Five things to check yourself before the next agency report lands.

The monthly report says impressions are up, engagement is up, and the campaign is performing well. The phone is not noticeably busier and you cannot tell whether those two facts are related.

Rather than asking whether the agency is any good, which nobody can answer in a meeting, check something narrower: can the numbers you were shown be reproduced from accounts you already control?

Five checks. An hour, once a quarter. It applies to us as readily as to anyone else you might be paying.

The five

What you actually appear for

Open Search Console and look at your queries rather than your traffic. The question is whether you appear for the things you sell, or only for your own business name. Brand searches are people who already knew you existed, so a report built largely on them is describing your existing reputation rather than any work done to widen it. Sort by impressions and read the top twenty phrases out loud. If most of them contain your company name, or the name of the town plus nothing you actually charge for, that is worth raising before anybody discusses strategy. It is also the single cheapest way to tell whether the work has been aimed at customers or at a dashboard.

Whether the profile chart is measuring ads

This is the check with the most teeth, and almost nobody knows to run it. Google's documentation states that Business Profile performance data includes views, searches and actions from Google Ads as well as organic results, while the Searches panel reflects organic results.

So a chart showing profile activity climbing through a month when ad spend also climbed is not evidence about your local SEO, however it is captioned. If that chart is the centrepiece of your report, ask which portion was paid. The counting quirks are worth understanding before the conversation rather than during it.

Whether anything is defined as a conversion

In GA4, check whether a key event exists at all. A property with none is measuring attention rather than outcomes, which means every number in the report is a proxy. If sessions are up and nobody can tell you what a conversion is, that is the finding.

Your own listing, on your own phone

Two minutes, done as a customer rather than as an owner. Is the phone number right and under your control. Does the Book button go where you expect, given that third-party links can appear without anyone adding them. Do the hours match reality this week. None of this shows up in a report, and all of it is visible to every customer.

A password reset

Trigger one on an account they manage and see whose inbox it reaches. If the answer is theirs, that outranks every metric on the page.

What you are really testing

None of these prove anybody is doing bad work. Search can be flat for months while genuinely good work compounds underneath, and a quarter with no movement is normal rather than damning.

What they test is reproducibility. A report you cannot rebuild from your own accounts is not a report; it is a summary you are being asked to take on faith, and the difference matters most when the relationship is going badly and everyone is describing the same period differently.

What a good answer sounds like

A capable agency will not be defensive about any of this, because the numbers are theirs too. The answers you want are specific: here is the organic share, here is what we changed, here is the thing that did not work.

The answer to be wary of is the one that reframes the question. If asking which portion of profile activity was paid produces a discussion about brand awareness and long-term positioning, you have your answer, and it did not require you to understand any of the underlying tools.

If you want a second opinion on a report somebody sent you, book a strategy call and bring it. We will tell you which parts are real, including when the honest answer is that your current agency is doing perfectly sensible work and the problem is somewhere else entirely.