Google Ads Conversion Tracking: A Five-Minute Setup Check

Google Ads conversion tracking is the link between what people do on your site and what the platform reports and bids on. A five-minute check means opening your conversion actions, comparing each one against real events in analytics, firing one test conversion, and confirming the status, value and ID land where they should.

In short:

• Open the conversion actions list and check the status of every action you rely on.
• Compare recent activity in Google Ads with the same events in your analytics tool.
• Fire one test conversion and confirm it appears with the right value and transaction ID.
• Check the conversion window and whether the action is primary or secondary.
• A change in reported numbers alone does not prove deduplication works; verify the events and identifiers.

What does the five-minute Google Ads conversion tracking check actually cover?

It covers the setup, not the strategy. In five minutes you can confirm that tags fire, that actions report, that values and IDs arrive, and that the actions feeding your bidding are the ones you think they are. You cannot audit attribution models or rebuild a broken data layer in that time, and you should not try.

The check works because most tracking failures are visible at the surface. A tag that never fires shows a status that never changes. A duplicate action shows two entries for one purchase. A bid strategy optimizing for the wrong event usually traces back to a primary/secondary split someone set months ago and forgot.

Run this check after any site release, after a tag manager update, and whenever reported conversions move in a way the traffic does not explain. Website conversion tracking breaks quietly. Nothing errors, the numbers just drift.

How do I check that Google Ads is recording conversions correctly?

Open the conversion actions list in Google Ads and compare the status and recent activity of each action with the same events in your analytics tool. Then complete a test conversion yourself and confirm it appears under the matching action with the correct value and ID.

Do it in this order:

1. Open the conversion actions list and read the status of each action. Anything inactive, unverified or newly created deserves a look before you touch bidding.
2. Compare recent activity per action against the equivalent events in your analytics tool. Match by date and by count, not by a vague sense that "it looks about right."
3. Complete one test conversion on the live site, using a real order or form submission you can identify later.
4. Find that test in Google Ads under the matching action. Check the value, the currency and the transaction or order ID.
5. Only after those four steps, look at the conversion window and the primary/secondary setting.

If the numbers do not line up, check the tag, the window and the primary flag before changing anything. Changing settings to "fix" a number you have not diagnosed usually creates a second problem on top of the first.

Where do the mismatches between Google Ads and analytics come from?

They come from counting rules, not from one tool being wrong. Google Ads attributes conversions to clicks and applies its own window. GA4 applies its own attribution models. The same purchase can appear in both systems and still be correct in both.

A conversion can also be recorded in both tools without being double-counted in Google Ads, because a repeated action carrying the same transaction or order ID is deduplicated. That is the mechanism to test. Compare a specific order ID across both systems before assuming something is broken.

For a deeper walkthrough of why campaign-level numbers diverge between the two systems, see how to fix the GA4 versus Ads gap with campaign URL tagging. If you suspect the tag itself rather than the settings, the Meta Pixel troubleshooting checklist covers the same diagnostic logic on another platform.

Which settings change what Google Ads reports and bids on?

Two settings do most of the work: the conversion window and the primary/secondary flag. Get these wrong and every downstream number is off in a way no tag fix will correct.

The click-through conversion window is 30 days by default. You can set it anywhere from 1 to 90 days, depending on how long your sales cycle runs. Short windows suit fast purchases. Longer windows suit considered purchases with several touchpoints, where the click that started the research deserves some of the credit.

Primary and secondary decide what counts. Primary actions feed the Conversions column and smart bidding. Secondary actions appear only under All conversions. If your bid strategy seems to optimize for the wrong event, check this split first. Custom goals have their own rules, so read them rather than assuming they behave like standard actions.

One trap worth naming: two separate conversion actions for the same purchase. That can distort the total, though the existence of two sources does not by itself prove double counting. Verify the events and the identifiers before you conclude anything.

How do I test a conversion end to end without corrupting my data?

Test with an identifier you can find again, and expect to filter it out afterwards. Use a real order number, a recognizable email or a marked test submission. Then search for that identifier in both systems.

What to confirm on the test:

• The tag fires on the actual thank-you page or on the event, not on the page before it.
• The value and currency arrive correctly, not as zero and not in the wrong currency.
• The transaction or order ID is present and matches what your analytics tool recorded.
• The action that received the test is the one you expected, not a similarly named duplicate.
• The test appears once, not twice.

If you run server-side uploads alongside the browser tag, the same transaction or order ID must be passed on both paths. A consistent ID is what allows deduplication to work. Without it, the two paths cannot be reconciled and you will see inflated counts that look like a tracking bug but are really an identifier problem.

Offline conversion tracking follows the same rule. When you upload conversions from a CRM, the identifier you join on has to match what the click or the browser tag recorded. Google offline conversion tracking fails most often at that join, not at the upload itself.

What does a clean setup look like across platforms?

The same four questions apply on every ad platform, even though the interfaces differ. Does the tag fire on the right event? Is the identifier consistent? Is there a duplicate action for the same outcome? Which actions feed bidding?

LinkedIn conversion tracking through Google Tag Manager follows the same pattern as Google's own tag: one container, one tag per event, and a clear rule about which events count. Conversion tracking in digital marketing generally fails at the same three points regardless of platform: the event fires in the wrong place, the identifier is missing, or two actions claim the same outcome.

Server-side paths deserve their own check. The Conversions API overview explains what a server-side feed replaces and what it does not, which matters when you are deciding whether a second source is redundancy or a duplicate.

Why do reported conversions change after a fix, and why is that not proof?

A change in the reported number is a signal, not evidence. Conversions can move because a tag started firing, because a window changed, because an action moved from secondary to primary, because a duplicate stopped, or because traffic shifted. Several of those produce identical-looking movements.

To prove deduplication works, verify the events and identifiers directly. Send one action with the same order ID twice and confirm it counts once. Send it with two different IDs and confirm it counts twice. That is a test. Watching the daily total drop is not.

Keep a short record of what you changed and when. When someone asks three weeks later why conversions fell, the answer is usually in that record rather than in the platform.

What should I check before trusting the setup?

Confirm that the tag fires on the actual thank-you page or event, that the same transaction or order ID is passed for both browser and server-side uploads, and that no duplicate action exists for the same purchase. Also check the conversion window and whether the action is primary.

Then keep the check routine. Run it after every site release, after any change to your tag container, and any time a bid strategy starts behaving oddly. The five minutes are cheap compared with a month of bidding on the wrong event.

If calls are part of your funnel, they need their own link back to campaigns; call tracking that ties calls to campaigns covers how to keep phone conversions from disappearing between systems. And if you are still stitching UTM parameters by hand, the minimal end-to-end UTM tracking setup removes a common source of false mismatches before they reach your conversion reports.

Next step

Open your conversion actions list now and check three things in this order: status, primary or secondary, and the window. Write down what you find before you change anything. If you want the audit and reporting side handled alongside the tracking check, the tracking and reporting features page shows what the platform covers.