Choose your words wisely

Hey there.

This newsletter is called Marketing That Matters. So it is probably overdue that I write an entire issue about words.

I have touched on this topic before, mostly from the angle of what not to say. The buzzword lists, the AI tells, the phrases that used to mean something and no longer do. But the more I work with early-stage companies, the more I think the real issue is not the words to avoid. It is the absence of words worth keeping. A proprietary vocabulary. Language that belongs to the company and could not have been generated by a model trained on everyone else’s.

That is now a competitive advantage. And most startups do not have it.

Let’s get to it.

Linguistic abstraction.

A few years back, I wrote briefly about a concept from Chip Wilson’s book, Little Black Stretchy Pants. Wilson is the founder of lululemon. I know Chip is a polarizing figure in the business world, but it is impossible to argue with his ability to build a brand that matters.

The concept was linguistic abstraction. He defined it as a common set of terms and definitions that all employees understand. Not quite culture. Not quite brand pillars. Not quite values. Something adjacent to all three, but more specific and more useful than any of them individually.

Here is how he described it:

“What are the five books that really define this company and really light people up? And then out of those, take 30 terms and definitions that are key to those five books and develop a linguistic abstraction, which you can grow a global company on and everybody knows and understands. It’s about speed of communication.”

Worth noting: He did not call this a brand glossary or an internal dictionary. He coined his own term for it: linguistic abstraction. That is itself a small lesson in what he is describing. He built a proprietary vocabulary for the concept of building a proprietary vocabulary. Intentional all the way down.

I treat it as the same exercise as building a brand glossary, and part of the same foundational work as mission, vision, values, and brand pillars (which all feeds into something called a Strategic Narrative). All of that work is really a language exercise at its core. You are deciding what your company believes and then finding the words to hold those beliefs. The glossary is where those words live.

Most companies fall short on one specific thing: the gap between internal language and external language. The goal is for them to be as close to identical as possible. Not in a literal, copy-paste sense. But rooted in the same vocabulary, the same beliefs, the same way of seeing the problem your company exists to solve. When that gap is large, it is usually a brand problem waiting to surface.

Nike tells the same story from a different angle.

At their best, Nike had an internal language so specific and so consistently held that three words became one of the most recognized phrases in advertising history. Just Do It was not (just) a tagline someone came up with in a brainstorm. It was the distillation of a worldview that ran through everything: how they talked about athletes, how they made products, who they chose to work with, what they refused to do. The language was consistent because the belief underneath it was consistent.

When that belief got replaced by channel strategy and ecommerce metrics, the words changed with it, and the brand followed.

The argument I keep coming back to is this: the words are usually a reflection of what is actually happening inside a company. When the thinking is sharp and the conviction is real, the language tends to show it. When the strategy loses its footing, the words go bland. The writing is the tell.

Both companies are now making deliberate moves back toward founder mindset and founder language. It will be interesting to see if the words lead the business back.

Here is why this matters more in 2026 than it did three years ago.

AI is producing the first draft of almost everything. Marketing copy, investor updates, press releases, website heroes, email campaigns. But the models are trained on the same enormous body of text — every startup playbook, every VC blog post, every tech platform website ever written. When you ask for a tagline or a company description without giving the model something proprietary to work with, it gives you the statistical center of how companies in your category already describe themselves.

Reimagining the future of work. Unlocking the potential of your team. The platform that brings it all together.

Clear thinking is now the scarce resource in marketing, not competent prose. The brands that stand out are not the ones with better prompts. They are the ones that bring something proprietary into the room before the model starts writing: a vocabulary that belongs to them, a way of talking about what they do that could not have been pulled from the same place that everyone else is drawing from.

That is the opportunity. And most startups are leaving it entirely on the table.

Here’s the exercise to get started.

There are two sides to the glossary: the words you use, and the words you do not.

The prohibit list is the one most founders find easier to start with. The short version: if AI would write it from a generic prompt, if a competitor in your space would say the exact same thing, if the words describe a category rather than a company, they do not belong in your glossary. Innovative, seamless, transformative, real-time, platform, holistic.

The affirmative list is harder and more important. These are the words and terms that are yours: words you have named, defined, and decided to own. They might be internal shorthand that has found its way into external copy, or a named methodology, or a specific way of talking about the problem your product solves that no competitor would be able to use.

A few questions worth sitting with first:

• What do we call the thing we do that nobody else calls it that? If the answer is that you use the same terminology as everyone else in your category, that is the gap, and it is worth taking seriously.
• What words do we use internally that our customers never hear? The most honest and specific language in most companies lives in Slack threads, in team meetings, in the shorthand between a founder and the early team. That language is usually the best raw material you have, and most companies never mine it.
• What would we have to believe to use the words we currently use? This one is harder, but it surfaces whether your vocabulary is genuinely connected to a worldview or just borrowed from the category because it was there and it was easy.
• What words have we stopped using because they feel wrong? The prohibit list is often already forming inside your company without anyone having written it down. Write it down.

The practical version is not complicated. It is just rarely done. Sit down with whoever knows the company best and list the words that are yours. Define each one precisely. Add the prohibit list alongside it. Then use the whole thing every time you sit down to write, or prompt. Give the model your vocabulary before you ask it for copy, and watch how different the output sounds when it has something real to work from.

Most companies treat language as the last thing to figure out — after the product and the pitch and the go-to-market — when it is actually the foundation all of those things sit on.

Your words are worth deciding.

Yours in marketing, Jeff