Google Ads Conversion Tracking Audit Checklist

A campaign can appear to generate a healthy volume of conversions while producing few worthwhile enquiries, duplicated leads or no clear revenue return. That is usually a measurement problem before it is a media problem. A Google Ads conversion tracking audit tests whether the actions reported in the platform represent genuine commercial progress - and whether the data is reliable enough to guide budget decisions.

For UK businesses spending meaningful sums on paid search, unclear tracking creates a costly chain reaction. Automated bidding learns from the wrong signals, reports overstate performance, and teams optimise campaigns around activity rather than better leads. The result is not simply imperfect reporting. It is budget being directed with less confidence than it should be.

What a Google Ads conversion tracking audit should establish

The purpose of an audit is not to count how many tags are installed. It is to establish a clear answer to four commercial questions: what counts as a conversion, whether it is recorded accurately, which campaigns influenced it, and whether the business can judge its value after the lead arrives.

Google Ads can track many actions: form submissions, telephone calls, purchases, newsletter sign-ups, brochure downloads and clicks on contact details. Not all of them deserve equal status. A completed enquiry form may be a useful primary conversion for a professional services firm. A click on a phone number may be helpful context, but it is not proof that a call connected or that the caller was a suitable prospect.

The audit should separate meaningful outcomes from early indicators of intent. This gives the account a stronger foundation for bidding, reporting and day-to-day optimisation.

Start with the conversion definition

A conversion definition should reflect the way the business actually grows. For an ecommerce advertiser, this is normally a completed purchase with an accurate transaction value. For a B2B company, it may be a sales-qualified lead, booked consultation or accepted opportunity rather than every form completion.

There is a trade-off. Waiting only for qualified leads can leave too little conversion volume for bidding systems to learn efficiently, particularly in lower-volume sectors. Treating every interaction as equally valuable, however, gives Google Ads permission to find more low-value interactions. A sensible approach often uses a primary conversion for the main commercial action, alongside secondary actions that help diagnose user behaviour without driving bidding decisions.

This distinction should be visible in the Google Ads conversion settings. If page views, button clicks and form submissions are all marked as primary, the account may be optimising for the easiest action rather than the most valuable one.

Check whether tags fire once, at the right moment

A tag that fires is not automatically a working conversion. It must fire when the defined action has completed, only once per event, and with the right information attached.

Thank-you page tracking can work well when every successful form submission reaches a unique confirmation page. It is less dependable where visitors can reload the page, return to it from their browser history, or reach it without completing a form. In these cases, event-based tracking triggered by a confirmed successful submission is generally more reliable.

Test the main conversion paths yourself. Submit each live form, make a test purchase where practical, and place calls through tracked call assets or a call tracking provider. Check whether the event appears in the relevant debugging tools and whether it reaches Google Ads as expected. Testing needs to cover desktop and mobile, because mobile forms, click-to-call actions and consent settings can behave differently.

Duplicate tracking is a common source of inflated results. It can happen when the same conversion is imported from Google Analytics and tracked directly by a Google Ads tag, when multiple tag containers run on the site, or when a thank-you page is reloaded. For lead generation accounts, this can make cost per lead look attractive while the sales team sees no corresponding increase in enquiries.

For purchases, transaction IDs matter. They allow Google Ads to deduplicate repeat signals and prevent a refreshed confirmation page from being counted as several sales. The transaction value, currency and product data should also be checked against the ecommerce platform, not assumed to be correct because sales are showing in the account.

Review conversion settings, attribution and values

Tracking accuracy depends as much on configuration as implementation. A Google Ads conversion tracking audit should review each action's count setting, value, conversion window and attribution model.

The count setting is particularly important. For a purchase, every transaction is normally valuable, so counting every conversion makes sense. For a lead form, one conversion per ad interaction is often more appropriate. Counting every event can overstate leads when the same person submits several enquiries.

Conversion values need the same scrutiny. Fixed values can be useful where a lead has a broadly consistent commercial worth, but they should be grounded in actual close rates and customer value, not selected to make reporting look more sophisticated. If values differ significantly by service, product or lead type, dynamic values or imported offline outcomes may give a truer picture.

Attribution is not a magic answer to causality. It is a rule for assigning credit among touchpoints. Data-driven attribution can be useful where an account has enough conversion activity and a meaningful customer journey, while last-click approaches may simplify reporting in smaller accounts. The key is consistency: stakeholders need to understand what the chosen model reports and avoid comparing figures from different models as though they are identical.

The conversion window also needs to match buying behaviour. A seven-day window can under-credit campaigns for a service with a six-week sales cycle. A 90-day window may overstate the immediate impact of advertising if prospects routinely return through other channels. There is no universal setting. The right period depends on the time between first click, enquiry and sale.

Follow the lead beyond the form fill

Website conversions are only part of the measurement picture for many London businesses. A form submission tells you that someone completed a form. It does not confirm that the phone number was valid, the company was in your target market, or the opportunity became revenue.

This is where CRM integration and offline conversion imports become commercially important. By passing qualified lead, opportunity or closed-sale outcomes back into Google Ads, the account can be assessed against lead quality rather than raw volume. Over time, this also gives bidding systems better signals.

The process does require discipline. The CRM must capture the original click identifier where possible, sales teams need clear lead-status definitions, and imported outcomes must be timely. If qualification is inconsistent, importing the data will not solve the underlying reporting problem. It may simply formalise it.

Where a full integration is not yet practical, start with a regular reconciliation. Compare Google Ads leads with CRM records by date, source and campaign, then review contact rates, qualification rates and sales outcomes. This often exposes campaigns that look efficient in-platform but create poor-quality demand.

Check consent, cross-domain journeys and reporting gaps

Privacy controls affect what can be observed. Consent banners, browser restrictions and ad blockers can reduce recorded conversions, especially where tags are prevented from firing until permission is granted. That does not necessarily mean paid media is underperforming. It means the reporting has a coverage gap that should be understood.

Consent Mode can help model some missing measurement where implemented correctly, but it is not a substitute for valid consent management or clean tracking. Review whether consent signals reach the Google tag, whether tags respect the user's choices, and whether the organisation's privacy process aligns with the way data is collected.

Cross-domain journeys also deserve attention. A user may click an ad, browse the main website, then complete payment through a separate booking or checkout domain. Without appropriate cross-domain configuration, the click information can be lost and the conversion credited incorrectly. Similar issues arise with embedded form providers, third-party scheduling tools and call tracking platforms.

Finally, compare Google Ads figures with analytics, the CRM and finance data. These systems will not match perfectly because they use different attribution rules, time zones and definitions. Large or unexplained differences, however, are a reason to investigate rather than a normal reporting inconvenience.

Turn audit findings into practical priorities

An audit should end with a prioritised action plan, not a list of technical observations. The immediate priorities are usually broken primary conversions, duplicate tags, incorrect values and conversion actions that are training bidding towards low-quality leads. These can materially affect budget allocation straight away.

The next layer includes CRM feedback, consent configuration, cross-domain tracking and more useful conversion values. These changes may take longer, but they improve the business's ability to judge performance beyond cost per click and cost per lead.

At Invaro Media, the useful standard is straightforward: reporting should show what is generating better leads, what is wasting budget and what should be prioritised next. If a conversion cannot support one of those decisions, it may still be an interesting metric, but it should not be steering the account.

A tracking audit is most valuable when it changes how decisions are made. Before increasing spend, ask whether the account is measuring an action your sales team would genuinely want more of. If the answer is uncertain, fixing that question is likely to be the most profitable optimisation available.