Why your Google Ads stopped working

The most common cause of an account that "stopped working" is that it did not, and the measurement did. A tag removed during a site deployment, a consent banner change, a conversion action edited, a thank-you page URL that moved — all of these produce a chart that falls off a cliff while the business carries on receiving the same enquiries.

This is worth checking first for two reasons. It is quick, and it is the only cause that is genuinely invisible from inside the account: everything looks correctly configured, the campaigns are running, and the conversions have simply stopped. It also matters more than it used to, because automated bidding optimises against the conversion signal. Break the signal and the bidding does not merely report badly, it starts making genuinely worse decisions within days.

1. 1.Compare Google Ads conversions against your CRM or inbox for the same period. If the CRM is flat and Ads has collapsed, it is tracking.
2. 2.Check whether the site was deployed around the date performance changed. Ask the developer, not the marketing team.
3. 3.Look at the conversion action itself — was it edited, paused, or its counting method changed?
4. 4.Check the consent banner. A change in defaults or a new CMP can stop tags firing for a large share of visitors overnight.
5. 5.Submit a test enquiry and watch it arrive. This takes four minutes and settles the question definitively.

Once tracking is eliminated, performance changes come from a short list. The table separates them by the signature each leaves, because they look similar in a summary chart and completely different one level down.

This single question narrows the cause faster than anything else, and it is skipped surprisingly often because the headline chart shows conversions and everybody starts there.

If impressions fell, the searches are not happening or you are no longer entering the auction. That is demand, budget, or a targeting change. If impressions held and clicks fell, you are entering the auction and losing it — a competitor, a position drop, or creative people have stopped responding to. If clicks held and conversions fell, the account is doing its job and the problem is on the site or in the measurement.

Each of those three leads somewhere completely different, and diagnosing the wrong one wastes weeks. An agency that responds to a performance drop by rewriting ads without establishing which of the three moved is guessing in an expensive direction.

Smart bidding is genuinely good and it is entirely dependent on the conversion signal it is optimising towards. When that signal degrades, the bidding does not stop — it continues confidently in the wrong direction, which is worse than stopping.

• If conversions stop being recorded, bidding sees a collapse in performance and pulls back, so spend and volume fall together and the account looks worse than it is.
• If a soft conversion is added — a page view, a phone click with no duration threshold — bidding optimises towards it enthusiastically and you get a great deal of the wrong thing.
• If conversions are counted twice, cost per acquisition halves on paper and bidding becomes more aggressive than the real economics support.
• If offline conversions stop importing, the platform loses sight of which leads became customers and reverts to chasing form fills.
• Any significant edit restarts the learning period, so an account changed weekly never stabilises and its performance looks permanently erratic.

A sequence rather than a list, because the order matters and each step eliminates a category before the next begins.

1. 1.Establish whether the business actually noticed. If enquiries are unchanged and only the dashboard fell, this is measurement and nothing else.
2. 2.Open the change history and read the last thirty days. Somebody frequently changed something and did not mention it — including auto-applied recommendations, which count as changes made by Google on your behalf.
3. 3.Determine which metric fell: impressions, clicks or conversions. Do not proceed until this is answered.
4. 4.Read the search terms report for the period. A shift in what you are matching against explains a surprising number of these.
5. 5.Check auction insights month over month. A new advertiser with a larger budget is a real and unfixable cause, and knowing that early prevents wasted optimisation.
6. 6.Only then change something, and change one thing.

Sometimes nothing is broken. A larger competitor entered, prices moved, demand shifted, or a category that was cheap became expensive because it got noticed. These are real and the correct response is commercial rather than technical.

The test is impression share against cost per click over several months. If impression share is falling while your cost per click rises, you are being outbid rather than mismanaged, and no amount of account optimisation reverses that. The options at that point are narrowing to segments where you can still win, improving conversion rate so you can afford a higher click price, or accepting a smaller share at a sustainable cost.

That is an uncomfortable conversation and it is the honest one. An agency that responds to a genuine auction shift by promising to optimise its way out is either not looking at auction insights or is unwilling to say what they show.