How to Get Honest Feedback on a Talk You Gave

Meera Datey — Thu, 09 Jul 2026 01:42:15 +0000

You gave the talk. The sprint review demo, the lightning talk at the meetup, the conference session you spent three weekends prepping. Someone hit record, and now there is a video file with your name on it.

You will watch it back once, and learn almost nothing. Not because you are lazy. Because reviewing your own recording feels productive and is actually a no-op.

Watching yourself present is like reading your own code right after you wrote it. You do not see the bug, because your brain is running the version in your head, not the one on screen. You know what you were trying to say in minute four, so you skim right past the twenty seconds of "um, so, basically" that everyone else sat through.

The audience did not get the version in your head. They got the actual bytes. Your intentions do not deploy.

Confidence is not the metric. Plenty of engineers demo with total conviction and still lose everyone to a wall of jargon at 190 words a minute. Sure-of-yourself and unintelligible are independent variables, and the second one is the one you cannot measure from the inside.

So what do people do with the recording? Mostly nothing. Maybe ping a teammate too slammed to watch it. The file rots, the habits repeat, and the next talk is a carbon copy of this one.

That is the waste, because the tape is the best teacher you will get. What you want is the code-review equivalent for the presentation itself, the you-standing-there-talking part, not the slides. An impartial pass that just tells you the facts: four minutes before you showed anything running, pace spiking on every slide you were unsure about, "does that make sense" seven times, filler count north of respectable.

That is not criticism. That is a diff. Facts you can act on, with nobody watching you read them.

And here is what is in it for you. The engineers who move up are not always the best coders. They are the ones who can stand up in a sprint review and make people actually get it. That skill compounds: every talk you learn from makes the next one land harder, and landing talks is how projects get greenlit and you get known for more than the commits.

You already did the hard part. PresenterPrep is that review pass for your delivery. Drop in the recording you already have, video or audio, no slides required, and you get a plain read of how you did: words per minute, filler count, whether your energy held or cratered, clarity, the two or three things that landed, the two or three to fix. Full transcript with timestamps, so you can jump to the part where you lost everyone. No dependency on anyone else's calendar.

Try uploading your old footage into PresenterPrep and get the feedback you deserve to actually get better. Not for anyone else, for the person giving the next talk, who is you a few sprints from now, wondering why it suddenly feels easier.

Record your pitch and hear it back at PresenterPrep →

You shipped the talk. Now read the logs.

Your pitch deck is done. Now what? How technical founders actually prepare to deliver it.

Meera Datey — Sat, 30 May 2026 07:34:19 +0000

You spent three weeks on the deck. The problem slide is tight. The market sizing is real. You even got the competitive matrix right.

And now you have a VC meeting on Thursday.

Most technical founders do one thing at this point: read the deck again. Maybe fix a font. Maybe tweak a number. Then show up on Thursday and hear themselves stumble for the first time in the actual meeting.

This is the preparation gap. And it hits technical founders harder than any other archetype — not because they're bad communicators, but because they've spent their careers optimizing for written precision. Code reviews, PRs, documentation. Asynchronous, careful, revisable.

A pitch is none of those things.

Why technical founders specifically struggle

When an engineer debugs, they read the code. When they review an architecture, they read the doc. When they prepare to pitch, they read the deck.

But reading is not the same as saying.

The gap shows up at specific moments.

Follow-up questions. A VC asks "why now?" You know the answer. But you've never said it out loud under pressure, so you say something close to the answer but not quite it. You hear yourself trailing off. The VC writes something down.

Slide transitions. On paper your story flows. Out loud you realize there's no natural bridge between your market slide and your business model. You've read over this gap fifty times without noticing it. Reading fills in the gaps automatically. Saying it out loud exposes them.

Confidence under pressure. Investors are not just evaluating your idea. They're evaluating whether you can attract talent, close customers, and hold a room. A founder who knows their deck cold but sounds hesitant when speaking it is sending a signal — not about the idea, but about the person.

This is not a soft skills problem. It's a practice problem. And technical founders solve practice problems with systems.

Step 1: Get a fresh read before any investor sees it

You've been building this product for months. That proximity is dangerous. You forget what it looks like to someone who has never heard of your company. You write "our proprietary data pipeline significantly reduces latency" and it makes total sense to you. To an investor reading it cold, that sentence raises three questions and answers zero.

The gaps in your deck are not gaps you can find yourself. You're too close.

Find a mentor or advisor who doesn't know your business well and ask them to read it cold. Watch where they slow down. Watch where they look confused. Those are the gaps.

If you don't have a mentor available, AI tools like ChatGPT or Claude can give you a useful outside read. Paste your slide content and ask: "What is missing from an investor's perspective? What claims need more support? Where does the narrative break down?" It won't replace a human who knows the fundraising world, but it catches the obvious gaps fast.

What you're looking for:

Slides where the narrative assumption is not stated (you implied "why now" without saying it)
Missing sections investors expect (team credibility, unit economics, use of funds)
Claims without support ("large and growing market" with no number behind it)
Language that only makes sense if you already know the product

Fix what you find. Then move to delivery.

Step 2: Actually say it out loud. Under pressure.

This is the step almost nobody does.

Not because founders don't know they should practice. They do. It just feels awkward to talk to nobody. So they read the deck out loud once, call it rehearsal, and move on.

Reading is not rehearsal.

Start simple: stand in front of a mirror and give the pitch. Full version, out loud, no stopping. This feels uncomfortable. That discomfort is useful information. The moments where you slow down, reach for a word, or lose your train of thought — those are the gaps in your delivery, not your knowledge.

Here is what changes when you actually speak it:

You discover the words that are not yours. Most pitch decks have language written to sound good on paper. "We are uniquely positioned to capture," "a paradigm shift in how X is done." You cannot say these out loud without sounding like a press release. You need to find the version you would actually say to a friend.

You find the pacing problems. Some slides have too much content for the time you have. You do not know this until you are mid-sentence and realize you are 40 seconds into a slide you need to be done with in 20.

You discover what you actually believe. When you say your market size number out loud, do you sound like you believe it? Investors hear this. The founder who has practiced sounds like they own the number. The founder who has not sounds like they are hoping nobody asks.

After you can get through the pitch cleanly, practice the hard questions. Generate a list of the toughest investor follow-up questions for your specific deck. ChatGPT or Claude are good at this — paste your deck content and ask for the 10 most likely investor questions. Then answer them out loud, one by one, until the answers sound natural.

The questions that trip up most technical founders:

Why now, not two years ago?
What happens if a well-funded competitor enters this space?
Why are you the right team for this?
Walk me through your unit economics.
What does a customer who churns look like?

These are not trick questions. They are standard. But if you have never said your answers out loud before Thursday, Thursday is the first time you will hear how you sound.

Step 3: Record yourself. Watch it once.

This is the part that feels most uncomfortable and produces the most improvement.

You do not need a professional setup. Open QuickTime on your Mac, hit new movie recording, and give the pitch to your laptop camera. Any screen recorder works. The point is to see what the investor sees.

Founders consistently overestimate how they come across when speaking about their own company. They are energized by the topic. They know every word. From the inside it feels confident and clear.

From the outside it often is not:

The pace is too fast (you know where you are going, so you rush)
Eye contact breaks toward the slide at the wrong moments
Filler words cluster around the transitions — this is where the uncertainty lives
The energy drops on the slides you are least sure about, and investors notice the drop

Watch the recording once. You will immediately know what to fix. You do not need a coach to tell you. You already know.

Do not try to fix everything. Pick the three most obvious things and record again. Repeat until you have one take where you sound like yourself at your best. Not a rehearsed version. Not a nervous version. The take where you would want an investor to see you.

The system

Technical founders are good at building systems. Here is the one for pitch prep:

Get a cold read from a mentor or AI. Fix what they find.
Stand in front of a mirror and give the full pitch out loud. Do it until it flows.
Generate the hard investor questions and practice answering them cold.
Record yourself on QuickTime. Watch once. Fix the three most obvious things. Record again.

This takes a few hours, not days. And it is the difference between showing up on Thursday hoping you remember the answer to "why now" and knowing you have said it out loud twenty times already.

Your deck got you the meeting. Your delivery keeps you in the room.