How many customer interviews do you need for brand research?

For an established mid-sized organization, a dozen or two well-chosen conversations usually surface the main patterns. The mix matters more than the count: recent wins, long-standing clients, lost prospects and, if the strategy targets a new market, buyers who do not know you yet.

Should we use focus groups to choose a new logo?

Generally not. Groups drift toward the familiar and the opinion of whoever speaks first, and people judge abstract marks poorly. Test identity directions in realistic applications with individual buyers, and ask what the work communicates rather than whether they like it.

Can we do brand research ourselves?

You can gather a lot of it internally, especially existing evidence such as reviews, sales notes and search terms. Customer interviews are the part that benefits most from an outside interviewer, because clients are more candid with someone who is not their long-time contact.

How long does brand research take?

For most established organizations, a few weeks: time to book and run interviews, gather existing evidence and write the findings. The scheduling of customer conversations, especially with lost prospects, is usually what sets the pace.

How to Run Brand Research Before Changing Your Identity

What the research is for

Brand research exists to answer a small number of questions that determine everything after it: who the brand is really for, why those people choose you, what nearly stops them, what they compare you to, and which parts of the current brand they recognize and value. If the research does not change or confirm a decision, it was a ritual.

So start by writing down the decisions it has to inform. For example: should we keep the name; are we positioning for the commercial clients or the residential ones; does the acquired practice stay separate; can we drop the old tagline. Each research activity should map to at least one of those. This also keeps the research proportionate. A refresh needs less than a rebrand (the difference is covered in brand refresh versus rebrand), and a team that cannot yet say which one it needs should let the research decide.

A practical scale for an established mid-sized organization: a few weeks, a dozen or two conversations, and a findings document short enough to be read in one sitting. That is enough to change minds. Much more than that and the research starts to become a way to delay the decision.

Start inside the building

Internal interviews come first, for two reasons: they surface what leadership assumes, which the external research will then test, and they find the people who know things leadership does not.

Leadership

Interview each senior leader separately. Group sessions produce agreement with whoever speaks first. Ask the same questions of everyone so the differences are visible:

Who is our best customer, and why do they choose us?
Who do we lose to, and why?
What will the business look like in five years that it does not look like now?
What about our current brand embarrasses you, and what would you fight to keep?

The disagreements are the finding. If the partners describe three different ideal clients, that is the most important thing the project will learn.

The front line

Then talk to the people who sell, quote, answer the phone and deliver the work. They hear objections leadership never sees, and they know which materials they quietly avoid using. Ask them what customers say in the first five minutes of a conversation, what questions come up again and again, and what they find themselves explaining that the brand should have explained already.

Talk to customers, including the ones you lost

This is the heart of the research and the step most often skipped, usually because it is uncomfortable. It is also where the leadership assumptions get tested.

Aim for a deliberate mix:

Recent wins, to learn what tipped the decision while it is fresh.
Long-standing clients, to learn what keeps them and how they describe you to others.
Lost prospects, who chose someone else. They are the most valuable and the hardest to book.
Former clients, if any left on reasonable terms.
Target buyers who do not know you, if the new strategy aims at a market you do not yet serve.

Conversations of twenty to thirty minutes are enough. Ask about their situation and decision, not about your brand directly: what were you trying to solve, what else did you consider, what made you choose, what almost made you choose differently, how would you describe us to a colleague. The brand insight comes out of the answers. Asking “what do you think of our logo” produces politeness, not information.

Who asks matters. In a small market, a client talking to the owner they have known for fifteen years will be kind rather than candid. Someone outside the company, whether a consultant or a trusted person from another part of the business, gets more honest answers. In New Brunswick, offer interviews in French as well as English; a francophone client answering in their second language gives you a thinner, more cautious version of what they think.

Use the evidence you already have

Most organizations sit on a pile of customer evidence nobody has read as a whole. Collect it before the interviews end, because it helps you spot patterns and check whether what you heard is typical.

Reviews on Google and industry sites. Read the words customers use, not the star ratings. Repeated phrases are often your real positioning.
Sales and CRM notes: reasons recorded for wins and losses, if anyone records them.
Proposal and RFP debriefs. Public bodies in Atlantic Canada will often give a debrief on request after a procurement; it is some of the most direct feedback a firm can get on how it is perceived against competitors.
Inbound inquiries: how prospects describe what they need in contact forms and first emails.
Search terms from Search Console and your ad accounts, which show how people describe the problem before they know your name.
AI assistant answers: ask a couple of the major assistants to describe your organization and recommend options in your category. The answer is a rough mirror of how the web describes you.
Your own materials: a year of proposals, ads, social posts, signage and the website, laid out side by side. The inconsistencies are themselves a finding.

Map the alternatives buyers actually weigh

Competitive research for a brand is not a list of every company in your category with their logos lined up. It is a map of the options your buyers actually consider, which the interviews will have told you. Often that list includes a national firm, a cheaper local option, an in-house solution and doing nothing.

For each real alternative, capture what they claim, what visual and verbal territory they occupy, and what buyers said about them. Then look for the gap: a claim that matters to buyers, that you can credibly make, and that nobody else is making well. That is the raw material for positioning, which the positioning framework for competitive markets turns into a structured statement. Our broader competitive analysis framework covers the marketing side of the same exercise.

Also note the visual conventions of the category. If every competitor uses navy and a stock photo of a handshake, the convention tells you what buyers expect and what would stand out. Whether to follow it or break it is a later decision, but you need to see it first.

Synthesize into findings, not a data dump

The output of research is a short document that leadership reads and argues with. A useful structure:

What we heard: the recurring themes, with real quotes (anonymized) under each.
Where we agree with ourselves: the internal beliefs the customer evidence confirmed.
Where we were wrong: the internal beliefs it contradicted. This section earns the research its budget.
The gap: the opportunity in the competitive map.
Equity to keep: the names, phrases, colours and cues customers clearly recognize.
Implications: what this means for each decision listed at the start.

Resist counting. Twenty interviews are not a statistically meaningful sample, and presenting them as percentages implies a precision they do not have. Patterns, strong signals and surprising single comments are what qualitative research gives you. Treat them as such.

Validate before you commit

Once the findings produce a draft positioning and, later, identity directions, test them before they become permanent. Validation is not asking a panel which logo they prefer. People are poor judges of abstract marks and good judges of whether something makes sense to them in context.

Test the words first. Put two or three versions of the positioning and core message in front of a handful of target buyers. Can they repeat back what you do and why you are different? Does any version feel untrue given what they know of you?
Test in real applications. Show identity directions on a proposal cover, a website header, a sign or a truck, not on a white board. Ask what kind of organization it suggests, and whether it looks like it belongs among the alternatives they would consider.
Test with staff. Could the front-line team explain the new positioning to a customer without a script? If not, it is too abstract.
Check the practical constraints. Does it work in French and English, at small sizes, in one colour, on a vehicle at a distance?

Validation should be able to change the outcome. If the team has already decided and the testing is for show, skip it and save everyone the time. Afterwards, the same evidence becomes the baseline for measuring whether the positioning is working.

In our brand strategy and identity engagements, nothing visual is shown until the findings and positioning have been agreed in writing. That is less about process purity than about saving clients from approving a look before they have agreed what it is supposed to say. If you are hiring outside help for this stage, how to evaluate a branding agency lists the research questions worth asking in the pitch.