Do you need an independent Google Ads audit?
Possibly, and it does not mean replacing anyone. An independent audit checks whether the conversion numbers used to judge your advertising reflect real enquiries, calls and sales. Your agency keeps running the campaigns.
No account access required · Nothing changes in your campaigns
- Not a pitch to take over your Google Ads.
- Not a review of your keywords, bids, budgets or ad copy.
- Not a report card on whoever currently manages the account.
- Not something that requires you to change agencies, or tell them anything.
- A check on whether the conversions being reported represent real customer actions.
- A check on whether real enquiries are reaching the reports at all.
- A written record of what was observed, and what could not be established.
- A document you can forward to your agency and ask about.
Why not just ask the agency to check it?
You can, and many will do it competently. The reason independent verification is worth having is structural rather than technical: the conversion numbers are the numbers their performance is judged by.
This is not an allegation about anybody. It is the same reason a company does not audit its own accounts, and the same reason a builder does not sign off their own inspection. Nobody searches as hard for faults in their own scoreboard, however capable and well-intentioned they are.
There is also a practical reason. Measurement setups are usually assembled in layers over several years by different people: a previous agency, a web developer, a plugin, an analytics contractor. The person managing the account today often inherited it and has no more visibility into why it was built that way than you do.
What this cannot tell you An external check cannot tell you whether your agency has already verified this. If they have, and the setup is sound, the check should confirm it and you have lost nothing but the time it took to ask.
What happens if the audit finds something?
You get a written description of what was observed, what it could mean commercially, and what would need account access to confirm. It does not assign blame, and that is deliberate rather than diplomatic.
Attributing a finding to whoever currently manages the account is frequently just wrong, for the reason above: most of these setups predate the current arrangement. A report that names a culprit is a report built on a guess.
What you can do with a description of the current state is have a specific conversation. "A conversion request appears to fire on page load, is that action being used as a bidding goal?" is a question any competent agency can answer in ten minutes, and the answer is often a perfectly good one.
What if the audit finds nothing wrong?
Then you have a written, independent confirmation that your measurement is sound, which is a useful document for both you and your agency. The fee for an authorised audit is not contingent on finding faults, specifically so that this is not a disappointing result to report.
This matters more than it sounds. A business that profits from finding problems has an obvious incentive to find them, and the only credible answer to that is a pricing structure and a reporting standard that do not reward it.
It is also the honest expectation to set. A meaningful share of setups are broadly fine, and the value of checking is knowing which kind you have rather than assuming.
Will this create a problem with our agency?
Usually not. A verified scoreboard is easier to be judged against than a disputed one, and a written confirmation that the measurement is sound is a document a good agency can use.
The free check needs no account access and looks only at what your website shows any visitor, so there is nothing to disclose if you would rather not raise it yet.
An authorised audit does need read access, which in practice means asking. Most clients simply do. If you would prefer not to, the check goes as far as public evidence allows and the report states exactly where that boundary fell.
What this cannot tell you One thing worth noting plainly: a strong negative reaction to the idea of independent verification is itself informative, whatever the explanation turns out to be.
Keep your agency. Verify the numbers anyway.
The free check needs no account access, changes nothing, and returns a written record of what was observed and what it could not establish.
What permissions would an authorised audit need?
Read-only access, granted by whoever holds admin rights, and revocable at any time. Typically Google Ads, the analytics property, and the tag manager container if one is in use.
Nothing is changed, no settings are touched and no campaigns are paused. If read access to any of those is refused or unavailable, the audit proceeds without it and the report states which questions consequently could not be answered.
Can you audit an account managed by a large agency?
Yes, and the size of the agency changes very little about the work. What it sometimes changes is how long the access request takes to move through their process.
It is worth saying that large and competent are not the same variable here. Measurement drift accumulates through staff turnover and inherited setups, both of which happen more in larger organisations, not less.
We run Google Ads in-house. Does any of this still apply?
Yes, and the structural argument is the same one. Whoever manages the advertising is judged by the conversion numbers, whether they sit in your building or invoice you monthly.
In practice in-house setups have a distinct pattern: nobody inherited a badly documented account from a predecessor, but nobody was ever given the measurement work as an explicit responsibility either. It sits between marketing and whoever maintains the website.
Can the check be done before we hire an agency?
That is arguably the best time to do it. You establish what the measurement currently says before anybody has an interest in how it reads, which gives you a baseline that is not disputed later.
It also changes the conversation you have with candidates. Asking a prospective agency what they make of a specific documented observation tells you considerably more about them than asking what their approach is.