What to Do Before an Investor Pitch
Published 8 min read
Short answer
Before an investor pitch, rehearse the questions rather than the deck. Spend the final fortnight recording your sixty-second version and the ten answers you are dreading, the day before on one clean run and proper sleep, and the last ten minutes on slow breathing and a deliberately slow first sentence rather than on changes.
Most pitch preparation is spent on the wrong half. Founders rework slides until the night before and rehearse the questions not at all, which is backwards: the deck is the part you control completely, and the questions are the part that decides the meeting. This is a timed plan built around that inversion - what to do with two weeks, with one day, and with ten minutes. It assumes the deck is finished, because if it is not, the answer to what to do before the pitch is finish the deck.
Two weeks out: build the question bank, not more slides
Write down the ten questions you are most hoping they will not ask. Not the ten most likely - the ten you are dreading, which is a different and much shorter list, and the one that predicts how the meeting goes. The overlap between the two lists is usually about half, and the non-overlapping half is where meetings are lost.
Then answer each one out loud, cold, timed, once. No rehearsing before the first take: the first take is the data. Most founders discover two things immediately - that answers they thought were thirty seconds run ninety, and that pace and filler rate on the dreaded questions are far worse than on the pitch itself, which they have rehearsed. That gap is the actual preparation gap.
The full treatment is in preparing for investor Q&A. The short version: an answer you have said out loud once is a different object from an answer you have thought about, and only the first survives being interrupted.
The ten that belong on almost every founder's list:
- Why now, and why has nobody done this already?
- What happens when the obvious large incumbent decides to do this?
- Walk me through the unit economics at your current scale.
- What is your actual retention, cohort by cohort?
- How much of that revenue comes from a single customer?
- What did you get wrong in the last twelve months?
- Why is this team the one to do it?
- What does this round buy you, in milestones rather than in months?
- What is the weakest part of this business?
- Who else is in the round, and on what terms?
The fortnight: one minute a day, same prompt
Record the sixty-second version of the pitch every day. Same length, same prompt, same time of day if you can, because the comparison is the entire value and a comparison between different prompts measures nothing. Ten minutes total across the whole fortnight, and it outperforms three long rehearsal sessions comfortably.
Watch four numbers and ignore the rest. Pace should sit between 130 and 150 words per minute and, more importantly, stay there across the whole clip rather than averaging into range by running slow through the safe part and fast through the ask. Filler rate under three a minute. A pause of half a second to a second and a half before each conclusion. Clarity holding on the last words of sentences, which is where it fails first.
Alternate: pitch on odd days, a dreaded question on even ones. The question days are the ones that move, because the pitch is already over-rehearsed and the answers are not.
| Number | Target | What a bad reading means |
|---|---|---|
| Pace | 130-150 wpm, stable across the clip | You accelerate into the part you are unsure of |
| Filler rate | Under 3 per minute | You are composing while speaking, not recalling |
| Pause before conclusions | 0.5-1.5 seconds | Every sentence lands at the same weight |
| Clarity at sentence ends | Under 5% unclear | You are outrunning your own articulation |
The day before: stop changing things
No new slides, no new numbers, no restructuring. A change made the day before has not been rehearsed and will be the sentence you stumble on, and the cost of stumbling is higher than the benefit of marginally better wording almost every time.
Do one clean run of the whole thing, out loud, standing, timed, without stopping to fix anything. Stopping to fix is the habit that produces a founder who can deliver the first two minutes beautifully and has never once delivered the ending. Then do one pass on the three questions you are still worst at, and stop.
Then sleep, properly. Sleep loss shows up in speech before it shows up anywhere you can feel it - pace rises, articulation degrades, and recovery from an interruption gets noticeably slower. It is the highest-leverage thing available on the last night, and the one most reliably sacrificed to slide edits.
Logistics worth checking the day before
Boring, and each one has ended a meeting badly for somebody. Ten minutes of checking removes an entire category of failure that has nothing to do with how well you speak.
- Test the actual video link on the actual device, signed into the account you will use. Share a slide and confirm the audio route while you are there.
- Have the deck as a PDF locally, not only in a cloud tab, and know which key advances it.
- Know your three headline numbers well enough to say without the slide, because the slide will be the one that fails to load.
- Confirm who is in the room and what they invest in. Being pitched a seed thesis by someone who only does Series B is a five-minute meeting.
- Set the room up: light in front of you, camera at eye level, phone off rather than face down.
The last ten minutes
Do not rehearse. Reading the deck one more time in the ten minutes before does nothing for recall and reliably raises your baseline speaking rate, which is the opposite of what you need walking in.
Slow breathing instead - out for longer than in, for about two minutes. This is not a relaxation ritual; it is the most direct available lever on the physiology that makes you speak too fast, and it works within a couple of minutes.
Then plan the first sentence, and only the first sentence. Say it slower than feels natural, because the opening is delivered at whatever pace your adrenaline sets and everything after it anchors to that. A deliberately slow first sentence buys back the entire first minute. Sounding confident in an investor pitch covers what a room reads off that opening, and the pitch practice tool is where the fortnight of recordings goes.
Questions
What should I do the day before an investor pitch?
One clean uninterrupted run of the whole pitch out loud, one pass on the three questions you are still worst at, and then stop. No new slides and no restructuring - an unrehearsed change is the sentence you will stumble on. Then sleep properly, because sleep loss degrades pace and articulation before you can feel it.
How long should I practise before a pitch meeting?
One minute a day for the final fortnight, recording the same prompt each time, plus one full run the day before. Under half an hour of practice in total. That beats three long rehearsal sessions, because the value is in the comparison between days, and a long session gives you one data point and a sore throat.
Should I memorise my pitch word for word?
No. A memorised pitch fails at the first interruption, because recall of a fixed sequence does not survive being knocked out of it, and investors interrupt. Memorise the sixty-second structure and the three headline numbers, and let the wording vary. The one thing worth rehearsing word for word is the first sentence.
What do investors judge in the first minute?
Whether you can explain the business clearly, and whether you sound like someone who knows it. That reads mostly off delivery, since they cannot yet evaluate the substance - pace, whether you pause before the important sentence, and whether the filler rate suggests recall or composition. It is not fair, but it is what a stranger has to work with in sixty seconds.
What should I do in the ten minutes before the meeting?
Slow breathing, out-breath longer than in, for about two minutes, then plan your first sentence and deliver it slower than feels natural. Do not read the deck again - it does nothing for recall and raises your baseline speaking rate at exactly the wrong moment.