Audit
Conversion tracking audit
Fourteen areas across GA4, Google Ads, the tag manager and the CRM, verified rather than assumed.
A conversion tracking audit checks that the numbers your decisions rest on are true. It covers fourteen areas across GA4, Google Ads, the tag manager, the data layer and the handoff into your CRM, and it ends with the question most measurement setups fail: whether every system agrees on what a conversion is. This is the audit that most often changes what the other three would conclude, which is why we run it first whenever the numbers and the sales have stopped agreeing.
Coverage
What does a conversion tracking audit cover?
Fourteen named areas across GA4, Google Ads, the tag manager, the data layer, consent and the handoff into your CRM.
The audit covers fourteen areas of the measurement stack. All of them are listed, because in this discipline the areas nobody thinks to check are exactly where the faults are.
GA4 property configuration and key events
Data streams, key events, and whether the property is set up to answer the questions the business actually asks of it. A default install answers almost none of them.
Google Ads conversion actions, sources and attribution
What the ad account counts, where each action comes from, which are Primary, and what attribution setting is applied. These four decide what bidding optimises toward.
Conversion imports and offline conversion handling
Where the real outcome happens after the click, in a CRM or a phone call or a signed contract, whether it makes it back into the platform at all, and whether it arrives in time to be useful.
GTM container structure and hygiene
Tags, triggers and variables, including the ones firing on pages nobody remembers and the ones left paused since a migration. A container nobody has pruned is a container nobody can reason about.
Data layer implementation and event naming
Whether the site pushes its own events with consistent names, or whether triggers are scraping the DOM and waiting to break the next time the markup changes.
Consent mode and privacy compliance
Whether consent defaults are set before the tag loader runs rather than after it, whether choices are honoured, and whether the modelling that depends on all of it is actually working.
Duplicate, missing and inflated conversions
The three failure modes, each with a different cause and a different fix. The most common is the same lead counted twice by two systems, which silently halves the reported cost per acquisition.
Cross-domain and subdomain tracking
Where a journey crosses a domain, a payment provider or a booking system, whether the session survives it. When it does not, every conversion is credited to the last hop rather than to the source.
Form and CTA event coverage
Whether every meaningful action on the site is measured, and whether the measured ones fire once, on success, rather than on click and on failure too.
Platform to CRM handoff and lead attribution
Whether a lead arrives in the CRM carrying where it came from, and whether the outcome ever travels back. This is where most businesses lose the ability to tell which campaigns produce revenue rather than forms.
Channel and UTM taxonomy
Whether campaign tagging is consistent enough to group, or whether the channel report is thirty spellings of the same thing. Taxonomy is boring and it decides whether any report can be trusted.
Server-side versus client-side coverage
What is measured where, what each approach is currently costing you in lost data, and whether a server-side move would actually fix the problem you have rather than the one in the vendor's case study.
The reporting layer
Whether the reports people act on are built on the definitions that were just verified, and whether anyone would notice if a feed stopped updating.
Definition alignment
The one that matters most: whether the ad platform, the analytics property, the CRM and the finance spreadsheet all agree on what a conversion is. When they do not, every meeting about performance is four people reading four different numbers.
Method
How is the tracking verified?
By triggering the events ourselves and following each one through every system it touches, rather than reading a configuration screen and believing it.
This is the same audit we run on an account before we agree to take it on. We use it to decide whether we can help, which is a stronger reason to make it accurate than any promise about it would be.
Read-only access, agreed first
You grant read-only access to the analytics property, the Google Ads account and the tag manager container, and to the CRM where lead attribution is in scope. We make no changes to any account during an audit, and access can be revoked as soon as the walkthrough is done.
Trigger the events, do not read the settings
We complete the actions ourselves and follow each one through every system it touches. A configuration screen tells you what somebody intended, which is not the same as what happens when a real person submits a real form on a real phone.
Reconcile the systems against each other
The platforms are compared to each other and to whatever the business counts as a sale. Disagreement between them is the finding, and the size of the disagreement tells you how much of the reporting can be trusted.
The judgement pass
Most free and cheap audits in this market are tool-generated scans: a crawler or a script runs a fixed rule set and formats the output. Ours is run by hand by the senior specialist who would do the work, which is why it reads like a person wrote it and why it can tell you which three findings matter out of forty.
The walkthrough and the specification
We take you through the findings on a call and leave you with a measurement specification: one definition of a conversion, and what each system has to do to honour it. Your own developers can implement it without us.
Deliverables
What do you receive at the end?
A written document, the evidence for each fault, a measurement specification, and one agreed definition of a conversion across every system.
- A written findings document covering all fourteen areas, ranked by impact and effort
- The evidence for each fault, including what we triggered and what each system recorded
- A reconciliation showing how far the platforms disagree, and where the difference comes from
- A measurement specification: one definition of a conversion and what each system must do
- An implementation order written so your developers can act on it directly
- A call to walk through it, and the document afterwards whatever you decide
Where this leads
- Marketing analytics services
What the ongoing work covers, if the audit concludes it is worth doing.
- Marketing analytics cost
Third-party market data, named and dated, with what each figure does not tell you.
- How pricing works
Why an audit is quoted against scope rather than listed, and what drives the figure for a given account.
The do-it-yourself version
The method is published rather than held back. If you would rather run it yourself, these are written so you can.
Fit
Who is a tracking audit for?
Anyone whose platform numbers and actual sales have stopped agreeing, and any account about to hand its budget decisions over to automated bidding.
- Anyone whose platform-reported conversions and actual sales have stopped agreeing
- Accounts running automated bidding, which is only ever as good as the signal underneath it
- Businesses whose leads land in a CRM with no record of where they came from
- Sites that have migrated, replatformed or changed consent tooling and never re-verified the measurement
- Teams about to build reporting, who would rather verify the inputs than automate a wrong number
When it is not the right thing to buy
If your measurement was verified recently and the numbers reconcile with your sales, you do not need this and we would rather say so. If the site has almost no traffic and one form, the setup is small enough that a fix is cheaper than an audit of it, and we will quote the fix instead.
If that is where you are, the free audit is the better place to start. It costs nothing and it will tell you whether there is enough here to be worth a paid one.
Questions
What do buyers ask about tracking audits?
Why the platforms disagree, what access is needed, whether consent mode breaks measurement, and whether we fix what we find.
Why do Google Ads and GA4 report different numbers?
Because they are answering different questions. They use different attribution models, different conversion windows and different counting rules, so a gap is expected and a specific size of gap is not. The audit establishes which differences are structural and which are faults, which is the only way to know whether a discrepancy is worth chasing.
What access do you need?
Read-only access to the analytics property, the Google Ads account and the tag manager container. Where lead attribution is in scope we also need read access to the CRM, because whether the outcome travels back is exactly the thing that cannot be checked from the platform side.
Does consent mode break measurement?
Not by itself, but implementing it in the wrong order does. Consent defaults have to be set before the tag loader runs, and a default applied afterwards is a default that never applied at all. That ordering fault is common, invisible in the interface, and one of the specific things this audit checks.
Do you fix what you find?
Only if you want us to, and it is quoted separately. The measurement specification is written so your own developers can implement it, because tracking work usually sits with whoever owns the site rather than with the marketing team.
Should this come before the other audits?
Usually yes, if the numbers are in doubt. Every finding in a Google Ads or CRO audit is derived from conversion data, so if that data is wrong the other audits produce confident conclusions built on a false floor. When we see both being considered we say which order we would run them in and why.
How does this compare to the free audit?
The free audit is a review that surfaces the biggest issues we can see and tells you which are worth fixing first, delivered on a call. This is the fourteen-area read, with each fault evidenced by what we triggered and what each system recorded, ending in a measurement specification. If you only need to know whether the numbers can be trusted, start with the free one.
The other audits
Google Ads audit
Eighteen areas of the account, read by hand by the person who would run it.
SEO audit services
Sixteen areas, read by a person, with the findings ranked rather than dumped.
Conversion rate optimization audit
Thirteen areas of the conversion path, ending in a test backlog you can actually run.
All four, and how they fit together, are on the audit services page.
Request the Tracking audit
Tell us what you are running and we will scope it, then quote against that scope in writing before anything starts. You keep the findings whatever you decide afterwards.