Google Ads audit: can you trust the conversion data?
Most Google Ads audits examine keywords, bids, budgets and campaign structure. Every recommendation that comes out of one shares a single assumption: that the conversion data used to judge performance is accurate. That assumption is rarely the thing being audited.
There are two different audits, and this is the second one
Both are legitimate work. They answer different questions, and one of them depends on the other being right.
- Campaign and ad group structure
- Keywords, match types and search terms
- Bidding strategy and budget allocation
- Ad copy, extensions and quality scores
- Audiences, placements and geographic targeting
This is what most people mean by a Google Ads audit, and it is not what this service does. If it is what you need, a campaign management specialist is the right person.
- What each conversion action actually counts
- Whether the ad click survives the journey to the enquiry
- Whether forms, calls and purchases are measured at all
- Whether the same customer is counted once or several times
- Whether the reports reconcile against what the business received
This is what an independent conversion audit does. It says nothing about your keywords or your bidding, and everything about whether the numbers those decisions rest on describe real customers.
Which audit should come first?
The measurement one, because campaign decisions are made using the numbers measurement produces. Optimising toward a conversion that does not represent a customer moves the account further from where it should be, faster.
This is a claim about order, not about importance. Campaign work is where the money is actually managed, and a good campaign manager will produce more value than a measurement check ever will.
But a campaign audit performed on unverified conversion data can only be as good as the data it inherited. Recommendations about which campaigns to scale, which to cut and where to move budget are all derived from the conversion figures. If those figures describe the wrong actions, every recommendation built on them inherits the error and looks well-reasoned while doing it.
What this cannot tell you An external check cannot tell you whether your conversion data has already been verified by somebody competent. If it has, this is a short conversation and a cheap one.
Can a Google Ads audit be done without account access?
Partly, and the boundary is worth being precise about. An external check can observe how your website handles a paid click, what conversion requests it sends and when, whether forms and calls appear to be measured, and how the page performs for someone arriving from an ad.
What it cannot see is anything that lives inside the account: which conversion actions are set as primary, what your campaigns are bidding toward, whether offline or imported conversions are in use, and how any of it reconciles against your actual sales.
So the honest description of what a public check produces is a list of things worth investigating, with the evidence for each, and an explicit statement of what could not be established. That is genuinely useful, and it is not the same as a verdict.
What this cannot tell you Anybody offering a definitive audit of your conversion accuracy without account access is overstating what the evidence supports. The limits are published on the methodology page rather than discovered later.
What if the campaigns turn out to be the real problem?
Then the audit says the measurement is sound and you have spent very little establishing it. That is a useful result: it removes the explanation that is hardest to see from inside the account, so the remaining explanations are the ones your campaign manager can act on.
A conversion audit that finds nothing is not a failed audit. It converts an unknown into a known, and the alternative is optimising campaigns while quietly unsure whether the scoreboard is right.
Start with the data the audit would rest on.
The free check needs no account access and returns a written record of what was observed, what it could mean, and where the evidence ran out.
What does a Google Ads audit normally cover?
Usually campaign structure, keywords, bids, budgets, ad copy and quality scores. Those are the levers a campaign manager can pull, so that is what a campaign-focused audit examines.
Every recommendation that comes out of that work shares one assumption: that the conversion data used to judge which campaigns are performing is accurate. That assumption is rarely the thing being audited.
Is this a full Google Ads account audit?
No, and it is worth being direct about that. This examines conversion measurement: whether the actions being counted as conversions correspond to real customer outcomes, and whether the connection between an ad click and that outcome holds up.
It does not review your keywords, your bidding strategy, your ad copy or your budget allocation. If those are what you need reviewed, a campaign management specialist is the right person and this is not that service.
How often should conversion measurement be checked?
Once, and then after anything that changes the website. There is no useful calendar interval, because the thing that breaks measurement is a change rather than the passage of time.
The events worth re-checking after are specific: a website rebuild or redesign, a new form or booking system, a consent banner change, a domain migration, a new landing page, or a change of whoever manages the advertising. An account that has sat untouched for two years behind a website that has also sat untouched is unlikely to have drifted.
Do you work with agencies rather than advertisers directly?
Yes, and it is a sensible arrangement for an agency inheriting an account. A measurement check at the start of an engagement establishes what the previous setup was doing before any of it becomes your responsibility to explain.
The report is written the same way regardless of who commissions it. It describes what was observed and what could not be established, and it does not assign blame, which is as useful to an agency being judged on inherited numbers as it is to the advertiser reading them.