Tracking that is wrong
Not firing, counting page views as conversions, or double-counting from two sources. Every optimisation decision made afterwards is guesswork dressed as data.
14 August 2026 · 6 min read
Three faults, in the same order, every time.
Almost every underperforming account I open has the same three problems. None of them are exotic, all of them are visible within an hour, and fixing them usually matters more than any bid strategy.
Short answer
The three most common faults in underperforming Google Ads accounts are, in order: conversion tracking that is broken or counting the wrong event, a search terms report full of queries that could never convert, and landing pages that do not continue the promise made in the ad.
01 The faults
The three faults, in order.
Check them in this order, because each one invalidates the diagnosis of the next.
Search terms nobody would buy from
Broad match plus an unattended account produces spend against queries adjacent to your business but not in it. Frequently a large share of the budget.
A page that contradicts the ad
The ad promises one thing, the page discusses another. The visitor has to work out whether they are in the right place, and Google charges you more for the mismatch.
02 Self-check
How to check each one yourself.
All three take under an hour and need no tools beyond the account itself.
Self-check
- Submit your own form. Then look for it in the account. It should appear exactly once, attributed correctly, within the expected window.
- Read the search terms report. Not a sample — a month of it. Most advertisers have genuinely never opened this screen.
- Check when negatives were last added. If the list has not changed in three months, nobody is managing the account.
- Click your own ad. On a phone. Does the page say what the ad said, and does the form submit?
- Compare conversions against reality. If the account reports forty leads and you received twelve, the tracking is wrong.
03 Order of work
Why bid strategies come last.
Bid strategy gets disproportionate attention because it is the most interesting part of the platform and the part vendors write about. It is also the part that matters least when the foundations are broken, because an automated bidding strategy optimises towards whatever it is told a conversion is.
If your conversion event is a page view of the contact page, automated bidding will efficiently find people who visit contact pages. It will do that very well. It will not find you customers, and no amount of strategy selection changes that — the instruction was wrong.
So the order is: make the numbers true, remove the waste, match the destination, and only then consider how bidding should work. That order is unfashionable and it is why most accounts improve substantially in the first fortnight without anything clever happening.
— Questions
Common follow-ups.
Submit your own form and follow it through. It should appear exactly once in the account, attributed to the right source, within the expected window. If it appears twice, not at all, or as a page view, the tracking is wrong.
It varies too much for a single figure to be honest, and any specific percentage you see quoted is almost certainly invented. What is consistent is that unattended accounts accumulate waste, and the search terms report shows you your own number in ten minutes.
With solid conversion data and a mature negative list, it can find valuable searches you would not have listed. In a new account with neither, it is the fastest way to spend a month’s budget on adjacent curiosity. Earn the freedom rather than starting with it.
It is frequently the cheapest improvement available. A homepage offers choices to a visitor who has already made one. A matched page raises conversion and lowers cost per click simultaneously, because Google scores relevance too.
— More
Related reading.
AEO & GEO · 6 min
Blocking Google-Extended does not remove you from AI Overviews.
This is the single most common technical misunderstanding in AI search right now, and acting on it incorrectly can remove a business from Google Search altogether.
ReadAEO & GEO · 8 min
Two mechanisms. Most GEO advice ignores both.
Almost every argument about generative engine optimisation comes from conflating two completely different things: what a model learned when it was built, and what it fetches when it answers.
ReadSEO · 7 min
One template, forty cities, six months of rankings.
It is still one of the most commonly sold local SEO tactics in India, it works briefly, and it is a named spam pattern in Google’s own documentation.
ReadFree consultation · No obligation
Want this applied |to your business?
The reasoning on this page is public because I would rather be judged on it than on a claim. If you want it applied, that is the service.
- An honest answer, including “do not bother”
- Findings are yours whether you hire me or not
- Fixed price agreed before work starts
- Same-day reply, all 7 days
Start here
Send your enquiry
Name, number, email, what you need.