The one-sentence company description that saved the pitch

Why the generic one-liner fails

The "AI-powered platform for X" sentence fails for a specific reason, and it is not that it is boring. It fails because it gives the listener nothing to hold. There is no customer they can picture, no task they can imagine being done badly today, no before-and-after. The sentence describes what you built from the inside, in the language you use with your engineers, and asks the investor to do the translation work themselves.

They will not do it. Not because they are lazy, but because they hear forty pitches a week and the founder who makes them translate loses to the founder who hands them the finished thought. Worse, a vague sentence forces the investor to reconstruct your company from the questions that follow, which means the first impression of your business is "this person could not explain it simply." Fair or not, that sticks.

There is a deeper failure too. The generic sentence is unrepeatable. Venture decisions do not happen in your meeting. They happen on Monday, in a partner meeting, when the person you pitched tries to describe you to four colleagues who were not there. If your sentence cannot make that trip intact, your company gets re-described by someone with less conviction and less information than you, and that watered-down version is what gets voted on. You are not writing a sentence to impress the person in the room. You are writing a sentence for the person who has to defend you when you are gone.

The formula

A company sentence that travels has three parts, in this order: the customer, the painful workflow, the changed outcome.

The customer is a specific person, not a category. "Logistics teams" is a category. "Freight brokers" is a person you can picture at a desk. The more specific, the more the listener trusts that you know who you serve.

The painful workflow is what that person does today that is slow, manual, error-prone, or humiliating. This is the part founders skip, and it is the most important part, because pain is what makes the outcome mean something. A workflow the listener can wince at does more work than any adjective.

The changed outcome is what is different after you exist. Not your features. The result in the customer's day or their numbers. "Covers twice the loads before lunch" is an outcome. "Streamlines their workflow" is a feature wearing an outcome's clothes.

The structure, then:

[Specific customer] used to [painful workflow]. Now they [changed outcome], because we [the one thing you do].

You do not have to say it in exactly that grammar. But every strong company sentence contains those four facts, and every weak one is missing at least two of them. Read your current one-liner and check which of the four are present. Usually it is just the last clause, the "what we do," floating with no customer and no pain attached to it.

Five rewrites

Here is the same move across five company types. The bad version on the left is the real default most founders reach for. The good version names the customer, the workflow, and the outcome.

B2B SaaS

Bad: "We're a collaboration platform that helps teams work better together."

Good: "Clinical trial coordinators track patient visits across spreadsheets, email, and a paper binder. We put all three in one place, so a coordinator can run forty patients without missing a visit window."

AI

Bad: "We're an AI-powered platform for customer support automation."

Good: "Support agents at insurance companies retype the same policy answers fifty times a day. We draft the answer from the customer's actual policy, so the agent edits instead of writes and closes tickets in a third of the time."

Marketplace

Bad: "We're the marketplace connecting buyers and sellers of industrial equipment."

Good: "Factory owners selling used machines list them in Facebook groups and wait weeks for a tire-kicker. We get a verified offer from a vetted buyer in 48 hours, so the machine sells before the floor space is needed."

Infrastructure

Bad: "We provide a scalable data infrastructure layer for modern engineering teams."

Good: "Data engineers spend their first hour every morning checking whether last night's pipelines broke. We catch the break and tell them what broke and why, so the day starts with a fix instead of a search."

Consumer

Bad: "We're a fintech app that helps people manage their money smarter."

Good: "People with irregular freelance income never know how much they can safely spend this week. We do that math from their actual account balance and upcoming bills, so they see one number they can trust."

Notice what every good version refuses to do. It does not say what it is. It says who hurts, how, and what changes. The "what we are" (a SaaS, an AI tool, a marketplace) becomes obvious from the story, so you never have to claim a category. Claiming a category is what you do when the story is not clear enough to imply one.

How investors actually use the sentence

Founders imagine the sentence as the opening line of a performance. Investors use it as a tool, and understanding the tool changes how you build the sentence.

First, the investor uses it to file you. In the first ten seconds they decide which mental bucket you go in, and a vague sentence lands in the worst bucket: "generic, will probably pass." A sharp sentence buys you the rest of the meeting because they are now curious rather than categorizing.

Second, they use it to test you. A good investor will repeat your sentence back, often slightly wrong, to see if you correct it cleanly or fumble. The repeat-back is a gift. It tells you exactly which part of your sentence survived and which part fell off in transit. If they repeat the customer and the outcome but drop the workflow, your workflow clause is weak.

Third, and this is the one that decides rounds, they use it to sell you internally. The sentence you say is the raw material for the sentence they say in the partner meeting. You are arming an advocate who is less informed and less convinced than you, and who has thirty seconds before someone interrupts. If your sentence needs context to make sense, it dies in that room. If it is self-contained, picture-able, and repeatable, it walks into the partner meeting and does your job for you.

This is why "repeatable by a stranger" beats "impressive when I say it." You will not be in the room where the decision happens. Your sentence has to go alone.

The artifact: a one-liner editing drill

Run your current sentence through this in order. Stop at the first failed step and fix it before moving on.

1. Name the human. Underline the customer in your sentence. Is it a person you can picture at a desk, or a category? If it is "teams," "businesses," "users," or "companies," replace it with the actual job title or situation. Pass only when a stranger could draw the person.

1. Find the pain. Underline the workflow that is bad today. If there is no "used to," no slow or manual or broken current state, your sentence is describing a feature in a vacuum. Add the painful before-state. Pass only when someone could wince.

1. State the outcome, not the feature. Underline what changes. If the change is "more efficient," "better," "streamlined," or "powerful," it is a feature in disguise. Replace it with a result in the customer's day or numbers: twice the loads, a third of the time, one number they trust. Pass only when the outcome is concrete enough to be wrong.

1. Cut the category claim. Delete the words "platform," "solution," "AI-powered," "operating system," "ecosystem," and "tool." If the sentence still makes sense, you never needed them. If it collapses, your story was carrying nothing and the category word was doing all the work. Rebuild the story.

1. The stranger test. Say it once to someone who does not know your company. Wait a beat. Ask them to say it back. Whatever they drop is your weakest clause. Rewrite that clause and run the test again. Pass only when the repeat-back keeps all three parts: customer, pain, outcome.

1. The Monday test. Imagine the person you pitched describing you to four colleagues who were not there, with no slides. Does your sentence survive that trip? If it needs you standing next to it to make sense, it is not done.

Do this once, seriously, and you will have a sentence that works in a cold email subject, the top of your deck, the first line of an investor update, and the first ten seconds of every meeting. One sentence, fixed once, repays itself in every conversation for the rest of the round.

Where this connects to running the round

The drill gives you a sentence. The harder question is which version of it works, and you cannot answer that from your own ear. You answer it from what investors repeat back to you.

That signal is sitting in your meeting notes right now, and most founders lose it. The investor said your company back to you in their words, and you moved on. RoundOS keeps that. When you log a meeting, you can mark the moment an investor repeated your description, and over a few meetings you can see which sentence keeps coming back intact and which one keeps getting re-described into something weaker. The sentence investors echo without prompting is your real one-liner. The one you have to keep re-explaining is the draft.

So the loop is: write the sentence with the drill, take it into meetings, and watch which version survives contact. Let the repeat-backs, not your taste, pick the final wording.