We were interrogated for thirty-six minutes in June by a consultant running between €50,000 and €80,000 a month in ad spend for a beauty brand. He never asked about our results. He asked about our process, nine times, in increasing detail, and every question landed somewhere that would have been awkward if we had not had an answer.
That call is the best guide to choosing an agency we have seen, because it was conducted by someone who already knew what he was buying. This article is his questions, roughly in his order, with what a good answer sounds like and what a bad one is hiding.
We are an agency. Read this with that in mind. It is also why we can tell you what the questions are actually testing.
1. "What is your creative process? Where do you start?"
He asked this first, before results, before pricing, before anything. That ordering is correct and almost nobody uses it.
What it tests is whether creative production is driven by a repeatable method or by whoever is available that week. The tell in a bad answer is that it describes outputs, not inputs. "We produce concepts based on your brand guidelines and best practices" is a description of a deliverable, not a process.
A real answer names its inputs and its sequence. Ours starts with data you already have and mostly ignore: post-purchase survey responses, review text, support tickets, clustered to find out how your customers actually describe the problem. Then a backward read of the ad account to find what already has signal and what has already failed. Then a diversity map of what the account is currently covering and what it is not. Concepts come last, and they come from the gaps. Our own version of that map is described in what creative diversity actually means.
Ask the follow-up: "Show me a brief you wrote for another client, with the client details removed." A process that exists produces artefacts. A process that does not produces adjectives.
2. "Where do your responsibilities stop? Do you write the video briefs?"
This is the sharpest question on the list and it exposes more agencies than any other.
There is a common shape where the agency handles media buying and strategy, and creative is "coordinated" with your team or a production partner. Every additional layer between the person reading the performance data and the person making the asset is a translation loss. You brief a brief, and what comes back is seventy percent of what was meant.
The honest answer is that if a service wants to be fast and efficient, it has to cover the brief. Not necessarily the filming, not necessarily every hour of production, but the document that turns a data observation into a creative instruction. That cannot be outsourced without losing the thing that made it worth doing.
Ask the follow-up: "Who physically writes the script for a video ad, and who do they report to?" If the answer involves three parties, expect three-party results.
3. "Is there a limit on how many creatives you produce? Is it tiered?"
He was probing for a rate card dressed as a strategy. The bad version of this business sells you fifteen assets a month because fifteen is what the tier says, regardless of whether your spend can evaluate fifteen assets.
A good answer inverts the question. Volume should be derived from your account: what your historical creative lifespan looks like, what your win rate is, whether the constraint is market speed or a technical problem in the account, and how much budget each new concept needs before it produces a readable result. Then the number falls out of the maths.
The practical trap on the buyer's side is assuming more is better. A creative that never accumulates enough impressions to be evaluated is not a test, it is a donation. At smaller budgets, fewer and genuinely more distinct concepts beat higher volume every time, which is the argument we make at length in creative volume without intent.
Ask the follow-up: "At my spend, how many concepts can we actually evaluate per month?" If they cannot answer with arithmetic, volume is being sold to you as a feature.
4. "How does reporting work? Weekly? Shared sheets?"
Reporting cadence is a proxy for whether the agency understands your scale.
If you are spending modestly, a weekly call produces a weekly conversation with nothing in it, and both sides start performing progress. If you are running serious volume, monthly is too slow to catch a problem while it is still cheap.
The answer we give: dashboards always, with the tool chosen to match what already exists in your stack rather than adding one. Weekly written reporting as standard. A short weekly call where volume justifies it, structured as a retrospective plus what is being tested next, not as a metrics recital you could have read yourself. Monthly, the longer analytical review.
Ask the follow-up: "Does the report cover creative, or only media?" A media-only report from an agency that claims to run creative strategy tells you which half they actually operate.
5. "What does onboarding look like, and how long until you are operational?"
He was switching agencies the following month, mid-year, which is the situation most of these conversations happen in.
The honest range is wide and the reason matters. A pure handover of an existing, functioning account is roughly two weeks to fully operational, sometimes less. Building creative capability from zero is a different project with a different timeline, and any agency quoting you two weeks for that is quoting you the easy half.
The risk in a mid-season switch is not the setup. It is the learning that lives in the outgoing team's head and never gets written down. Ask the outgoing agency for their testing history and their creative performance data, in writing, before you give notice. After notice, the quality of that handover is entirely dependent on goodwill you no longer have leverage over.
Ask the follow-up: "What do you need from my current agency, and what happens if we do not get it?" A team that has done this before has a list.
6. "Do I really have to give you ad account access, or can I send screen recordings?"
He asked this directly, and the hesitation behind it is completely reasonable. Handing a stranger access to your business manager is uncomfortable, and it exposes to your internal team that you are shopping.
The honest answer is that access makes the work faster but is not always mandatory for a first analysis. What is mandatory is data that links the quantitative to the creative. A purely quantitative export with ad IDs and no way to see the assets is the most common gap, and it makes creative analysis impossible. If you can produce a report that connects spend and performance to the actual creative, that can be enough to start.
Ask the follow-up: "What is the minimum you need to do the analysis, and what do you lose without full access?" An agency that says access is non-negotiable for a first look is either inflexible or does not have a real analysis method.
7. "Have you worked in my category before?"
This came from a different call, a US apparel brand, and it is the oldest question in the category.
The honest answer is that vertical experience matters much less than people believe, with two real exceptions. Regulated categories, where the constraint is compliance, and categories with genuinely unusual purchase mechanics, like considered B2B or highly seasonal hardgoods. Outside those, the transferable skill is diagnosis, and the category-specific knowledge is a few weeks of reading your reviews and your support inbox.
What actually predicts fit is spend range and business model. A brand with first-order economics that only work on repeat purchase needs a very different engagement from one with a 60% margin and a single-purchase product. An agency whose entire experience is at €5,000 a month will not have seen the problems that appear at €80,000, and the reverse is also true: teams built for large accounts often cannot operate economically at small ones.
Ask the follow-up: "What is the smallest and largest monthly spend you currently run?" That answer tells you more about fit than a client logo ever will.
8. "How do you price, and can you give me a number now?"
Two prospects out of three ask us for a number on the first call and we do not give a full quote, because scope genuinely varies. That is true and it is also, from the buyer's side, only half acceptable.
The best version of this objection came from a UK founder: "It would be really helpful to have some indicative starting points, because what I do not want to do is waste your time and then get the quote and go, whoa, that is way out of my budget." She is right. Refusing to give any range at all protects the agency, not the client.
So here are the ranges we see in the European and US market for DTC brands, stated as ranges and not as anybody's rate card.
- Media management, monthly retainer. Roughly €2,000 to €4,000 a month for a small single-channel account, €4,000 to €8,000 for a multi-channel account at meaningful spend, €8,000 and up where creative strategy and production are included. Senior operator time is the cost driver, not the media budget. For how the media budget itself should be sized, separately from the fee, see our ecommerce marketing budget guide.
- Creative production, per batch or per month. Commonly quoted per asset batch. Entry-level batches of a dozen or so mixed static and video assets start in the low hundreds and rise sharply with production complexity. Fully produced video with talent is a different economy entirely.
- Project fees. For a defined build such as a funnel rework or a launch, quoted as a project rather than a retainer.
On the model itself, one point is worth being direct about. A fee calculated as a percentage of ad spend puts your agency's revenue in direct conflict with your interests, because their incentive is for you to spend more, whether or not more spend is the right call. A flat monthly fee that does not move with your budget removes that conflict. It means an agency that recommends cutting spend takes a hit to results and none to revenue, which is the correct arrangement.
Milestone components are a reasonable middle ground: a base fee, with an uplift unlocked at an agreed performance threshold.
Ask the follow-up: "Does your fee change if I double my budget?" If it does, you now know what advice you will be getting.
9. "Do you offer a trial? What are the terms?"
Reasonable question, and the answer separates confidence from lock-in.
Look for two things in the contract. A notice period you can live with, thirty days being common and reasonable for creative services, longer being defensible only where the agency is carrying real fixed cost on your behalf. And clear ownership of everything produced: creative files, ad account structures, audiences, and data. Assets you paid for should leave with you.
The free audit that most agencies offer, including us, is the real trial. It is not a favour. It is a paid-in-kind sales asset, and both sides know it. That is fine as long as it produces something you can use even if you never sign.
Ask the follow-up: "If I take the audit and do not hire you, do I keep it?" The answer should be yes without hesitation.
What a good free audit looks like
Since it is the most common entry point, it deserves its own standard. A useful audit shows you the reasoning, not just the findings. It should tell you something you did not already know about your own account, quantify the opportunity rather than assert it, and be specific enough that you could hand it to your current agency and they could act on it.
An audit that is thirty slides of platform benchmarks and one slide of recommendations is a pitch deck. An audit that opens with your own data and ends with an argument you can disagree with is analysis.
Key takeaways
- Ask about process before results. Case studies are selected. Process is structural.
- The scope question is the one that exposes most agencies: find out who physically writes the creative brief, and how many parties sit between the performance data and the asset.
- Creative volume should be derived from your spend and your account's win rate, not from a service tier.
- A fee tied to a percentage of ad spend creates a conflict of interest. A flat fee that does not move with budget removes it.
- Any agency should be able to give you an indicative range on a first call. Refusing entirely protects them, not you.
If you are running this evaluation right now, start with the audit. We will show you the reasoning and you keep it either way. Book a discovery call
