Google Ads conversion tracking not working: what should you check first?
If Google Ads conversion tracking is not working, check one website enquiry in order: action status, real form test, then campaign goal and reporting column. Our reading: a tag firing in a test does not show that a real ad click will get credit; fix the first evidenced failure before changing bids.
On this page9 sections
What exactly has stopped working?
Choose one website enquiry that should be measured, then identify its conversion action and source. Our suggested starting question is: “Did this completed form fail to record, or are we looking at a different action, goal or date?”
A campaign with no enquiries is not automatically a tracking failure. Conversely, a healthy-looking account total can hide the loss of one important form route. For no impressions, use the delivery diagnosis; for qualified outcomes uploaded later, use the offline-conversion guide.
- Define the outcome. Describe the completed enquiry the business expects, rather than assuming a button click is the same thing.
- Identify the route. Record the action name, source and actual form destination. If the source is a GA4 import, investigate that route instead of installing another website tag as a blind fallback.
- Write the symptom. Distinguish no recorded activity, a sudden stop, missing campaign credit and a number lower than the business’s own submissions.
- Save the basis. Record the account, selected dates, conversion column and observed time. Keep customer details out of the diagnostic note.
This sequence is our investigation method. It does not assume that every completed website enquiry came from an ad, or that every form submission should be used for bidding. For a phone-call route, use the separate call tracking guide, which has its own acceptance test.
What does the conversion status actually establish?
Read the conversion action’s status and its explanation before editing the tag. As of 19 Oct 2026, Google’s troubleshooting page places the action table in Summary within Goals, with a Status column.
Some status details are gradually rolling out and may not be visible in your account yet
| Google’s overview meaning | Our next check | |
|---|---|---|
| Active | The action is working as intended. | Check whether it is the action and reporting scope you expected; this label does not identify every business enquiry. |
| Needs attention | There are action items, but recording has not stopped entirely. | Read the particular alert before treating this as complete failure. |
| Misconfigured | Conversions have stopped recording entirely because of a setup error or a broken tag. | Follow the identified setup issue and capture a test result. |
| Awaiting conversions | No conversions recorded in the last 7 days. Google names recent creation under 48 hours, paused campaigns and low traffic as typical explanations. | Establish whether an expected enquiry occurred; absence of activity is not proof of a missing tag. |
| Removed | The action was manually deleted or archived and no longer tracks data. | Check whether this was intentional and which action now measures the outcome. |
Google says to select the status to see alerts and use View details or View all issues for more information. Its shorter troubleshooting page also uses tag labels such as Inactive and Unverified; these are not a reason to invent a one-to-one mapping to the overview table.
Our reading: save the exact message shown for the affected action. A label copied from an older checklist is weaker evidence than the explanation attached to the action you are investigating.
How do you test the form without guessing the cause?
Test the same journey a visitor must complete, and record the form outcome beside the tag evidence. Google’s troubleshooting instructions separate using Google Tag Assistant to complete a test conversion from checking that the tag is correctly placed and fires on the intended page. They also direct Tag Manager users to check the Conversion Linker setup.
Our suggestion is to ask the site owner to test the real landing page, actual form and completion state. A successful form message and an observed tag event answer different questions; record each rather than replacing them with “tracking checked”.
| Evidence from the test | Our interpretation and owner | |
|---|---|---|
| Form never completes | No accepted enquiry reaches the completion state. | Give the reproducible form failure to the website owner; a campaign edit cannot repair that route. |
| Form completes; expected tag evidence absent | The business journey succeeds, but the expected measurement is not observed. | Give the actual destination, completion state and tag evidence to the tag/site owner. Do not diagnose a consent issue from this symptom alone. |
| Expected tag observed | The test reached the measurement step. | Continue to the action, goal and reporting checks. |
For Google Tag Manager users, Google describes one case: if an action shows as Misconfigured and you cannot find a Conversion Label to enter into Tag Manager, check whether the action was created automatically.
Google says auto-created goals cannot be mapped to Tag Manager conversion tags, and its remedy is to create a manual conversion action to get the Conversion ID and Conversion Label. Our suggestion: apply that remedy to the documented case rather than creating replacement actions for every warning.
A test is not a reason to bypass the site’s consent choices or include personal lead information in screenshots. If a specialist needs a separate consent or cross-domain investigation, hand over the evidence and leave the cause unresolved until it is established.
Why can a tag fire while the conversion column stays empty?
Recording and reporting have separate conditions: the action, campaign goal, selected column, attribution date and processing time all matter. Check those conditions before assuming a second tag is needed.
Under Google’s primary/secondary guidance, primary actions are reported in Conversions and are used for bidding as long as the standard goal they are part of is used for bidding.
Secondary actions are used for reporting in All conversions but not for bidding, even if their goal is used for bidding. The one exception is a secondary action that is part of a custom goal, which is used for bidding.
The short troubleshooting page also says, in Goals → Conversions → Summary, to ensure the actions are set to Primary. The primary/secondary page instead distinguishes key actions from observation-only actions, and says some actions cannot be set as either primary or secondary. Our suggestion: choose the business-enquiry action deliberately, rather than treating every supporting event as a bidding outcome.
The troubleshooting page says Google Ads attributes conversions to the date of the click, which may cause discrepancies with other platforms. Our reading: a comparison against the form’s completion date can answer a different question. The reporting guide explains the columns and date basis in more detail. To separate a report difference from a tag fault, see the Google Ads vs GA4 conversions worksheet.
Google’s freshness descriptions do not give one deadline
The freshness page says: Data freshness for conversions varies significantly depending on your attribution model and conversion type.
The short troubleshooting page says reporting can take 3–24 hours. The data-freshness comparison table gives 3 hours for Google Ads tracking under last-click and 15 hours under other attribution models; Analytics imports are 12 / 24 hours.
The same freshness page also includes last-click conversions in a table labelled hourly refresh. These descriptions differ. Our reading: use the documented route and model as context, save when the check was made, and do not promise that every conversion appears by one fixed deadline.
Our suggestion: keep the current state as “tag observed; report attribution not yet established” until the reporting evidence supports a stronger conclusion. That is more useful than marking the whole investigation fixed.
What does a useful tracking incident record look like?
A useful record identifies the missing proof and the person who can obtain it. This invented repair-service example shows the difference between an observed failure and a proposed explanation.
| Invented observation | Decision or missing proof | |
|---|---|---|
| Expected outcome | A visitor submits a repair enquiry on the website. | The business treats the completed submission as the action to inspect. |
| After a website change | The form now completes inside the page; the tester sees its confirmation message. | The form works in this test; the tag evidence remains a separate observation. |
| Tag evidence | The expected conversion event is absent from the captured test. | Ask the tag owner to check the actual completion state and current implementation. The website change is a lead to investigate, not proven causation. |
| Reporting evidence | The chosen action has no recent recorded activity. | Keep the action/source and date basis with the case; a status alone does not isolate the technical cause. |
| Acceptance condition | A repeat test observes the intended event; the responsible operator then checks reporting on its documented basis. | Keep test success and attributed advertising results separate in the sign-off. |
The example contains no client data and promises no recovery rate. It does not reconstruct an earlier lead or measure every production visitor.
Our suggestion is to record other relevant releases or account edits alongside the incident. Without that context, “it stopped after the redesign” can become an unsupported causal explanation.
What can Variafy help inspect?
Variafy offers read-only Google Ads access after your beta invitation. It reads conversion-tracking settings, enabled conversion actions, their counting and attribution settings, and available conversion totals and last recorded days.
Those readings can help identify the action and missing evidence for this investigation. They do not run the website form in a browser, establish every individual visitor’s path, repair tags or change conversion settings. An unavailable optional reading remains missing evidence.
Our suggested investigation request is: “Identify the intended account and website action. State its available settings, observed conversion period and missing readings. Separate a setup warning from an empty reporting period, and draft the next check without changing anything.”
This is a reading workflow. Variafy does not change Google Ads.
What should you send the website owner next?
Send a reproducible failure and an acceptance check, not a request to “fix Google Ads”. Keep the diagnosis open when the evidence only narrows the problem.
Our suggested handover
“Expected outcome: [completed website enquiry]. Action/source: [name and source]. Tested route and time: [page, form and completion state]. Observed: [form result and tag evidence]. Google action message: [exact message]. Still unknown: [reporting or technical evidence]. Please check [specific missing step], then repeat the same route and return the evidence.”
Once that check passes, confirm the intended goal and reporting basis before making campaign decisions. If recording is established but enquiries are unsuitable, move to the search-term decision guide or the wider account audit; more tag changes do not answer lead quality. If you suspect poor-quality traffic, the invalid clicks guide covers a separate, bounded check.
How we checked
Checked on 19 Oct 2026 against 4 public Google Ads Help pages: Conversion Tracking: Troubleshooting conversion tracking status; Troubleshoot your conversion tracking status; About primary and secondary conversion actions; About data freshness. Status availability, goal exceptions and differing freshness descriptions retain their stated scope.
The investigation order, incident record and handover are our suggestions. The service example is invented and uses no client data. Variafy is our product.
Questions
Why is Google Ads conversion tracking not working?
Identify one expected website action, read its exact status and test the actual form journey. If the event is observed, check the action, campaign goal, reporting column and date basis before assuming the tag is broken.
Does Awaiting conversions mean my tag is broken?
No. Google describes no recorded conversions in the last 7 days and names recent creation, paused campaigns and low traffic as typical explanations. Test an expected conversion and read the action’s own explanation.
Why does Tag Assistant show a conversion but Google Ads shows zero?
Our reading: a test can establish tag firing without proving credit to an advertising interaction. Check the intended action and campaign goal, the selected conversion column, click-date reporting and route-specific processing time.
Should every conversion action be primary?
Not necessarily. Google’s troubleshooting page says to ensure actions are Primary. Its primary/secondary guide reports primary actions in Conversions, using them for bidding if their standard goal is used. Secondary actions report in All conversions, with a custom-goal bidding exception. Our suggestion: make the business-enquiry action primary and supporting actions secondary, checking custom goals.
How long should I wait for Google Ads conversions to appear?
Google’s pages describe different timing by source and attribution model, and their freshness descriptions are not fully aligned. Record the route and check time rather than treating a general range as a guaranteed deadline or evidence of a broken tag.