Google Ads conversion tracking is only useful when the numbers can be trusted. This QA checklist gives you a repeatable way to verify lead, purchase, and revenue tracking before launch, then monitor the setup monthly for broken tags, duplicate conversions, missing values, attribution changes, and consent-related gaps.
Overview
A campaign can receive clicks and still produce unreliable reporting if a conversion action is misconfigured. Common causes include a tag placed on the wrong page, an event firing more than once, an imported analytics event configured differently from the original event, or a purchase conversion that records no transaction value.
A useful conversion tracking QA process checks four connected layers:
- Definition: the action represents a meaningful business outcome.
- Collection: the Google Ads tag, Google tag, analytics event, or integration receives the signal.
- Configuration: counting, value, attribution, and primary or secondary status match the reporting objective.
- Reconciliation: platform data is compared with the website, CRM, payment system, or analytics records.
Complete this checklist before a new campaign goes live, after a website or checkout release, and during a recurring PPC conversion tracking audit. For related measurement work, see the UTM Builder and Campaign Tracking Guide.
What to track
1. Conversion action definitions
Begin with a conversion inventory. Record each action's name, source, trigger, business purpose, counting method, value method, and destination system. Separate actions that look similar but have different value, such as a completed lead form, a phone call, a quote request, a completed purchase, and an account registration.
Confirm that every conversion action is still needed. Retired forms, old thank-you pages, test events, and duplicate imports can inflate totals. Decide which actions should guide bidding and optimization and which should remain available for observation. The correct choice depends on the account's measurement plan, so document the reasoning rather than relying on default settings.
2. Tag installation and firing
For website conversions, inspect the Google tag and event-specific implementation or the tag-management configuration. Test the exact user journey, not only the destination URL. A lead conversion may depend on a form submission, an embedded form provider, a redirect, or an AJAX response. A purchase conversion may depend on a confirmation page or a transaction event generated by the checkout system.
Use a browser debugging or tag-preview method to confirm that the expected event fires once, after the intended action, and with the correct parameters. Check different paths where relevant: mobile and desktop, logged-in and guest checkout, validation errors, payment failure, and a successful retry. A page-view conversion should not fire simply because someone visits a page that is also accessible through navigation.
3. Enhanced conversion inputs
If enhanced conversions are part of the setup, verify that the implementation can identify the intended first-party information after the user submits a qualifying action. Review the data flow, consent behavior, matching configuration, and diagnostic feedback available in the platform. Do not treat an enabled setting as proof that usable signals are being received. Record whether enhanced conversion diagnostics show activity, warnings, or insufficient data.
4. GA4 and imported events
When a GA4 event is imported into Google Ads, validate both systems. First confirm that the event fires in the analytics property with the expected event name and parameters. Then check that the corresponding Google Ads conversion action uses the intended counting and value settings. An event can be present in GA4 but absent from Google Ads, or it can be imported alongside a separately installed Google Ads tag for the same action.
Compare event definitions carefully. “Generate lead,” “form submit,” and “purchase” should not be assumed to mean the same thing across tools. Imported analytics conversions may also have different reporting timing or attribution behavior than website-tag conversions. Record the source of truth for each action in your conversion inventory.
5. Values, currency, and transaction IDs
For purchases and revenue-generating leads, test whether the value is present, numeric, and assigned to the right transaction. Confirm currency handling when customers can pay in more than one currency. If the setup sends a transaction ID, check that the same order cannot be counted repeatedly after a refresh, back button, or return to the confirmation page.
For leads, document whether the value is fixed, assigned by lead type, or added later from CRM data. Avoid recording a nominal value merely to fill a field if it does not support a meaningful optimization decision.
6. Consent and measurement gaps
Test the experience under the consent states that your site supports. Note what happens before consent, after consent, and when a visitor declines optional measurement. Depending on the implementation and visitor context, some conversions may not be directly observable or may be modeled later. Treat a lower platform total than the internal system as a reconciliation question, not automatically as a tagging failure.
Platform-by-platform checklist
| Platform or source | Verify | Record |
|---|---|---|
| Google Ads website tag | Conversion action, trigger, firing count, value, currency, and status | Test URL or event, timestamp, expected result |
| GA4 import | Source event, event parameters, import mapping, and duplicate action risk | Analytics event name and Ads conversion name |
| Google Tag Manager | Tag trigger, variables, consent conditions, and version published | Container version and preview result |
| CRM or payment system | Lead or order created, value, currency, and unique ID | Internal record ID and test amount |
| Microsoft Ads or other ad platform | Separate tag, goal definition, revenue setting, and UET or equivalent event | Platform-specific test result |
Cadence and checkpoints
Before launch: review the conversion inventory, publish the intended tag version, run a controlled test, and confirm that the test appears in the relevant diagnostic view. For purchases, compare the order ID and value with the payment or commerce record. For leads, confirm that the submitted record reaches the intended inbox or CRM destination.
Within the first reporting cycle: check whether the conversion action moves from a setup or diagnostic state into normal activity. Compare clicks, sessions, leads, orders, and revenue across systems using the same date range and timezone. Early data may not align perfectly because systems can process or attribute events differently, but unexplained zeroes and sharp discrepancies deserve investigation.
Monthly: review conversion volume, conversion rate, cost per conversion, conversion value, and value per cost alongside tag diagnostics. Check for newly created actions, changes to primary or secondary status, altered values, and campaign exclusions. Revisit the PPC Budget and Bid Management Workflow before making bid or budget decisions from the data.
Quarterly or after a major release: repeat a full journey test. Include landing-page changes, form-provider changes, checkout updates, consent-management updates, domain changes, analytics property changes, and tag-manager publishing history. A small technical change can affect measurement without changing the visible conversion page.
How to interpret changes
Do not diagnose a tracking problem from one metric alone. Start by locating the first point where the expected journey diverges.
| Observation | Likely checks | Next action |
|---|---|---|
| Clicks exist, but no conversions appear | Trigger conditions, consent state, event name, landing path, and reporting delay | Run a live test and inspect the tag or imported event |
| Conversions doubled after a release | Two tags, two imports, repeated confirmation-page firing, or duplicate triggers | Compare implementation sources and transaction IDs |
| Conversions exist, but value is zero | Missing value parameter, wrong variable, currency mismatch, or fixed-value setting | Test a transaction and document the intended value method |
| Ads conversions are lower than CRM leads | Consent, attribution, click identifiers, offline import timing, or lead-quality filters | Reconcile by date, source, status, and unique lead ID |
| GA4 and Google Ads totals differ | Different attribution, counting rules, timezones, event definitions, or import status | Compare configuration before changing tags |
Keep a test-results log rather than relying on memory. A practical entry includes: date and time, tester, device, browser, consent state, campaign or test source, conversion action, expected event, received event, value, currency, transaction or lead ID, diagnostic message, and remediation owner. Mark each result as passed, failed, or inconclusive. “Inconclusive” is useful when platform processing or modeled measurement prevents an immediate confirmation.
When to revisit
Revisit this checklist before every campaign launch and at least monthly while advertising is active. Run an additional review whenever conversion volume changes unexpectedly, revenue stops importing, a form or checkout is redesigned, a tag container is published, a consent experience changes, or a new analytics-to-Ads integration is added.
At the end of each review, save the test log, current conversion inventory, implementation version, and reconciliation notes. Compare the new results with the previous month instead of starting from a blank page. If a discrepancy remains, isolate one conversion path and one platform at a time; changing several tags or settings simultaneously makes the cause harder to identify.
Once tracking is verified, use the data carefully. Pair conversion quality with keyword and campaign analysis, maintain an appropriate negative keyword maintenance workflow, and document any adjustment that changes what the account counts as success. Reliable measurement is not a one-time installation task; it is a recurring control that protects every budget, bid, and optimization decision that follows.
