosis revealed two separate, unrelated problems:
1. A billing failure on the Google Ads account, the client's card declined on July 22, cutting ad delivery roughly 40%. Confirmed from the account's billing log, correcting the client's own assumed date.
2. A Google Ads attribution failure starting five days earlier (July 17), completely independent of the billing issue. Both Google Ads conversion paths collapsed together on this exact date while GA4 continued recording purchases normally. That pattern, both ad-platform paths failing in lockstep while GA4 stayed healthy, pointed specifically at the click-attribution layer.
Root cause: the client's separate Cookiebot subscription had lapsed on the same failed card. With no consent banner rendering, no visitor could grant consent, so ad storage stayed permanently denied. No click cookie meant Google Ads could not tie any sale back to its ad click, even though the sale was recording correctly in GA4 the whole time.
The client restored the Cookiebot subscription. The banner resumed immediately, consent states updated, and no tracking rebuild was required. The entire attribution failure was caused by a third-party billing lapse that was invisible from Google Ads and GA4 directly.
This is exactly what ongoing monitoring is for. A silent, unrelated billing lapse on a third-party tool, invisible from the ad platforms, quietly breaking attribution for over a week before anyone noticed. A monitored setup catches this in days. If you want that coverage for your own account, start with a free audit.
He always delivers 140%! It's great work every time. This is my third or fourth time working with him, and he goes above and beyond every single time.
If you run ecommerce and cannot say for certain that your revenue tracking is accurate, you are making budget decisions on numbers you have not verified. That is the first thing we check.
Get the Same for Your Business →Suspect Your Tracking Has the Same Problem?
Duplicate events and hidden GA4 sources are more common than most teams realize. We trace every active source, verify deduplication at the transaction level, and give you one number you can trust. Start with a free audit.
No payment to start · 24-hour response · Findings before implementation