The questions that separate an agency that will move your revenue from one that will send you a very attractive report.
Agency selection is usually decided on chemistry and a deck. Both matter, and neither predicts results. These are the questions that do — and, more importantly, what a good and a bad answer to each actually sounds like.
If you want the surrounding context first, what an agency actually does and what one costs are the two pieces to read before the meeting.
1. Who actually does the work?
The people in the pitch are frequently not the people on the account. Ask who will be doing the work day to day, how much of their time you are buying, and whether you will meet them before you sign.
The follow-up is the one that matters: how many other accounts does that person carry? A strategist with six and a strategist with twenty-five have the same job title and are selling entirely different products. A good answer names people and numbers. A bad one describes a "team" without either.
2. What will you measure, and against what baseline?
An agency that cannot describe how it will judge its own performance in the first month is not going to produce an uncomfortable report in the sixth. Agree the baseline before work starts, not after.
Insist that the baseline is written down and shared. Baselines have a way of becoming retrospectively flexible once results are in, and the argument is unwinnable if the only record of the starting position is a slide from a meeting nobody minuted.
3. What would make you tell us to stop?
A good answer exists. "If your average order value cannot support paid acquisition at these CPCs, we would tell you to fix pricing before spending" is the kind of thing you want to hear. An agency that will never advise against its own services is selling capacity, not judgement.
This question also reveals fee structure honestly. An agency paid a percentage of media spend has a structural reason to recommend more media; that does not make the advice wrong, but you should know the pressure exists before you weigh it.
4. Who owns the accounts and the data?
Ad accounts, analytics, the site, the content. If leaving means starting from nothing, you are not a client so much as a tenant. Ownership should be yours from day one and it should be in the contract.
Get specific: ask whether the Google Ads account will be created under your own Google account with them granted access, or inside their manager account. The second is portable only with their cooperation, and the conversion history that makes an account perform does not always survive the move. What to own before you sign covers the full list.
5. Show me work in my situation, not my sector
Sector experience is overrated; situation experience is not. A team that has taken a business from one location to twelve understands your problem better than one that happens to have worked with a competitor.
Ask for a case where it did not work, and what they changed. Every agency has several. The ones worth hiring will tell you about one without being pressed, and the account of what they learned is more informative than any success story.
6. What happens in the first thirty days?
A specific answer — audit, measurement, quick wins, a plan — suggests a repeatable process. A vague one suggests you will be the pilot.
Look for measurement early in that sequence. An agency proposing to start with campaign launches before establishing what is currently happening has committed to being unable to prove its own contribution later.
7. What is the notice period, and why?
Long lock-ins are sometimes justified by genuine set-up cost. Often they exist because the work does not hold up to a monthly decision. Ask which it is, and expect a straight answer.
Where a longer term is genuinely warranted — SEO and content both take four to twelve months to prove — a reasonable compromise is a longer initial term with defined checkpoints, rather than a rolling commitment with no review built into it.
The question to ask yourself
What will be true in twelve months if this works? If you cannot answer that in a sentence with a number in it, no agency can be held to it, and the engagement will be judged on how the reports look rather than on what happened to the business.
Agreeing that sentence before you sign is the single most useful thing you can do — and a good agency will help you write it rather than let you sign without one.
Common questions
- What should you ask a marketing agency before hiring them?
- Ask who does the work day to day and how many accounts they carry, what will be measured and against what recorded baseline, what circumstances would make them advise you against their own services, who owns the ad accounts and analytics data, what happens in the first thirty days, and why the notice period is set where it is. Specific answers to all six are a strong signal; vagueness on any of them is worth pressing.
- What are the red flags when choosing a marketing agency?
- Guaranteed rankings or guaranteed lead volumes, reporting built around impressions and engagement rather than leads and cost per lead, senior people who disappear after the pitch, refusal to give the client ownership of ad accounts and analytics, no defined baseline, and a long lock-in with no review checkpoints.
- Should a marketing agency guarantee results?
- Be wary of any guarantee about rankings or lead volume, since neither is within an agency’s control. What can reasonably be committed to is process: a defined scope of work, an agreed measurement baseline, reporting on stated metrics and a review point at which continuing is a genuine decision.