There’s a particular phone call I’ve had more than once. A business owner, half apologetic, says: “Google says the ads are converting. But we’re just not seeing it in the sales.” They’ve usually been told they’re imagining it. They almost never are.
Ad platforms count what they’re told to count. If the account has been told the wrong thing — and a surprising number have — the dashboard will report success while the bank account reports the truth. Here’s how to find out which one is lying to you, in five checks you can do yourself.
1. What does your account actually call a “conversion”?
In Google Ads, go to Goals → Conversions. Look at the list of conversion actions and ask one question of each: is this actually a sale?
You’re looking for things like “Page view”, “Time on site”, “Clicks to call”, “Directions”, “Add to basket” marked as primary conversions. Any of those set as primary means Google’s bidding is chasing them — and reporting them — as if they were revenue. The worst case I’ve seen this year: a retailer’s dashboard claimed €24 back for every €1 spent, while their till showed 29 cents. The account had been told a page view was worth €1, so Google went and bought page views — tens of thousands of them, as cheaply as it could find them, mostly from the other side of the world. It was doing exactly what it was asked. It was asked for the wrong thing.
2. Are the conversion values real?
Even where the action is right, the value often isn’t. If every “purchase” records at £1, or at some ancient default, the return figure is fiction. Your average order is a number you know — if the platform’s “conversion value ÷ conversions” doesn’t look like it, the tracking is wrong.
3. Does the spend show up in your till?
This is the only comparison that matters, and almost nobody does it: monthly ad spend on one side, and on the other, sales from your store’s own records — Shopify, WooCommerce, your EPOS. Not Google’s number. Not Analytics’ number. The till’s number. If spend went up for three months and the till didn’t move, you have your answer, whatever the dashboard says.
4. Where did the clicks actually come from?
In the account, look at search terms, placements and — for Performance Max — the location data. Cheap clicks are cheap for a reason. If you’re a Dublin shop and half your traffic is from countries you don’t ship to, or from placements no customer of yours has ever visited, the machine has been optimising for volume because volume is what it was told to value.
5. Do Google Ads and Analytics even agree?
Google Ads and GA4 will rarely agree with each other, and neither is the truth about revenue — your store’s records are. Use Analytics for shape (which channels bring people who actually buy) and the till for money. When paid traffic converts at a fraction of your free search traffic’s rate, that’s not an advertising industry mystery. It’s a targeting problem with a paper trail.
What it looks like when it’s fixed
That same retailer — the 29-cents one? The history told a different story once the junk was stripped out: two years earlier, with tracking pointed at real purchases, the same account had returned over €2 back per €1. The channel wasn’t broken. The instructions were. The fix was measured in settings, not in budget.
If this sounds familiar
I do this examination as a fixed-price piece of work: the Forensic Ads Audit — £695. It’s done at the data level through the Ads API, not by skimming screens; every finding in the report carries the query that proves it, so nothing rests on my say-so. It’s independent — you can act on it with me or hand it to anyone — and if we do go on to rebuild or run your ads, the £695 is credited in full.
The Forensic Ads Audit, and what it covers →
Or if you’d rather sanity-check things yourself first: run the five checks above. Check number one finds the problem more than half the time.