How we used AI to prep for our board meeting
Before you get in the hot seat in front of a difficult client, stakeholder, or board member, build your confidence by asking AI to prepare you.
👋Hi, it’s Greg and Taylor. Welcome to our newsletter on how to make high-stakes professional and personal decisions in your 30s.
Last September, I gave our Q3 board deck to ChatGPT and Claude and asked for feedback.
I was mostly curious. I didn’t expect much from it. The results – especially from Claude – stunned me. Now, in preparation for every high-stakes meeting, I ask AI, “What do you think?”
You should be doing this too. Working in a fast-paced environment is all about smart ideas and plans – delivered with confidence. AI will build your confidence 10x going into a meeting, because you’ll have thought of all the angles your audience might bring up.
If you think you’ve presented enough to skip this step, you’re wrong. I’m a six-time CEO, and I still feel more confident knowing what AI thinks of the deck before I present it.
Here’s how to do it.
Greg
How we prompted AI to act as an experienced board member
We uploaded our entire board deck to Claude and ChatGPT, and we didn’t anonymize the data, though we probably would if we were a larger, public company. Our POV is that Section is small/inconsequential enough that our financial data is a drop in the AI data training bucket.
Here’s our prompt (feel free to borrow it):
Please review this deck and give me feedback as if you’re a board member. Give one set of feedback as if you’re an aggressive, growth-minded board member, and another set of feedback as a more conservative, risk-averse board member.
What are your top 3 takeaways and your top 3 concerns after reviewing the deck?
After reviewing the deck, what are some next steps or recommendations you would give to the CEO and the COO?
How AI did vs. the board
We gave the deck and our prompt to four LLMs and compared their performance against the human board.
Long story short – Claude was excellent. It asked 91% of the questions our human board asked, was appropriately ambitious, and could foresee business implications 12-18 months out.
Our favorite insights from Claude:
It questioned the plan’s dependence on closing a small number of deals. Claude said: “Achieving the AI Academy enterprise revenue target by September becomes very important to extend the runway. The dependence on a few key enterprise deals and distribution partnerships carries risk in hitting those revenue goals.”
It asked about our ability to differentiate. Claude said: “As demand for AI-related training grows, the competitive landscape is likely to become increasingly crowded. The deck does not provide a clear strategy for how the company plans to differentiate itself from existing or emerging competitors in the AI education space.” That was right – our board ultimately asked us this a few times – and we hadn’t included it in the first version of our deck.
It was concerned about focus. Claude said: “I would push leadership to really double down on just the most critical initiatives required to prove out product-market fit and a repeatable sales model for the AI Academy with enterprises over the next 6-12 months. Everything else should be deprioritized until you have that solid foundation in place to then accelerate growth.” We then asked which projects it would deprioritize, and it provided a solid list – very helpful. We are juggling a lot of products right now to hit our revenue target, but this effort is distracting from building more sustainable enterprise value.
In comparison, ChatGPT was pretty good, but more generic than Claude. And we wouldn’t waste time with the others (Gemini and Copilot) for now – they mostly regurgitated the deck.
Our takeaways and advice
We’ve done this three times now, and every time AI is almost as good as the board. It’s not better – but keep in mind, we’re lucky. We have a great startup board. Most founders and CEOs don’t.
If you work at a startup in any role - but especially as a founder/CEO or exec, I think it is startup malpractice not to be asking AI for help. It doesn’t replace conversations with your human stakeholders. But it does dramatically increase the quality of these conversations.
Every startup leader should be asking themselves “what does AI think?” before any major decision or presentation. No excuses.
To the next 10 years,
Greg & Taylor
Why is 100% A-?