How much did we spend?
The total, how it changed and where the money was allocated.
You do not need to assume your agency is dishonest. You do need enough independent visibility to understand what was spent, what was tracked and what deserves an explanation.
Most agency relationships are not a simple choice between complete trust and complete distrust. An agency may be skilled, honest and still produce reports that focus on platform metrics rather than the financial reality of the business. A business owner should be able to verify the story without becoming an advertising expert.
The total, how it changed and where the money was allocated.
Purchases, qualified leads or another agreed business outcome—not only clicks or impressions.
Performance compared with the target return, margin or acceptable acquisition cost.
A like-for-like comparison with the previous period and an explanation of the main causes.
The creative, targeting, keyword, landing-page or budget changes being evaluated and how success will be judged.
The amount of data available, tracking limitations and any attribution overlap that could affect the answer.
Reach, impressions and clicks may have improved while the cost of a sale or lead deteriorated.
Platform attribution does not account for margin, refunds, repeat customers or duplicated claims across channels.
A return can look positive while still being below the level your business needs to break even.
Learning periods are real, but the agency should define what evidence or date will trigger a decision.
More spend is not automatically good if the account is already acquiring customers at an unacceptable cost.
Missing values, duplicated conversions and different attribution windows can materially change the story.
AdSpend Lens is designed to connect read-only to Meta and Google Ads, calculate findings from synchronised data and present the evidence in plain English.
No. Advertising data is genuinely complicated, and an agency may be producing good work while communicating it badly. The point is to make the evidence understandable so performance can be discussed properly.
At minimum: spend, the outcomes that matter to the business, tracked value where relevant, cost per result, comparison with an agreed target, changes over time, important tests and any tracking limitations.
No. ROAS can be useful but may ignore margin, new versus existing customers, lead quality, attribution overlap and the campaign’s intended purpose.
That can be reasonable. Ask what amount of spend, number of conversions or time period will be considered enough to judge it, and what decision will be made at that point.
Use independent account data to understand the report, recognise uncertainty and ask for clear explanations.