The Market for Words

I spent years building in 3D technology and VR, which means I watched "the metaverse" get minted, watched it peak, and watched it die. I watched investors who could not describe a single use case say "spatial computing" with total confidence, and saw decks in every sector add a metaverse slide.

Then I watched the same firms delete the word, the way you exit a position you would rather not discuss.

Language is a market.

Words get minted by people doing something, they appreciate as the association spreads, and then they go to zero once everyone owns them.

A word can be useful while it makes you sound different, but every firm that copies it makes you sound less different. Once everyone says it, the word does the opposite of its job - it makes you sound like everyone.

In venture, sounding like everyone is expensive, because your words are how you get priced before your returns exist. LPs + founders read them to decide whether you are worth a meeting.

You already know this curve. Venture capital firms exist to sell their ability to recognise value before the market does. The words you use to describe yourself deserve the same discipline.

So this week… how to avoid buying your vocabulary at the top, and where the mispriced words are now.

This is the market for words.

Your words have value

In the most measurable sense, this is clearest in the public markets.

In the famous “A Rose.com by Any Other Name” paper, Cooper documented abnormal returns of 74% in the 10 days around companies announcing Internet-related name changes during 1998-99. The announcement effect was similar regardless of the firm's actual level of Internet involvement. In their study window, they found no subsequent reversal. The association could pay without much substance behind it.

The rename trade never dies, it just updates the word.

Long Island Iced Tea became Long Blockchain, +289% in a day in 2019 and Allbirds became NewBird AI, +582% in 2026.

The same applies to funds, mutual funds that renamed towards hot investment styles attracted 28% abnormal inflows over the following year, without improved performance.

ESG hype shows the full round trip, adding an ESG term to a fund's name was worth an estimated 8.9% in extra inflows in the first year. Then the cycle turned - nearly 1,600 funds holding over $1 trillion have renamed, and by late 2025 most renamers had removed their ESG words.

Additionally in the majority of cases, the portfolio was unchanged while the name traded around it.

Words get bought and sold. The uncomfortable finding is that it works - the fashionable name kept attracting money long after it stopped describing anything.

In venture, the same trade runs quieter. There is no share price to spike and no audit to fail, and, worst of all for firms, the only people who can see the trade happening are the insiders they most need to impress.

The clearest tell is the trend-hopping. The same firms went from Web3 experts to AI experts without breaking stride, and then, “oh no, we really meant physical AI”, a la robotics, and the 100 other iterations of the same memes.

It’s hardly surprising, VCs are parroting these lines not because they believe it, but because (like e.g. .com and ESG before it) it sells to LPs and makes it easier to move dollars.

1 of the largest firms in VC even calls itself a “full-stack AI transformation company”.

Sometimes it is subtle

Emergence Capital's services essay and playbook now use "AI-native" where earlier archived versions used "AI-enabled".

The boom in "services are the new software" makes that distinction commercially attractive. An AI-assisted services business and a business built around AI from the ground up invite different expectations.

Sure, this is okay. Firms should develop their thinking, and tbf Emergence was publishing about machine-learning businesses back in 2017.

BUT the assumption is that the label changed, because this is the phrase that the market was paying for.

The problem is while an LP considering AI alongside other investment categories may find "AI-native" a useful way to place the fund. A founder who has spent years inside the problem is going to want more.

I mentioned my 3D tech background, when firms today write about spatial computing or robotics, I can usually distinguish familiarity with the vocabulary from depth in the subject within a paragraph.

I suspect most specialists recognise this inside their own sector. The phrase you chose to reassure a generalist audience can be exactly the one that exposes you to the expert one… and founders are always the experts.

Your words reveals when you arrived

You can watch these conversations move through the language.

Using the X api and twexapi.io you can track phrases across AI, infrastructure and investing. They give us a record of the conversation, which is a small window into investors adopting a thesis.

To a generalist you might sound current. But to someone who has actually built things and you may you sound like you just arrived.

Take taste. Paul Graham was writing seriously about it in 2002. Two decades later, "taste is the moat" became a claim of commercial defensibility.

Today, if you are tweeting "taste is the new moat!", well, you look like someone's exit liquidity. We even have counter trades of “taste is not a moat”…

That does not make taste worthless. It makes repeating the sentence a poor substitute for explaining decisions that ‘thesis’ produced. Talking about ‘taste’ is free vs having it costs deals. PG’s essay predates YC by 3 years, and "very exacting taste, plus the ability to gratify it" became an investment filter he then ran money through.

Your words can credit someone else

Likewise, "Forward Deployed X" is now part of the AI pitch for VCs and their accomplices.

Palantir had made forward-deployed engineering an institutional role in a recruitment posting from September 2013. Now AI companies have an implementation problem, the role is fashionable again, and the category gets explained through Palantir. a16z's essay on this trend 13 years later is literally titled The Palantirization of Everything.

Every firm that says forward deployed is doing free marketing for Palantir. You borrowed the word to explain your thesis, by strengthening theirs.

Likewise, I have written before about European VCs adopting Silicon Valley's nationalist marketing playbook to argue for independence from Silicon Valley. The broader behaviour applies to more than Europe.

American Dynamism is the venture-native version.

Katherine Boyle used the phrase in a May 2021 General Catalyst essay. a16z then hired her, announced a named practice, and in April 2024 launched a $600 million American Dynamism fund.

Today other funds explain their own positions by reference to it. In April 2025, Karman Ventures described its thesis as aligning with a16z's American Dynamism and celebrated five portfolio companies appearing on a16z's list.

Karman already invested here but reached for a16z's frame to explain them.

Likewise Earth's On Climate Dynamism explicitly refers to a16z's definition while making its own climate argument.

Every one of those references is a small advert for a16z - borrowed word also arrives with baggage the borrower does not choose. The UK now has a Dynamism Fund whose remit covers talent, science and institutions - explicitly not defence related.

BUT as the market mostly learnt the word through a16z's defence and reindustrialisation arguments, you may bring those associations anyway. a16z's own remit is broader than defence, but the shorthand can be narrower than the thesis, which is a problem for everyone else touting all flavours of “dynamism”.

Different buyers, different reasons

For an emerging manager words can buy admission. Familiar vocabulary makes an unfamiliar team legible to an allocator: a recognised market, a category to be placed in, a strategy the reader has heard of. The risk is becoming legible as an interchangeable manager.

Likewise, an established firm can buy currency: a new phrase makes an old organisation look present-tense, or gives new work a category.

There are 2 ways to make a word pay. Be early, or go all in.

Air Street was early. It called its founders and investors "AI-native" in October 2020 and has since gone on to raise the largest solo GP fund in Europe ($232M). The word was cheap when they bought it.

Emergence arrived late but went all in, and it is working. Yes, the same Emergence whose label swap we showed - what has followed is the commitment. The phrase is now in all their essays, their playbook, their positioning, repeated until firm and phrase travel together, the same play I wrote that EssenceVC ran with "infrastructure."

That is the test the trend-hoppers fail.

Changing your mind is fine. Investors are supposed to change their minds. What breaks a firm is changing the words and hiding the change - deleting the crypto thesis, publishing the AI one, and hoping nobody remembers.

Say what changed your mind. "We spent 3 years in VR and here is what it taught us about robotics" is a stronger position than pretending the VR years never happened.

Over the last year or so we've worked with 2 firms going through a thesis repositioning and we're really pleased with the results (aka deals and LP $$).

When the market is long jargon

Sometimes the opportunity is in the ordinary words that professional language has made unfashionable.

Venture firms almost never say "money." They say capital, and they deploy it, because deploying capital sounds strategic and handing over money sounds like what it is. "Backing founders" does a different job - it moves the spotlight from the money to the relationship. None of these is dishonest. But each one steers your eye away from the transaction.

Which is why plain speech now stands out.

Lowercarbon's homepage talks about companies that "make real money". Ribbit's archived homepage put it more directly "It takes money to change money."

A venture firm is allowed to say what the money is for!!

“Conviction” deserves the same treatment… "We have strong beliefs" is a less glamorous claim, but it would make you notice how little has been said. Which beliefs? What decisions follow? How would a founder experience the difference?

I looked for 6 expressions in homepage text for the same 47 firms in 2016, 2020 and 2026. Conviction appeared on 4 firms' pages in 2016 and 14 in 2026!!

If these are just words and not tied to evidence, imo they are super weak…

e.g. the 2016 CRV homepage, conviction is beside a concrete claim about often reaching a funding decision within a week. The word has something to answer for. That gives a founder something concrete to evaluate.

Likewise, “asymmetric opportunities” can describe a real investment structure. But the prestige of the expression spares a writer from explaining whether the edge is entry price, limited downside, exceptional upside or something else…

Plain language forces that choice. 

Replacing every technical term with a plain synonym would miss the point. But playing with words is a way to find out what you really mean + what you are prepared to defend.

Before you buy the word

Take the paragraph your firm uses to explain itself. The homepage/deck/the thing a partner says on a call.

For each fashionable word, write the belief or behaviour it is supposed to describe.

• If you have conviction, what do you do that another investor would not?

• If you are AI-native, where does that show up in the team/portfolio/how you assess a company?

Then ask who your words makes the reader think of. If the answer is another firm, every time you say it, you are marketing for them.

None of this means banning words. It means making each one answer for something. Air Street can say "AI-native" because 2020 backs it up. Bedrock can say "narrative violations" because they minted it. A word is not a problem because it is fashionable - it is a problem when there is no work behind it.

Your vocabulary is a portfolio.

Laurie, Refinery Media

If you made it all the way through, thanks so much for reading. A thousand VCs now open this every week. If it helped you think differently about marketing, Venture, or storytelling, please send it to someone in your orbit.

Read more from our top posts:

• AI Visibility for VC Firms

• What Should your GP be posting?

• What Would it Look Like if Bechmark Did Marketing