How AI Helps Leaders Accelerate Business Growth
Want to make smarter, faster business decisions with AI? Wondering how leaders can effectively leverage AI without getting lost in technical details?
In this article, you'll discover a practical framework for becoming an AI-driven leader who can harness artificial intelligence to enhance strategic thinking and decision-making across your organization.
Begin With Two Sticky Notes
The first step in becoming an AI-driven leader isn't to delegate AI to someone else—it's to become competent with AI yourself.
First, grab a sticky note and write, 'How can AI help me do this?' and put that note on your computer or desk. This reminder will change your perspective as you go about your work throughout the day. Instead of asking, "What can I do with AI?" you'll start asking, "How can AI help me do this?"
On a second sticky note, write: "Context, Role, Interview, Task"
Geoff Woods uses this CRIT framework for 99% of his AI prompts, and it will provide you with a structured approach to interacting with AI effectively.
-
Context: Provide background information
-
Role: Assign AI a specific role
-
Interview: Have AI ask you questions to gain more context
-
Task: Specify what you want AI to accomplish
3 Ways to Use AI to Drive Business Results
Pitch Deck Reviews
About an hour before presenting an important sales pitch deck to a large company, Woods opened ChatGPT and followed his CRIT framework.
First, he gave ChatGPT details about context and role:
Context: I'm about to give this proposal to [company name]. Here is the proposal [uploaded PDF]. I'm about to meet with the executive and their chief of staff. We have an hour allocated for this. Role: I want to be able to role play with you as them so that you can give me feedback on what I do well, what I don't do well, and the top changes I can make. Your role is to act as both people.
After describing the personalities of both the executive and chief of staff in detail, Woods continued with the interview and task instructions:
Interview: When I confirm, I'll switch to voice mode and give you the proposal. Task: When I say, "All done," I want you to challenge me, ask questions, and raise objections. I don't want you to be nice. I want you to make this a ten out of ten tough.
Woods then practiced his entire pitch with ChatGPT's advanced voice mode, which interrupted him with questions and objections. When he finished saying "all done," ChatGPT provided feedback highlighting a critical oversight—he hadn't outlined a return on investment. Woods immediately amended his deck to address this gap.
Financial Statement Analysis
When reviewing his financial statements, Woods opens ChatGPT and applies his CRIT framework:
Context: Here are our financials [drag and drop]. Role: You're a strategic CFO who's a world-class expert in telling a CEO the top five non-obvious insights about their business based on their financials that they don't know that they should know. Interview: Ask me up to five questions, one at a time, to gain deeper context. Task: Give me the top five non-obvious things about my business that I should know as a CEO.
This approach has become a monthly rhythm for Woods because it helps him see patterns and insights he might otherwise miss.
Board Presentation Reviews
When a CEO in his network was struggling with a hostile board that consumed 50% of the executive team's time, Woods helped develop a novel solution.
First, Woods pulled the executive team together on Zoom and used ChatGPT with his CRIT framework.
Context: I'm the CEO of this company. We have a hostile board. Every quarter is an absolute bloodbath, and 50% of my executive team's time is gobbled up by board shenanigans and distractions. We've been working toward an exit in the next twenty-four months, but I just got a call from the chairman saying if I don't rehab the relationship in the next six months, we're all fired. Role: Your role is to act as an HR professional with deep expertise in creating personality profiles. Interview: "Interview me. Ask me one question at a time. You get to ask as many questions as you want until you understand one of our board members, Susan, on a deep enough level so you can accomplish the task of creating a personality profile for her. Task: Create a personality profile for Susan.
Then, the AI conducted the interview, asking the executive team questions about Susan. As the executive team responded, Woods literally typed their responses word for word.
After this process, the AI created a personality profile, which the executive team reviewed critically: "They said, here's what I like about it. Here's what I don't like about it. Here are the top changes we need to make."
They repeated this process for each board director, creating personality profiles for every board member.
Then, they built a custom GPT that included these profiles with instructions:
Your job is to be our AI board. Every quarter, we'll give you our deck before the meeting. Your task is to review every slide as every director and simulate how you think they will react.
The AI quickly identified potential sections of the presentation that could derail the board meeting. In one case, it predicted that a particular board member would get distracted by details on slide eight, leading to a 30-minute detour. Then, the AI gave the team a suggestion to modify the presentation in a way that would appeal more favorably to the director.
Finally, the team implemented an AI note-taker in the meetings. They upload the transcript and tell the AI to compare its simulation to what actually happened in the meeting, then adjust the personality profiles based on the reality.
Within two board meetings, the board told the CEO, "These are the best decks we've ever had and the best meetings we've ever had. What changed?"
Today's advice provided with insights from Geoff Woods, a featured guest on the AI Explored podcast.
Watch the full interview YouTube P.S. Add
into your contacts list. Use Gmail?
We publish updates with links for our new posts and content from partners. Your information: Email:
Opted in on: 2021-09-06 17:03:43 UTC.
