How to Choose a Branding Agency (and Spot the Red Flags)

Most selection processes evaluate the wrong thing. A portfolio shows you work that was approved by a different client with a different problem, filtered to show only the successes, and photographed well. It is close to useless as a predictor.

Here is what predicts better.

Judge the questions, not the pictures

The strongest signal in any first conversation is what they ask you. Agencies that will do good work interrogate the business before discussing the deliverable: what does a customer cost you, what is your close rate, who do you lose to and why, what have you already tried.

Agencies that will not do good work start talking about their process, their awards, and how they would approach the design.

If nobody asks you an uncomfortable question in the first meeting, you have met a supplier, not a partner.

Ask who is actually doing the work

The pitch team and the delivery team are frequently different people. This is not automatically bad — but you should know before signing, not in week two.

Ask directly: who will be on this, what percentage of their time, and who is the most senior person who will touch it after the kickoff? Get the answer in the contract if the answer matters to you.

Read the case studies backwards

Skip the visuals. Look for what the client's problem was and what changed afterwards. A case study that describes a "bold new visual language" and never mentions a business outcome is a portfolio piece, not a case.

Then ask for the one they cannot publish. Every experienced agency has projects under NDA and projects that went sideways. How they talk about those is the most information you will get in the whole process.

The six red flags

1. A guaranteed outcome. Nobody can promise a position on a search results page or a percentage lift before understanding your market. Certainty offered this early is a sales technique.
2. No questions about budget or timeline, ever. Either they will discover the mismatch late, or they intend to.
3. The proposal arrives within a day, fully formed. It was written before they understood you, which means it was written for someone else.
4. Design begins before strategy is agreed. See why this order matters.
5. They cannot name what they would not do. An agency that takes every kind of client has no position of its own, which should worry you given what you are hiring them for.
6. Ownership is vague. Who owns the files, the fonts, the source code and the domain when this ends? If the contract does not say, assume the answer is not you.

Why the cheapest and the most expensive proposals often fail the same way

Both tend to be scoped without understanding. The cheap one wins by excluding the parts that are hard to see — copy, migration, post-launch work, the real cost drivers — and recovers the margin in change requests. The expensive one bundles everything anyone might need, including things you do not, because nobody asked enough questions to know.

The proposal you want is the one that reflects a conversation. It should contain something you did not ask for and an explanation of why, and something you did ask for that they have argued against.

Run a paid first step

The best de-risking mechanism available: hire for a small, paid, defined first phase — a strategy sprint, an audit, a single page — before committing to the full project.

You learn how they work, how they handle feedback, whether they hit dates, and whether the senior person stays involved. They learn your business well enough to scope the rest honestly. It costs a fraction of the project and prevents the expensive version of the same discovery.

And before you brief anyone, write the brief properly. Half of what looks like a bad agency in month three was an unclear brief in week one.