Nine questions to ask before hiring an agency

Choosing an agency is difficult because the pitch is the product. You are shown a deck by the people who are best at selling, then handed to whoever has capacity. By the time you find out which one you hired, you are three months and a retainer in.

These are the questions that get past the deck. We wrote them knowing we will be asked them too.

1. Who is actually doing the work?

Ask for the names of the people who will run your account day to day, and whether they are in the room today. It is common for a senior team to win the work and a junior team to deliver it. That is not automatically wrong — juniors have to learn somewhere — but you should know it before you sign, not after.

A good answer names people, explains who does what, and is honest about seniority. A bad answer stays at "our team" for the whole conversation.

2. What would make you turn this work down?

An agency that has never declined a project is telling you something. Every honest shop has categories it is bad at, budgets too small to serve properly, and clients it should have refused.

A good answer is specific and slightly uncomfortable. A bad answer is that they can help anyone.

3. How will we know in three months whether this is working?

You want a number, named now, that you will both look at later. Not "increased visibility" — an actual metric with a baseline.

The important half is what happens if the number does not move. Ask directly: what would you do if we get to month three and this has not worked? An agency that has thought about it will have an answer. One that has not will tell you it always works.

4. What do you need from us?

Engagements fail on the client side more often than anyone admits — approvals that take three weeks, a subject-matter expert who never has time, access that never arrives.

A good agency will tell you what it needs from you and what happens if it does not get it. If nobody mentions your obligations, the eventual delay will be framed as your fault anyway; better to have it named up front.

5. Do we own everything?

Ask specifically about:

• Ad accounts — set up under your billing, with you as owner
• Analytics and tracking — your property, not a sub-account of theirs
• Website code and designs — yours, including source files
• Content — yours, including anything drafted but unpublished

The test is simple: if we part ways next month, what do we walk away with? If any answer is "our platform" or "our proprietary system," you are renting your own marketing.

6. How do you charge, and does that change what you recommend?

Percentage-of-spend pricing pays an agency more when you spend more. That is a real conflict, not a theoretical one — it makes "increase the budget" the easiest recommendation in the room.

Flat retainers have their own failure mode: an incentive to do the minimum that keeps the invoice paid. Neither model is disqualifying. What matters is whether the agency can explain its own incentives without becoming defensive.

7. Can I speak to a client who left?

Everyone will give you a happy reference. The useful conversation is with someone whose engagement ended.

Some agencies genuinely cannot arrange this, and that is fair. But watch the reaction to the question. Discomfort is more informative than the answer.

8. What did you get wrong recently, and what changed?

You are testing whether the agency runs a real post-mortem or just moves on. A specific failure, honestly described, with a process change that came out of it, is the strongest signal you can get in a pitch.

Vagueness here — "we're always learning" — usually means nothing is written down.

9. What would you do first, and why that?

Ask for the first thirty days in order. A considered answer starts with finding out what is true — an audit, access, a baseline — before doing anything visible.

Be careful of anyone who arrives with a full solution before seeing your data. It means the plan was written before they met you, and you will be getting the same plan as everyone else.

The pattern underneath all nine

Every question above is really the same question: will this agency tell me something I do not want to hear?

That is the whole difference. Channels can be learned and tools can be bought. An agency that will say this will not work, do not spend the money is rare, and it is worth more over three years than a better media buyer.

The questions are useful in both directions, incidentally. Sitting on our side of the table, the clients who ask them are consistently the ones the work goes well with — because they are the ones who will engage with an honest answer when we give one.