You're a founder of an early-stage startup. For the past few years, you've bootstrapped the business, but the time has come to begin scaling core functions, which requires outside capital. So you pitch. You've got the story, some market adoption data, a small but impressive team, a great idea, and a lot of emotion tied up in this endeavor that has captured your imagination and attention for months, if not years.
Putting all that into a pitch deck that your dream investors will find compelling might be as daunting as building the business itself – maybe more. We're here with some good news. Recent studies show that pitch decks built with help from Generative AI are performing better with investors than those developed solely by the founder and startup team.1
These findings shouldn't be all that surprising. LLMs and ChatGPT are trained on large datasets. VCs continue to produce content like How to Design a Better Pitch Deck articles because they see thousands of presentations that indicate that founders can benefit from these 'how-to' documents.2 That content goes into the datasets feeding the Generative AI, making it possible for someone to prompt GPT4 to give them a winning pitch deck structure and even to use AI to design and populate the slides.
Fed by best practices and thousands of examples, GPT4 can deliver (in seconds, for free) a deck structure that reflects best practices, is unemotional, and has the correct outline and content to compel the VC who receives it. (Of course, do your homework and use intelligent prompts. Don't just rely on the first output that the technology gives you.)
We brought together two vice presidents from our investment team, Dustin Drees and Stephen Joy, to discuss why AI-built decks might perform better and how using these resources will change pitching for founders and VCs. We've summarized that conversation below.
Professionalism: Generative AI gives you good scaffolding for building the right kind of presentation. It'll identify the right types of information, data visualization, and storytelling elements for what you're striving to convey so the pitch follows best practices.
Structure: Generative AI gives you the benefit of having the 'wisdom of the crowd' for your pitch structure. Consider a teacher who gives excellent, straightforward instructions for an essay: three paragraphs, one quote, a clear intro, and conclusion sentences. The student who follows those instructions will be more successful. In other words, take in a few suggestions to make sure the instructions are sound, and then follow them. It will take some mental lift off the VC, allowing them to move quickly through the document and capture the information they need.
Specificity: The gatekeepers at a VC want to see certain information in pitch decks. If you ask ChatGPT to give you that outline, it makes sense the deck will be more successful. (For example, ask, "You are an analyst at a top-tier VC firm who reviews thousands of pitch decks for technology startups. Give me the outline of a pitch deck that prompts you to contact the founder for a meeting.”)
Bias-free: As you build a deck to pitch an innovation you've put all your energy, time, and even personal money into, you might be challenged to decide how much of your knowledge to share and how much personal storytelling is warranted. AI is not subject to human bias and likely will have an easier time conveying the most critical information in 10 slides.
Most VCs fund about 5 out of every 1000 companies they review. Those are challenging odds for VCs and founders. The hard truth is that at the start, VCs are looking for reasons to unqualify decks quickly before spending time in them. Most VCs will skim through a deck on a first pass to confirm the business is relevant to what interests them and in line with their investment thesis. If it is, they'll spend more time on it. The good ones will get you a meeting.
An AI-formatted deck will be more consistent with an expected format so the VC can digest it quickly and without too much effort. In that way, AI removes some of the barriers to entry by making the information more digestible. This is a benefit to the founder. Knowing there are initial gateways (business relevance, presentation quality), a founder can build a deck that maximizes the quality, compelling information the VC gets on the first pass, so the second step is more likely.
Unless the information is immediately compelling, a deck that looks terrible will get passed over immediately. It tells us that the founder didn't do their homework. If every deck looks great, we either have to refine the screening process or take more meetings. If you have such a great deck that it gets you in front of many investors, good for you. Do everything you can to get meetings with good investors.
AI is driving up the average quality of documents across a lot of industries. With respect to pitch decks, pervasive AI use might mean that the one-time 'standouts' get pulled down to the average. Or it might mean that once mediocre pitches become standouts. Those scenarios might be a downside for founders who have a track record of being good at raising money.
The 'A' players who understand what's needed to get to 'Yes' and how to deliver should not allow AI to do too much for them. But AI can be an advantage for 'C' and 'B' fundraisers to level up and compete with the 'A' players. If you are a founder who knows that pitch deck building is not a strong skill, AI might be a good way to put your best foot forward and get the opportunity to meet with the VC.
The biggest advantage for a founder is that a polished, hyper-relevant deck may get you in front of more investors. Just remember that if the business isn't ready or it doesn't match the VC's investment thesis, even a stunning AI-generated deck will only hide issues until a slightly later phase.
The pendulum may swing the other way. It improves the table stakes. It's a tool that should enhance efficiency and product output. If founders and VCs know that AI is generating most pitch decks, it might make a founder think deeper about the idea. They may find other ways to stand out, like making the deck more intellectually engaging or highlighting more differentiators.
If I were a technical founder great at building products but raising money wasn't my thing, I'd try using AI to build my deck. I'd A/B test a few to get feedback about how the pitch deck helps me to stand out.
Founders should know how to use this resource and be at least 'GPT polished.' But it's up to you to leverage the data to tell a compelling story to the right investors. And if your business isn't relevant to the investor's thesis, the market is off, or the platform isn't good, even the most polished pitch deck will fall flat.
.jpg)
