A store was spending on ads every day. Orders were coming in. But Google Ads showed almost nothing.

So the first instinct was to blame the ads. New creatives. New audiences. New campaigns.

The real problem was the tracking. The purchase event wasn’t firing properly, and the platform was optimising on broken data.

This is the single most common story I see in Google Ads accounts, and it’s the reason I fix tracking before I scale any budget. If Google Ads not converting is your problem, read this before you change anything else.

The symptom and the disease

Here’s how the problem usually shows up. Ads Manager shows few or no conversions, but the business is getting enquiries or orders. Cost per conversion looks terrible, so budgets get cut, and performance gets worse. The team changes creatives and audiences, but nothing improves, because the ads were never the problem. Reports to management look bad, and confidence in the channel collapses.

The disease underneath is usually one of these: conversions aren’t recorded, they’re recorded twice, they’re recorded with the wrong value, or they’re recorded for the wrong action. In every case, the platform learns from bad data, and every decision made on that data makes things worse.

Check 1: Is anything being recorded at all?

Start in Google Ads under Conversions and ask three questions. Does the conversion action exist, and is it set to count the right thing? A purchase should count every purchase. A lead form should count every submission. Is the conversion action receiving conversions in the last 30 days? Zero conversions over weeks, while your business receives orders, is the clearest possible signal. And is the tag actually on the site? Use Google Tag Assistant or Tag Manager preview mode to see what fires when you complete the action yourself.

If nothing fires, nothing else matters. Fix the tag first.

Check 2: Run a real test order

This is the check I run on every account, and it takes ten minutes. Go through your own checkout or lead form exactly as a customer would. Watch in real time whether the conversion event fires, in Tag Manager preview, in GA4 DebugView, and in Google Ads. Check the value recorded: a purchase of PKR 10,000 recorded as PKR 1, or as no value at all, corrupts every revenue based decision the platform makes. And check for duplicates. One order should equal one conversion, not two or three.

Don’t trust a setup you haven’t tested yourself. A tag that looks right in a dashboard and fails on a real checkout is more common than you’d think.

Check 3: Deduplication, especially with multiple tools

Many accounts run Google Ads conversion tracking, GA4 imported conversions, and sometimes a third tool, all at once. If the same purchase is counted in two places and both are set as primary conversions, Google Ads sees double the results and optimises for a reality that doesn’t exist.

The rule is simple: decide one source of truth for each conversion type, set it as primary, and set the others to secondary or remove them. Then verify with a test order that one purchase equals one recorded conversion.

The same logic applies to Meta Ads, where Pixel and Conversions API events must be deduplicated with matching event IDs. The principle is identical: one real world event, one recorded conversion.

Check 4: Compare Ads Manager with GA4

Google Ads and GA4 will never match exactly. They attribute conversions differently, they handle sessions differently, and privacy settings affect them differently. A small gap is normal.

A big gap is a warning. If Ads Manager reports 10 conversions and GA4 reports 100, or vice versa, something is broken: a tag firing on the wrong page, a conversion counted on a click instead of a purchase, a form that submits without validation, or a thank you page that can be reached without converting.

Investigate the gap before you trust either number. Usually the truth is in the middle, and usually the cause is embarrassingly simple once found.

Check 5: Are you tracking the right action?

Sometimes tracking works perfectly, but it tracks the wrong thing. I’ve seen all of these: counting a button click as a lead, when most clickers never submit the form. Counting a page view of the thank you page, which can be reached by refreshing or by old bookmarks. Counting every form start as a conversion instead of every completed submission. Optimising for traffic when the goal is sales.

If the platform is optimising for a shallow action, it will find you more of that shallow action, efficiently and uselessly. Make sure the conversion action matches the business outcome you actually want.

Only after tracking: look at the ads

Once tracking is verified, and only then, work through the account in this order.

Search terms first. What searches actually triggered your ads? Irrelevant searches burn budget fast, so add negatives aggressively. Then landing pages: is the click landing on a page that matches the ad and makes buying easy? Slow pages and generic homepages kill conversion rates. Then offers and pricing: is there a reason to buy now, and is the price competitive? No ad can fix an offer nobody wants. Then ad copy: does the ad match the search intent, and does it say what the customer gets and why they should choose you? And finally bidding and budgets: is the budget large enough to generate meaningful conversion data, and is the bidding strategy matched to your goal?

Notice the order. Tracking, then waste removal, then pages and offers, then ads. Most people do it backwards, and that’s why most accounts stay broken.

The scaling rule

Here’s the rule I follow, and the one I recommend to every client: never scale spend on an account whose tracking you haven’t verified in the last 30 days.

Scaling with broken tracking doesn’t just waste the extra budget. It teaches the platform’s automated bidding to chase the wrong signals, which makes the account harder to fix later. Ten minutes of checking saves weeks of wrong decisions.

Scale gradually on verified data. Increase budgets on campaigns with proven cost per lead or cost per sale, watch the numbers for a week, then scale again. Sudden 5x budget jumps destabilise even healthy accounts.

It also helps to set a simple rule for yourself before you scale: decide in advance what cost per lead or cost per sale would make you pause and investigate, and what number would make you scale further. Writing those thresholds down before the money moves keeps decisions calm and consistent. Without them, it’s easy to react to a single bad day by cutting a campaign that was actually performing, or to keep funding a campaign that has quietly stopped working. Good tracking gives you the numbers. Clear thresholds turn those numbers into decisions you can defend.

When tracking is fine and the ads really are the problem

To be fair, sometimes tracking checks out and the ads genuinely are the problem. Here’s how to tell the difference.

If conversions are verified, firing once, with correct values, and the account still produces clicks without sales, then look hard at intent. Are your keywords too broad? A business selling premium services that bids on generic informational searches will get curious clickers, not buyers. Tighten match types, add negatives, and build ad groups around buying intent searches specifically.

Next, look at the landing experience on a phone, since that’s where most clicks land. If the page takes more than a few seconds to load, if the form asks for too much information, or if the price is hidden until the last step, fix those before touching the ads again. I’ve seen conversion rates double from a shorter form alone.

Finally, check the offer honestly. If competitors offer free delivery, a clear guarantee or a lower price, and you offer none of those, no ad copy will close the gap. Ads amplify an offer. They cannot invent one.

The point of this section isn’t to excuse bad ads. It’s to make sure you fix the right thing. Tracking first, then intent, then pages and offers, then ad copy. In that order, every time.

A real-world pattern

The pattern repeats across industries. An account looks like it’s failing. The team wants to rebuild everything. We check tracking first and find one of three things: conversions not firing, conversions double counted, or the wrong action being optimised for. We fix that, wait two to three weeks for clean data, and suddenly the account that looked dead is performing. The ads barely changed. The data did.

That’s why tracking isn’t a technical detail. It’s the foundation of every paid media decision. Without it, you’re not managing campaigns. You’re guessing with a budget.

The bottom line on Google Ads not converting

If your Google Ads spend isn’t converting, resist the urge to change creatives, audiences or budgets first. Run the checks in this post: confirm conversions fire, run a real test order, fix deduplication, compare Ads Manager with GA4, and confirm you’re tracking the right action.

Most of the time, the ads were never the problem. The data was. Fix the data, and the ads finally get a fair chance to work.

Frequently Asked Questions