Paid search conversion tracking should be tested like any other production system. This workflow gives you a repeatable conversion tracking QA checklist for verifying tags, URL parameters, consent signals, lead data, phone calls, values, deduplication, offline imports, and platform reconciliation before launch and after every meaningful change.
Overview
A campaign can appear healthy while its conversion data is incomplete or misleading. A landing page redesign may remove a form event, a tag deployment may fire twice, a redirect may strip UTM parameters, or a consent setting may prevent signals from being recorded as expected. These failures affect bidding, budget decisions, keyword evaluation, and ROI reporting.
A useful PPC conversion tracking test separates the process into four questions:
- Did the user action occur? Confirm that the form submission, purchase, call, download, or other intended action can be completed.
- Did the tracking mechanism fire? Inspect the relevant browser event, tag, data layer value, or server-side request.
- Was the conversion assigned correctly? Check the campaign, keyword, click identifier, UTM parameters, value, currency, and timestamp where applicable.
- Did the destination platform receive and report it? Compare the ad platform, analytics property, CRM, call system, and backend records without assuming they will match exactly.
Run this workflow before a new campaign, landing-page release, analytics migration, tag change, consent-management update, form redesign, or offline conversion import. For broader account reviews, pair it with the Google Ads Account Audit Checklist.
What to track
1. Conversion definitions and ownership
Start with a plain-language inventory of every conversion action. Record the action name, business purpose, source system, primary or secondary reporting role, value method, counting rule, and owner. Examples include completed purchases, qualified lead forms, phone calls, booked consultations, trial starts, and downloads.
Do not test only whether a tag fires. Test whether the event represents the business outcome you intend to optimize. A button click may be useful as a diagnostic micro-conversion, but it should not automatically stand in for a qualified lead or completed sale.
2. URLs, click identifiers, and UTMs
Use a controlled test click and inspect the landing URL before and after redirects. Check that expected parameters survive navigation and that the final page does not introduce malformed values. Depending on the platform and setup, this may include an ad click identifier and campaign parameters such as source, medium, campaign, term, and content.
Use consistent naming rather than creating a new convention for each campaign. The UTM Tracking and ROI Measurement guide can serve as the reference for naming rules and UTM QA. Verify that parameters are captured by analytics and, where appropriate, stored with the lead record.
3. Form submissions and lead data
Submit a test form with a recognizable test value. Confirm that the form validates correctly, the success state appears once, the conversion event fires once, and the lead reaches the intended CRM or inbox. If a thank-you page is used, refresh it and test direct access so you can determine whether the conversion is vulnerable to duplicate page views.
Check field mapping, lead status, timestamp, source data, click identifiers, and consent status. A lead appearing in the CRM without its campaign details is not a complete attribution test.
4. Purchases, values, and currencies
For transactions, compare the order identifier, revenue amount, tax or shipping treatment, currency, and product or plan details across the checkout system, analytics, and ad platform inputs. Test a known amount and document whether the platform receives gross revenue, net revenue, or another defined value. Never leave the value definition implicit when it affects automated bidding or ROI tracking.
5. Calls and offline conversions
For call conversions, test the expected path: ad interaction, landing-page call button, forwarding number if used, call record, and conversion qualification rule. Confirm what makes a call count, such as duration or status, rather than assuming every call is a valuable conversion.
For offline imports, use a test record or a controlled production-safe sample. Verify the click identifier or matching key, event time zone, conversion name, value, currency, import status, and rejection reason. Keep a record of which events were imported so a QA conversion cannot be mistaken for a real business result.
6. Consent and duplicate behavior
Test the relevant consent states allowed by your implementation: consent granted, consent declined, and—where applicable—no choice made. Record which tags or signals fire in each state and whether the result matches your approved setup. Avoid treating a browser test as legal advice; consent behavior should be reviewed against your organization’s requirements.
Run the same action twice and inspect the results. Look for duplicate tags, repeated data-layer pushes, duplicate CRM records, repeated thank-you-page conversions, and multiple imports of the same offline event. A unique transaction or lead identifier is often useful for detecting duplication, but the exact method depends on the implementation.
Reusable test-case matrix
| Test case | Expected result | Evidence to save |
|---|---|---|
| Ad click to landing page | Final URL retains required attribution data | Initial and final URLs |
| Form submission | One valid event and one lead record | Event log, CRM record, timestamp |
| Purchase | Correct order ID, value, and currency | Order record and platform diagnostic |
| Phone call | Call follows the defined qualification rule | Call log and conversion status |
| Consent declined | Behavior matches the approved consent design | Consent state and tag output |
| Repeat success-page visit | No unintended duplicate conversion | Event count and transaction ID |
| Offline import | Record matches and processes once | Import status and source record |
Cadence and checkpoints
Use a three-stage cadence so QA is part of campaign operations rather than a one-time launch task.
Before launch or deployment
Review the conversion inventory, confirm the intended destination URLs, inspect the tag or event configuration, and test each important path in a clean browser session. Include mobile and desktop where the experience differs. Capture screenshots, event details, and test IDs in a shared log.
Immediately after a change
Repeat the affected tests after publishing a new tag, changing a form, editing a checkout, modifying consent behavior, switching analytics configurations, importing campaigns, or changing conversion settings. Do not assume that an unrelated-looking change cannot affect attribution: redirects, templates, and shared tag rules often serve multiple campaigns.
Weekly or monthly monitoring
Review conversion volume, conversion rate, reported value, cost per conversion, lead quality, and the share of records with usable source data. Compare trends by campaign, device, landing page, and conversion action. A sudden change is a prompt for investigation, not automatic proof of improved or degraded performance.
Quarterly control review
Reconcile the conversion inventory with the live account. Remove obsolete actions from operational documentation, confirm that naming remains consistent, review access and ownership, and rerun representative tests. Update this QA checklist when platform interfaces, tag-management processes, consent implementation, or reporting definitions change.
How to interpret changes
Begin with the point where the data diverges. If the user action is missing, investigate the page or form. If the action occurs but the tag does not fire, inspect the trigger, data layer, consent state, and browser conditions. If the tag fires but the platform does not report the conversion, check the request, conversion identifier, click matching, processing status, and reporting delay. If the platform reports the conversion but the CRM does not, investigate lead routing and field mapping instead of changing campaign bids.
Common patterns provide useful starting points:
- Conversions fall immediately after a release: compare the release timestamp with tag, form, redirect, and consent changes.
- Conversions rise while leads do not: look for duplicate events, repeated success-page loads, or a broader trigger.
- Analytics and ad-platform totals differ: compare date range, attribution settings, time zone, conversion action, counting rules, and processing status before labeling the data broken.
- Campaigns receive clicks but no attributed leads: test parameter persistence, click-ID capture, form submission routing, and landing-page conversion events.
- Revenue is present but ROI looks wrong: verify currency, value definition, order deduplication, refunds, and the date used for attribution.
Set a documented reconciliation tolerance appropriate to your reporting use case rather than expecting every system to produce identical totals. Escalate material discrepancies with evidence: test timestamp, browser or device, URL, event name, transaction or lead ID, consent state, and screenshots or diagnostic output.
When a discrepancy affects spend or bidding, pause automated changes until the measurement issue is understood. A keyword optimizer or bid optimization tool can act on incorrect conversion signals just as efficiently as it acts on correct ones. For recurring account checks, the Cross-Platform Ad Anomaly Monitoring Playbook offers a complementary approach to alerts and investigation.
When to revisit
Use this workflow before every campaign launch and material tracking change, then schedule a monthly review for active accounts and a broader quarterly audit. Revisit immediately when conversion volume changes unexpectedly, source data disappears, lead quality shifts, revenue no longer reconciles, a platform migration occurs, or a consent and privacy implementation is updated.
Keep a versioned QA log with five fields: date, change tested, scenarios run, evidence collected, and result or follow-up owner. This turns conversion tracking QA from an informal spot check into a repeatable control. At the next review, compare the current test matrix with the previous version and add any failure mode discovered in production.
Practical next step: copy the test-case matrix into your campaign operations document, assign an owner for each conversion action, and run one controlled test for every high-value path before making new bid or budget decisions. Then connect the findings to your paid search reporting process using the Paid Search Reporting Dashboard Template.
