The Brand Voice Process We Use With Clients (And Why It's Worth the Time)

Most brand voice exercises fail for a boring reason: everyone in the room likes every nice word. Ask a leadership team whether they want to be seen as innovative, customer-focused, and trustworthy, and you'll get three enthusiastic yeses and zero actual decisions. Skip the hard part, and you end up with a brand voice that sounds like every other company's About page, because that's exactly how it was built.

If you're a CEO deciding whether this is worth your team's time, here's the honest answer: it's worth it if you treat it like a real decision with real tradeoffs, and it's a waste of a week if you treat it like a wordsmithing exercise. The difference is the process.

Watch the full recording below if you want to see this run live on a real client.

Step 1: Gather real signal, not opinions

Before anyone picks a single word, pull together the raw material: customer interviews, sales calls, founder conversations, and a few pieces of writing your team genuinely admires. A customer interview outranks a blog post you found online. A founder explaining, unscripted, how they actually think about the business outranks almost everything else. This is the human signal that no AI tool can manufacture for you, and skipping it is the single biggest reason brand voice projects end up generic.

Step 2: Distill it into real tensions, not a word cloud

Cluster the raw material into themes, and prune the words every competitor already claims, like innovative, cutting-edge, and customer-centric. More importantly, surface the tensions instead of averaging them away. Take "innovative." Does that mean you build exactly what customers ask for because you listen closely? Or does it mean you push back on what they ask for because you believe you know a better way? Both are legitimate positions for a company to take. They lead to very different brand voices, and somebody has to actually choose.

Step 3: Force a vote, don't take a survey

Don't ask your team to rate each word on its own; everyone rates every nice word highly, and you learn nothing. Force a choice instead: given these four words, which is most like us, and which is least? Repeat it across the shortlist. This produces an actual ranking instead of a pile of ties, and it works better as an anonymous exercise than a conference room discussion, since nobody has to defend their answer out loud in front of the CEO.

Step 4: Turn the results into something usable

The output should be a handful of paired statements, not fifty adjectives: direct, not blunt. Confident, not hyped. Pragmatic, not dry. The second half of each pair matters as much as the first. It tells anyone writing for you exactly where the line is, which is the part most brand guidelines skip entirely.

Step 5: Lock it

This is the step people skip, and it's the one that makes everything before it worth doing. An unlocked voice guide is a document nobody reads twice. A locked one becomes the actual instruction set behind every email, landing page, and ad from that point forward, whether a person or an AI tool is drafting it.

What this buys you

Skip the process and you'll end up with the same generic list every other company has. Run it properly and you get something genuinely hard to copy, because it wasn't built from a prompt. It was built from what your customers actually said and the calls your team actually made.

Before you spend a week of your team's time on this, it's worth knowing whether your ICP is actually locked enough to make it pay off. Find out in the T2D3 Growth Workshop.