Conversion Tracking Audit: Verify the Outcome End to End

A conversion tracking audit tests whether a real business action is recorded once, with the correct definition and useful context, in every system that depends on it. It should follow the signal from the website or app through tag management, analytics, advertising and the CRM or commerce platform.

A tag firing is evidence of one technical step. It is not proof that the lead exists, the transaction is correct or the bidding system receives a suitable goal.

Quick answer: Inventory every conversion, document its business definition, complete controlled tests, inspect tags and network requests, reconcile analytics and advertising settings, verify CRM or order receipt, test duplicates and consent states, then prioritise fixes by their effect on optimisation and commercial reporting.

1. Inventory conversion actions across systems

Export or list conversion-like actions from:

• website or application code;
• Google Tag Manager;
• GA4 events and key events;
• Google Ads conversion actions;
• other advertising platforms;
• call-tracking tools;
• booking platforms;
• CRM workflows;
• ecommerce and payment systems;
• dashboards and data warehouses.

For each action, record name, ID, source, trigger, business definition, primary or secondary status, counting rule, value, currency, attribution, owner and last validation.

Duplicate concepts often hide behind different names. contact_submit, generate_lead and “Website lead” may all describe the same form. Conversely, one name such as “conversion” may combine purchases, phone clicks and newsletter signups.

Do not delete or edit anything during diagnosis. Preserve a baseline and change history.

2. Define what one conversion count means

Write a sentence for each action:

One generate_lead means the server accepted a completed consultation form and returned a unique lead ID.

Then answer:

• Can the same person create several valid actions?
• Should each count or only one?
• Can refresh or browser back repeat it?
• Does a click occur before validation?
• Does the CRM receive the same action?
• Is spam excluded later or at source?
• When is value known?
• Which time zone records it?

If stakeholders cannot agree on the sentence, the conversion is not ready to guide optimisation.

Separate diagnostic events, technical conversions and commercial outcomes. A form success can be technically valid but commercially unqualified. Both stages can be measured, but they should not be treated as interchangeable.

3. Run a controlled end-to-end test

Create a test matrix with unique timestamps and identifiers. Complete each real path:

• valid form submission;
• validation failure;
• duplicate click;
• page refresh;
• direct thank-you-page visit;
• mobile and desktop;
• consent choices;
• booking confirmation;
• connected and missed call where measurable;
• purchase, cancellation and refund;
• cross-domain handoff;
• CRM qualification update.

Observe each layer:

1. The user sees a success or error state.
2. The application or data layer exposes the correct event.
3. GTM or the tag fires once.
4. The network request contains permitted values.
5. GA4 or the destination receives the event.
6. Google Ads or another platform records the intended conversion.
7. The CRM or commerce system contains the underlying outcome.

Record evidence and discrepancies. A reusable verification record is more valuable than an undocumented “looks good.”

4. Check duplication, omission and inflation

Also check orphaned tags, paused but imported conversions, duplicated GA4 and Google Ads measurement, and global goals unintentionally applied to every campaign.

5. Audit settings and attribution context

A Google Ads conversion tracking audit must review more than tag firing. Check:

• primary versus secondary action optimisation;
• campaign-specific or account-default goals;
• counting method;
• conversion value and currency;
• attribution model;
• click and view-through windows where applicable;
• enhanced or offline conversion status;
• imported versus native source;
• inclusion in reporting columns;
• diagnostics and recent changes.

For GA4, check event naming, key-event status, custom definitions, data filters, unwanted referrals, cross-domain settings and property time zone.

Do not force platforms to match by changing settings blindly. Differences can be legitimate because platforms attribute and process events differently. Reconcile the base event and document each reporting lens.

6. Verify lead source and downstream quality

A lead tracking system needs stable source data and an identifier that connects the web submission with the CRM record. Preserve click IDs or campaign parameters where permitted, and store a lead ID that can support reconciliation.

Check:

• source and medium captured at the right scope;
• campaign and landing page;
• first-touch and latest-touch definitions;
• lead ID continuity;
• duplicate CRM records;
• qualification reason;
• sales stage and timestamps;
• customer value;
• consent and retention.

Lead source tracking should not overwrite the original source every time the visitor returns unless that is the defined model. Store separate fields for distinct attribution questions.

When returning offline outcomes to ad platforms, validate match rates, event timing, deduplication and goal settings under current requirements.

7. Prioritise fixes and establish monitoring

Prioritise:

• Critical: false primary conversions, missing purchases, personal data sent improperly, duplicate transaction value.
• High: broken lead IDs, wrong campaign goals, missing qualified outcomes, major consent-order errors.
• Medium: inconsistent naming, underused parameters, cross-domain attribution gaps.
• Low: cosmetic report differences with documented causes.

Every recommendation needs an owner, implementation step and retest. After release, compare event counts with source systems and annotate the change date.

Monitor sudden count shifts, new conversion names, (not set), unassigned traffic, failed tags, CRM handoff failures and conversion-action setting changes. Schedule revalidation after form, checkout, consent, tag or site releases.

Growthjunction’s analytics and conversion tracking service audits the full route from the user action to the sales outcome so campaigns are not optimised against a false signal.