I have sat through more startup pitches than I can honestly count. Accelerator demo days, agency innovation reviews, university pitch competitions, and at least one living room where somebody's cousin was the only investor in the building. Most of those pitches failed for the same three reasons, and not one of the three was "the deck needed a nicer font."
Founders spend weeks on slide transitions and thirty seconds on the one question that actually decides the outcome: what evidence do you have that this problem is real and somebody will pay you to solve it? Get that part right and a plain deck with typos still lands. Get it wrong and the best design in the world just makes the rejection prettier.
I coach founders through customer discovery and non-dilutive funding, and I have watched pitches die in the first ninety seconds and watched others survive a room that walked in wanting to say no. The difference was never the slides. Here is what actually makes a pitch work, drawn from the ones that did.
What investors are actually buying
Investors are not funding your idea. Ideas are free and everyone in the room has already heard three versions of yours this month. What they are funding is evidence that you understand a real problem better than anyone else in that room, and that you can build a repeatable way to make money solving it. The Y Combinator pitch framework boils this down to seven questions worth answering before you write a single slide: what you do, how big the market is, how much progress you have made, what you understand that others do not, how you make money, who is on the team, and what you are asking for.
The "why now, why you" test
Two questions stop more pitches than anything else. Why does this need to exist now, and why are you the person who should build it. If your honest answer to either one is "no particular reason," fix that before you fix your slides. A good "why now" points at something that changed recently, a cost that dropped, a regulation that shifted, a behavior that only became normal in the last few years. A good "why you" is not a resume recitation. It is the specific, sometimes uncomfortable, experience that means you see this problem more clearly than a founder without it.
The pitch that survives a real conversation
A pitch that holds up under questions covers six things and no more: the problem, in language a stranger would recognize from their own life or job. The evidence you have that the problem is real, meaning customer conversations, pilot data, or early revenue, not assumptions. The size of the market you are actually going after first, not the total addressable fantasy you will grow into eventually. The business model, meaning who pays, how much, and how often. The team, and specifically why this team can execute against this problem. And the ask, stated plainly, with what you will do with the money and what happens next.
Skip any one of those six and the room fills in the blank with their own doubt, which is always worse than whatever you would have said. Silicon Valley Bank's guide to pitch decks lands on a similar structure, and for good reason. Investors have seen thousands of these. They are not grading your creativity. They are pattern-matching against every pitch that came before yours, and the pattern that survives is the one that answers questions before they get asked.
The three questions that kill most pitches
Three questions sink more pitches than anything about the product itself. Who is your real competition, and most founders answer this one badly by naming other startups instead of the actual alternative, which is usually a spreadsheet, a part-time employee, or doing nothing at all. Who absolutely needs this, meaning a specific, narrow customer who is in enough pain to buy today, not a broad market you hope will eventually care. And what happens if you are wrong, which is the question that separates founders who have actually pressure-tested their own thinking from founders who are pitching a hope.
Answer those three honestly and most of the rest of the pitch takes care of itself, because you have already done the thinking that the good questions in the room are trying to draw out of you.
Rehearsing it without embalming it
Practice the pitch out loud, to actual people, until it stops sounding like a hostage note read off an index card. That takes longer than founders expect, usually fifteen to twenty run-throughs before the language starts sounding like something you would actually say to a person instead of something you memorized to survive a meeting.
Here is the part founders get backwards: the goal of rehearsal is not to lock the pitch into one perfect script. It is to know the material well enough that you can answer a curveball question without losing your place, and confident enough that you can drop the parts that are not landing in that specific room. A pitch that only works word-for-word is a pitch that breaks the moment someone interrupts you, and someone always interrupts you.
When the problem is not the pitch
Sometimes founders rewrite the deck for the fifth time and the pitch still keeps failing. At that point the deck usually is not the problem. The evidence underneath it is. If you cannot answer specific questions about who your customer is, what they currently do instead of your product, and how much the problem costs them, no amount of slide polish fixes that gap. That is a customer discovery problem, not a pitch problem, and it is worth reading through how many customer interviews you actually need and what those first conversations should sound like before you touch the deck again.
I have watched founders spend three months perfecting a pitch for a business model the market had already told them, clearly, it did not want. The pitch was fine. The business underneath it was not. Fix the evidence first. The pitch is just the messenger.
Frequently asked questions
How long should a startup pitch be?
A live pitch runs 10 to 15 minutes with room for questions, and the deck behind it should stand on its own in under 3 minutes of silent reading. If an investor cannot tell what you do, who pays for it, and why you in the time it takes to skim ten slides, the pitch is too long or too vague.
What is the biggest mistake founders make when pitching?
Leading with the product instead of the evidence. Founders describe features for five minutes before mentioning a single customer, a single dollar of revenue, or a single validated problem. Investors are not evaluating your product tour. They are evaluating whether you have proof anyone besides you wants this.
Should I get an NDA before pitching my idea?
No. Investors and program reviewers will not sign one, and asking signals inexperience more than it protects anything. Ideas are cheap and rarely stolen. Execution is what is actually hard to copy, so protect your traction and your team instead of your slide deck.
Do I need a working product before I pitch?
No, but you need evidence. A working product helps, but a founder with 60 structured customer conversations and a clear problem statement can out-pitch a founder with a polished demo and no evidence anyone will pay for it. Evidence of a real problem beats a proof of concept nobody asked for.
How do I pitch to a non-traditional buyer, like a government agency?
Translate your pitch from investor language to mission language. Government buyers and agency sponsors are not asking about your exit or your multiple. They are asking whose operational problem this solves, who owns the budget to buy it, and what happens if they do nothing. Answer those three questions and the rest of the pitch follows.
Should the pitch deck be different from the pitch itself?
Yes. The spoken pitch is a conversation you can adjust in real time based on what the room cares about. The deck is a document that has to work when you are not in the room, forwarded to a partner who was not there. Build the deck to stand alone, then use the live pitch to go deeper on whatever the deck cannot fully explain.
I built an entire talk around this because the pattern repeats so often it stopped surprising me. The pitch deck is not the business. It is a document that either earns you the next conversation or does not, and the fastest way to earn it is to walk in with more evidence than the founder before you.