Nothing Is Being Recorded
Start with the base tag or container, consent behavior and whether the expected event fires at all.
Check collection
Conversion measurement debugger
How to Fix Conversion Tracking When the Numbers Look Wrong
When conversions are missing, duplicated or inconsistent, do not rebuild every tag. Identify where the chain breaks: the user action, event, tag, GA4, Google Ads conversion action, value, consent, attribution or downstream lead outcome.
A form can submit successfully while no event is sent. GA4 can receive the event while Google Ads is not configured to use it. A purchase count can be correct while revenue is wrong. The fix depends on which layer failed.
Start with the symptom
What Conversion Tracking Problem Are You Actually Seeing?
Different symptoms point to different layers. Route the problem first so a missing GA4 event is not treated like a bidding problem or a revenue error like duplicate counting.
GA4 Records It but Google Ads Does Not
Check the Ads conversion action, source, linkage and the conversion goal the campaign actually uses.
Check Ads setupConversions Are Counted More Than Once
Look for duplicate tags, repeated triggers, reloads, duplicate events or overlapping conversion actions.
Check duplicationCount Looks Right but Revenue Is Wrong
Check dynamic value, currency and transaction ID rather than rebuilding the event itself.
Check valueConversions Exist but Google Ads Optimizes Poorly
Audit Primary and Secondary actions, campaign goals and whether the selected outcome is commercially meaningful.
Check bidding goalDefine the business outcome
Decide What Should Count as a Conversion Before You Fix the Tags
Technical implementation cannot rescue a bad conversion definition. Decide which actions represent business value before you debug event firing or campaign goals.
Separate Business Outcomes from Diagnostic Events
A purchase, qualified enquiry, booking or confirmed appointment may be the business outcome. A form start, CTA click, scroll or page view can help diagnose behaviour, but that does not mean it should be used as the bidding outcome.
For Lead Generation, Do Not Stop at Form Submission
A raw form submission, qualified lead and converted lead are different outcomes. If lead quality matters, measurement should eventually connect the website action with CRM or sales outcomes where the systems allow it.
For eCommerce, Purchase Must Carry the Right Transaction Data
The purchase event needs more than a count. Validate transaction ID, value and currency so deduplication and revenue reporting represent the actual order.
Measurement chain
Trace the Signal From Business Action to Bidding Goal
A conversion can fail at several points without the earlier layers being broken. Trace the path in order so you know exactly where the signal stops or changes.
Business OutcomePurchase, qualified lead, booking
User ActionSubmit, pay, call, book
Event / dataLayerEvent name and parameters
Google Tag / GTMTrigger and firing
GA4 / Google AdsReceipt and conversion action
Goal / Value / ConsentPrimary, value, deduplication
Bidding / CRM OutcomeOptimization and downstream quality
Primary debugging asset
Use This Conversion Tracking Debugging Tree
Follow the chain in order. If one layer fails, fix it before judging the next layer. If it works, move forward.
1
Did the real business action happen?
No: fix the user journey or business process.
Yes: check whether the expected browser or dataLayer event occurred.
2
Did the expected event occur?
No: inspect trigger logic, selectors, form callbacks, SPA behavior or event creation.
Yes: verify the tag fired exactly once.
3
Did GTM or the Google tag fire?
No: inspect trigger conditions, consent and tag configuration.
Yes: verify the event parameters.
4
Did GA4 receive the event?
No: debug collection with Realtime, DebugView and Tag Assistant.
Yes: move to the Google Ads conversion action.
5
Did Google Ads receive and use the intended conversion?
No: check source, status, linkage and goal configuration.
Yes: inspect Primary/Secondary status, campaign goals, duplication and value.
6
Is the action Primary or Secondary as intended?
Wrong: correct action optimization and campaign goal relationship.
Correct: check duplicates, value, currency and transaction ID.
7
Are consent or cross-domain journeys affecting collection or attribution?
Yes: validate consent defaults/updates, firing order and cross-domain continuity.
No: decide whether a downstream qualified or converted lead should be imported.
Step 1
Confirm the Base Measurement Setup Is Working
If the base Google tag or GTM container is missing, duplicated or loading in the wrong environment, debugging a purchase or lead event is premature. First make sure the measurement infrastructure is present on the pages where the conversion journey happens.
Use Tag Assistant before editing tags. It can show whether the container is connected, what fires, and whether consent settings affect expected behavior.
CheckGoogle tag or GTM container
Confirm the correct container or tag, the correct production environment and page coverage across landing, form, checkout or booking steps.
EvidenceTag Assistant session
Confirm connection, firing sequence and obvious implementation errors before changing configuration blindly.
FixConsent-related blocking
If the expected tag does not load or fire because of consent behavior, correct the consent implementation rather than bypassing it.
Step 2
Confirm the Business Action Fires the Correct Event
A successful user action and a successful measurement event are separate things. Test the real action. Submit the form, complete the booking or place a controlled order. Do not infer event health from page load alone.
If the action succeeds but no expected event appears, inspect the trigger, custom event, selector, thank-you logic or SPA state change that is supposed to create the signal.
CheckReal user action
Perform the actual conversion path and confirm the site's success state.
EvidenceTrigger and event sequence
Look for wrong selectors, callbacks, repeated custom events, reload behavior or SPA transitions.
FixEvent parameters
Send the names and business-relevant parameters the outcome requires. Keep the implementation specific to the site rather than copying a generic event recipe.
Step 3
Check Whether GA4 Receives the Event Correctly
Once the event is generated and the tag fires, verify collection in GA4. Realtime is useful for a quick sanity check. DebugView is better for inspecting event sequence and parameters during a controlled test.
Use Realtime for a Quick Sanity Check
Confirm that the controlled visit and event reach the correct property. Realtime can show whether collection is alive, but it should not be the only debugging layer.
Use DebugView for Event-Level Testing
Inspect the sequence of events and confirm the expected event name and parameters appear when the action occurs.
Check Whether the Event Is the Intended Key Outcome
GA4 receipt does not automatically mean Google Ads is optimizing toward that event. Treat analytics collection and advertising goal configuration as separate layers.
Step 4
Check Whether Google Ads Is Receiving and Using the Right Conversion
A conversion action can exist in Google Ads without being the action a campaign actually bids toward. Validate status, Primary or Secondary configuration, and the campaign's selected conversion goals separately.
Check the Conversion Action Status
Review whether the action is recording and whether diagnostics show an unverified, inactive or needs-attention state. Use the current Google Ads diagnostics because exact status wording can change.
Understand Primary vs Secondary
Primary actions can be used for bidding when the associated goal is selected. Secondary actions are generally observation-only and appear in All conversions, with the custom-goal exception.
Do not make form starts, page views or other easy micro-actions Primary merely to create more conversion volume. If the campaign optimizes toward an easy action that is weakly connected to revenue, bidding can become efficient at the wrong outcome.
Also check the campaign's actual goals. Account-level defaults and conversion-action configuration do not guarantee that every campaign is optimizing toward the same business outcome.
Duplicate measurement
Why Are Conversions Being Counted Twice?
Duplicate conversions usually come from repeated measurement, not from the business receiving two genuine outcomes. Identify whether the duplication happens at implementation, trigger, event or conversion-action level.
Check for Duplicate Tag Implementations
A common pattern is a hardcoded tag plus the same measurement through GTM, multiple containers, or the same conversion tag installed in more than one place.
Check Whether the Trigger Fires More Than Once
Form callbacks, page reloads, SPA state changes or duplicate custom events can fire the same conversion multiple times from one business action.
Use Transaction or Order IDs Where Appropriate
For purchases, a unique transaction ID helps deduplicate repeated measurement of the same order. Stable identifiers also matter when offline lead systems need reliable deduplication.
Also check whether multiple Google Ads conversion actions measure the same business outcome. Having several actions is not automatically wrong, but bidding or reporting can be distorted if the same purchase or lead is counted through overlapping actions.
Revenue accuracy
Why Is Conversion Value or Revenue Wrong?
A correct conversion count does not prove the economic data is correct. Validate value, currency and transaction ID separately from event firing.
Validate Value and Currency
Check that dynamic order value is passed from the real purchase or booking data and that the currency code is correct. Static fallback values can make every order look identical even when revenue differs.
Validate the Transaction ID
Each completed order should use the appropriate unique identifier. Reusing or omitting it can create duplicate or unreliable purchase measurement.
Compare Platforms with the Source of Truth
Use the store, CRM or payment system as the business reference. GA4 and Google Ads may legitimately differ because of attribution and reporting logic, so compare patterns rather than demanding identical totals.
Under-measurement
Why Might Google Ads Record Fewer Conversions Than Your Business Receives?
Fewer observed conversions do not automatically mean the business received fewer outcomes. Collection can be affected by tag behavior, consent, identity matching, offline conversion timing and whether the final business outcome occurs outside the website.
Check Tag and Consent Behavior
If consent implementation changes when Google tags load or what they send, observed conversion totals can change even while the business continues receiving leads or purchases.
Evaluate Enhanced Conversions Where Appropriate
First-party data can improve measurement and matching where the implementation and business case fit. Do not treat it as a guaranteed way to recover a fixed percentage of missing conversions.
Check Whether the Outcome Happens Offline or Later
A raw website lead may convert to a qualified lead or customer days later in a CRM or sales system. That downstream outcome requires a different measurement path from the original form submission.
Multi-domain journeys
Does the Conversion Journey Cross More Than One Domain?
A visitor may start on the main website and finish on a booking engine, checkout, payment domain or application portal. Map that journey before assuming the conversion itself is missing.
Broken cross-domain measurement can fragment the session or change attribution even when the final conversion event still fires.
CheckMap every domain
List the website, checkout, booking, payment or application domain used before the final conversion.
EvidenceCross-domain configuration
Verify GA4 and Google tag or GTM continuity across the relevant domains.
FixReferral and session breaks
Determine whether the problem is missing measurement or misattribution caused by a broken journey.
Consent-related measurement
Is Consent Mode Changing What You Can Measure?
Consent Mode can change tag behavior and observed data. Treat it as a measurement layer, not as a reason to bypass user choices. This section explains measurement behavior only and is not legal advice.
Basic and Advanced Consent Mode Behave Differently
With Basic Consent Mode, Google tags are blocked until the user interacts with the consent banner and grants consent. With Advanced Consent Mode, tags can load with denied defaults and send cookieless signals when consent is denied, then adjust when consent changes.
Verify Consent Signals and Firing Order
Use Tag Assistant to inspect consent defaults and updates, CMP integration and the relevant signals such as ad_storage, ad_user_data, ad_personalization and analytics_storage. A wrong order can create measurement gaps even when the banner appears to work.
Do Not Treat Modeled Conversions as a Guaranteed Percentage
Modeling depends on eligibility, thresholds and implementation. Do not promise that Consent Mode or modeling will restore a fixed percentage of conversions.
Lead quality measurement
Are You Tracking Form Submissions or Real Qualified Leads?
If the business optimizes for every form submission equally, the platform cannot distinguish a low-intent enquiry from a genuine sales opportunity unless better downstream signals are available.
Separate lead, qualified lead and converted lead
Define the stages that matter to the business. A raw lead is the website action. A qualified lead is one the business considers commercially relevant. A converted lead is the downstream customer or completed outcome.
Use enhanced conversions for leads when the business case fits
Google can use hashed first-party data and downstream outcomes to improve matching. Keep the website lead and CRM outcome connected with a stable process rather than treating the form submission as the final signal.
As of June 15, 2026, Google documents migration of offline conversions import and enhanced-conversions-for-leads uploads toward the Data Manager API. Build new workflows around the current Data Manager direction instead of relying only on legacy upload instructions.
Bidding quality
Is Google Ads Optimizing Toward the Wrong Conversion?
Tracking can be technically correct while bidding still learns from the wrong business outcome. Audit the actions used for optimization, not just whether they fire.
Audit Which Actions Are Primary
Primary actions should represent outcomes you are willing to let campaigns optimize toward when the associated goal is selected. Keep diagnostic actions separate unless there is a deliberate reason to bid toward them.
Check Campaign-Specific Goals and Custom Goals
Campaign configuration can differ from account defaults. A Secondary action included in a custom goal can still be used for bidding, so inspect the actual goal logic instead of relying on the label alone.
Do Not Inflate Conversion Volume with Easy Actions
Page views, CTA clicks and form starts may be useful diagnostics. Treating them as final outcomes can push Smart Bidding toward users who are likely to perform easy actions rather than become customers.
If measurement is trustworthy but Google Ads performance is still weak, move to diagnose why Google Ads are not working. At that point, the next investigation belongs to campaign delivery, traffic quality, landing-page performance and business economics rather than tracking.
End-to-end QA
Validate Conversion Tracking End to End Before You Trust the Data
A successful test should prove the whole chain, not just one platform screen. Run one controlled visit and follow the outcome from the site action through collection, reporting and bidding configuration.
- Use a controlled test visit.
- Perform the actual conversion action.
- Confirm the website success state.
- Confirm the expected event or trigger.
- Confirm the tag fires once.
- Confirm the event parameters.
- Confirm GA4 receives the event.
- Confirm Google Ads receives or recognizes the intended conversion action.
- Confirm Primary or Secondary status and campaign goal configuration.
- For purchases, confirm transaction ID, value and currency.
- For cross-domain journeys, confirm continuity.
- For leads, confirm CRM delivery and downstream qualification where implemented.
After measurement is verified
If Tracking Is Correct, Diagnose the Next Business Problem
Once the measurement chain is trustworthy, stop treating every weak result as a tracking issue. Move into the problem that the verified data is actually showing.
Traffic Arrives but Qualified Leads Are Weak
If events are correct but the website still produces too few or poor-quality enquiries, diagnose why your website is not generating leads.
Google Ads Is Still Underperforming
If conversion actions are reliable but campaigns remain weak, move to Google Ads performance diagnosis.
The Funnel Is Measured but Conversion Is Weak
If the traffic is relevant and tracking is trustworthy, move into conversion-rate improvement instead of changing tags again.
If the apparent problem is actually a fall in acquisition rather than measurement, use the website traffic drop diagnostic framework.
Fast lookup
Conversion Tracking Diagnostic Matrix
Use this table to identify the first layer to inspect. It is a routing tool, not proof of causation.
| What You See | Likely Layer to Check First | Best First Check | Next Action |
|---|---|---|---|
| GA4 shows no site activity | Base measurement / consent | Tag Assistant | Verify tag or container and consent behavior |
| Pageviews work but lead event does not | Event / trigger | GTM Preview or event debugger | Test the actual conversion action |
| GA4 has the event but Ads does not | Ads conversion action / linkage / goal | Google Ads conversion diagnostics | Verify action source and configuration |
| Conversions are too high | Duplicate firing / duplicate actions | Tag Assistant + event sequence | Remove duplicate measurement |
| Purchase count is high after reloads | Deduplication | Transaction ID | Send a unique transaction ID |
| Count is correct but revenue is wrong | Value / currency | Event parameters | Fix dynamic value or currency |
| Tracking changed after CMP launch | Consent | Tag Assistant consent verification | Check default and update states plus firing order |
| Journey breaks at booking or checkout domain | Cross-domain | Session / referral path | Configure cross-domain measurement |
| Many conversions but poor bidding outcomes | Goal configuration | Primary / Secondary + campaign goals | Optimize toward meaningful business outcomes |
| Form leads track but quality is unknown | Offline outcome measurement | CRM | Measure qualified or converted leads |
Avoid destructive debugging
What Not to Do When Conversion Tracking Looks Wrong
Bad troubleshooting adds more variables. Keep the system stable while you identify the failing layer.
Do not rebuild every tag before identifying the broken layer.
Do not make every micro-event a Primary conversion.
Do not optimize campaigns using data you have not validated.
Do not remove or bypass consent controls just to increase measured conversions.
Do not count thank-you page reloads as new purchases.
Do not compare GA4 and Google Ads as if they must always report identical numbers.
Do not change attribution settings to hide a collection problem.
Do not stop at form submissions if lead quality drives the business.
Residual questions
Conversion Tracking Troubleshooting Questions
Why does GA4 show conversions but Google Ads does not?
GA4 receipt proves the analytics event arrived, not that Google Ads has a usable conversion action for it. Check the Ads conversion action source, linking or import setup, status, Primary or Secondary configuration and the goal the campaign actually uses.
Why is Google Ads counting conversions twice?
Check for hardcoded plus GTM implementations, repeated triggers, thank-you page reloads, duplicate custom events and multiple conversion actions measuring the same outcome. For purchases, verify that the same order is not sent more than once without proper deduplication.
Should a form start be a Primary conversion?
Usually only if you intentionally want bidding to optimize toward form starts. For most lead-generation setups, a form start is better treated as a diagnostic action while the meaningful lead or qualified outcome is used for optimization.
Why do GA4 and Google Ads conversion numbers differ?
The platforms use different attribution and reporting logic, so identical totals are not required. First confirm that the same business action is measured correctly in both systems, then compare timing, attribution settings and reporting scope rather than forcing the numbers to match.
How do I test whether a conversion tag is firing?
Use a controlled test and Tag Assistant or GTM Preview. Perform the actual action, confirm the expected event occurs, verify the tag fires once, then confirm GA4 and Google Ads receive the intended signal.
Do I need a transaction ID for purchase tracking?
For ecommerce purchase measurement, a unique transaction or order ID is an important part of reliable deduplication. It also gives you a stable reference when comparing analytics and advertising data with the store's order system.
Can Consent Mode reduce observed conversions?
Yes. Consent implementation changes what tags can collect and send. Basic and Advanced Consent Mode behave differently, and modeled results depend on eligibility and thresholds. Validate consent defaults, updates and firing order before concluding the business stopped receiving conversions.
Should I track qualified leads instead of form submissions?
If lead quality drives the business, measure both. Keep the raw form submission for funnel visibility, then connect qualified or converted lead outcomes through the CRM or offline measurement process where practical so bidding can use a more meaningful business signal.
Commercial handoff after diagnosis
Still Cannot Find Where Your Conversion Tracking Is Broken?
If the user action, event, tag, GA4, Google Ads goal, duplicate logic, value, consent and downstream outcome have all been checked but the measurement chain is still unclear, the next useful step is a structured tracking diagnosis rather than another tag rebuild.