It's about time you stopped using real-time

Hi there,

Something has shifted in startup marketing over the last couple of years, and I don’t think it’s fully been named yet.

AI didn’t create bad marketing writing. That was already well underway. The paradigm shifts, the reimaginings, the relentless revolutionary-izing of completely ordinary software products. That slop existed long before anyone had a ChatGPT or Claude account. What AI did was take every bad habit already present in the corpus of startup marketing copy and accelerate it by a factor that’s hard to even quantify.

Faster, cleaner, more confident, and somehow even more empty.

The tells have changed too. I’ve written about the em dash. I’ve written about “it’s not an X, it’s a Y.” The models keep evolving, the tells keep shifting, and the underlying problem stays exactly the same: words that were already losing meaning are now being drained completely.

Psychologists call it semantic satiation. Say a word enough times and the brain stops processing it as meaningful. It becomes noise. “Real-time” got there a while ago. AI just made sure nobody noticed until it was too late.

Today’s subject: two words that appear in almost every deck, every hero, every press release. What they mean, where they came from, and why the only real move left is to get specific or get out.

Let’s get to it.

A quick note on grammar

“Real-time” is hyphenated when it modifies a noun: a real-time dashboard, a real-time alert. No hyphen when it functions as a noun or adverb: the data updates in real time.

Where the word came from

Open almost any startup website right now. Read the hero. Scroll the pitch deck you got last week. Skim the press release in your inbox.

Somewhere in there, guaranteed: “real-time.”

Real-time data. Real-time alerts. Real-time visibility. Real-time insights.

The phrase has a precise origin. In computing, real-time meant something specific and demanding: a system capable of processing input and responding within a defined, critical time window. Milliseconds. The kind of constraint that matters in aircraft navigation. Medical devices. Financial trade execution. Things where the gap between input and response has consequences.

That’s a meaningful claim. The words earned their place.

Then they moved into marketing. And the way words tend to work in startup marketing, the phrase got borrowed by people whose products didn’t quite meet the original definition. Then by people who were further still… and then by everyone. Now it shows up in decks for platforms that sync overnight and websites for tools that pull weekly reports.

That’s not a new phenomenon. That’s just what happens to useful words when enough people use them without earning them. “Disruptive.” “Innovative.” “Platform.” “Real-time” is the latest addition to the list, except with one additional problem the others don’t have.

It sounds precise while being completely vague, which is worse than just being vague.

“Innovative” is obviously empty. Nobody reads it and assigns meaning. The brain filters it out before it even registers. That’s annoying, but mostly harmless.

“Real-time” still feels like a technical specification. It carries the ghost of something concrete. So the reader processes it for a half-second as if it means something, and then it dissolves. That half-second is the problem. You’ve made someone do work for nothing.

An investor who has read two hundred decks knows what “real-time analytics” means. It means: we have a dashboard. A journalist who covers health tech has seen “real-time patient data” enough times that it tells them nothing about what your product actually does.

Credibility doesn’t come from sounding technical. It comes from being specific enough that someone can check you.

The rule

If you’re going to say “real-time,” answer this question first: real-time means how fast, exactly?

If the answer is milliseconds, or seconds, or something your product genuinely delivers and your competitors don’t, say that. “Alerts within 30 seconds.” “Data that updates every five minutes.” “A live feed.” Those are claims. They’re checkable. They differentiate.

If the answer is “same day” or “pretty quickly” or “well, nightly, but faster than the old way,” then you don’t have a real-time product. You have a fast product, or a frequent product, or a more-current-than-before product — so say that instead. None of those are embarrassing, and several of them are genuinely useful. All of them are more credible than “real-time” because they’re honest.

And if you reach for the phrase and you can’t answer the question, that’s your signal. Drop it. Describe what you actually do.

The irony of the AI era is that the bar for standing out has never been lower on effort and never higher on specificity. The models are very good at producing confident-sounding marketing language. What they’re not good at is knowing what your product actually does in a way that’s true, precise, and different from everyone else’s.

That part still requires a human. Probably you.

“Real-time” had a good run. It earned its meaning once. We’ve been spending it down ever since.

Get specific, or stop using it.

Yours in marketing, Jeff

(For the record: I have almost certainly used “real-time” in a client deliverable. This series is as much about me trying to get better as anything else.)