Eleven questions that separate a paid-social agency from a slide deck

You should read this knowing who wrote it. We sell the thing this post helps you buy, which means the honest way to write it is to include the questions we'd struggle with — so those are marked, and there are three of them.

The questions below are ordered by how much they reveal per minute spent. The first four will tell you more than an hour of case studies.

1. What does it cost, and can you tell me now?

Not a range dependent on a discovery call. A number, or a clearly-stated method for arriving at one.

Good answer: a price, or "percentage of spend at this rate, with this minimum" — stated immediately. Bad answer: "it depends on scope, let's book a call to explore." Pricing opacity is rarely about complexity; it's about wanting to price you after they've read how much you want it. Ours are on the pricing page because the alternative told us something we didn't want to be true of us.

2. How many creatives will I get per month, in a number?

Then do the arithmetic in front of them: the creatives you need per week is how many you want live divided by their effective life in weeks. If their monthly output is below your replacement rate, the relationship will feel like fatigue for as long as it lasts. That's the production maths, and it doesn't care how good anyone's portfolio is.

Bad answer: "unlimited." Unlimited is a queue with a marketing name. Ask what the queue depth is at month three with their current client count and watch what happens.

3. Who actually does the work, and what else are they on?

The pitch is senior. The delivery frequently isn't. Ask for the name and seniority of the person who'll touch your account weekly, and how many other accounts they hold.

Good answer: a name, a level, a number of accounts. Bad answer: "you'll have a dedicated account manager" — which describes who emails you, not who does the work.

4. What happens to my accounts, assets and pages if I leave?

Good answer: you own everything, it's all in your business manager already, and leaving is an access change. Bad answer: anything involving a handover process, an asset-transfer fee, or work living in their ad account rather than yours. Ask this early — it's the cheapest question here and it occasionally ends the conversation on the spot.

5. Show me a campaign that failed and what you did.

Everyone has losers. Only some will show you one. What you're testing isn't the failure — it's whether the diagnosis was mechanical ("we misread the audience overlap and consolidated too late") or atmospheric ("the market was tough that quarter").

6. What would make you tell me to stop spending?

Good answer: a specific threshold — a margin, a payback period, a spend level below which the maths doesn't work. Bad answer: the idea that more spend is always the answer. An agency paid a percentage of spend has a structural reason never to reach this conclusion, which is worth naming out loud to their face and watching how they take it.

7. How will I know what changed in my account, and can it be undone?

Good answer: a log you can read, with reversibility. Bad answer: a monthly report summarising outcomes with no record of actions. You should never have to reconstruct what someone did to your account from a performance graph.

8. What's the contract length, and what's the exit?

Six and twelve-month lock-ins are common and are sometimes justified — a genuinely long ramp is a real thing. But a lock-in is a bet that you'll want to leave before the work speaks for itself. Ask why it exists and listen for whether the answer is about your outcome or their revenue predictability.

Some context for how much weight to put on this one. A 2025 opt-in survey of 376 agency people — mostly North American, mostly firms under 25 people — found a quarter reported their typical client engagement lasts under twelve months, while about a third exceeded three years.^[1] Short relationships are normal in this segment, which cuts both ways: a lock-in is protecting against something real, and it's still protecting them rather than you.

These are the ones to use on us, because our honest answers are not the strongest possible ones.

9. Show me results from a brand like mine.

Our honest answer: we can't yet. We're new, we have no permissioned client case studies, and the client-results section of our site is deliberately empty rather than filled with borrowed logos. What we can show is our own account — the self-audit we ran on ourselves, including the dead lead form and the case-study claims we killed, and our fifteen launch concepts published with their scores, losers included.

If proof from a comparable brand is what you need to move, an established agency is genuinely the safer choice and you should take it. We'd rather say that than manufacture a case study.

10. Do you manage every channel I might want?

Our honest answer: no. Creative for Meta is core; management is Google today with Meta rolling onto the same change-ledgered system. If you need hands-on management across several platforms from day one, that's a real gap and you should weigh it. We'd rather lose the deal than sell a channel we can't run properly yet.

11. How much of this is done by AI?

Our honest answer: a lot of the production. The engine drafts at volume and scores every asset 0–100 before a human sees it; a senior human who didn't make the asset signs off everything that ships, and nothing touches live spend without approval and a reversible log entry.

If your objection is to AI being involved in production at all, we're not going to talk you out of it — and an agency that tells you no AI touches your work, in 2026, is either mistaken or being careful with words. Ask anyone, including us, exactly where the human judgement sits and what it can override.

• Price on the page or a straight number on the first call.
• Creative volume as a number, checked against your replacement rate.
• A named senior person and their account load.
• You own the accounts, assets and pages — leaving is an access change.
• A readable, reversible log of every change to your account.
• A stated condition under which they'd tell you to stop spending.
• An honest gap, named without being asked. Everyone has one; only some will tell you.

And the one that isn't a question: sometimes the answer is nobody. Under about £3k/mo on ads, the maths for any agency is difficult and a founder with taste and modern tools will usually do better alone. We say so on our comparison page, where every alternative gets the case for choosing it over us.