|
The first time I built a ghost deck was in Basic Consulting Readiness training at McKinsey, and it was uncomfortable. I had always been an "bottom-up" analyst. I didn't want to plan the high level. I wanted to dig into the data and figure it out as I went. That works when you are doing solo work, but isn't great when you are on a team. A ghost deck is a presentation you write before you do the analysis. It's simple: Each page gets a full-sentence title that states what you expect to find, and under it a sketch of the chart or a few bullets on the data that would prove it. The title is where the you are forced to think through what you might be trying to argue at the end of your analysis. You might write a title: "[X]% of our shoppers already pay a rival for delivery." Then a mockup of the data that would prove it. It feels like making things up, and people who love the data (me included) resist it. But it's the fastest way to see where a project is heading and where the holes are, before anyone spends a week in a spreadsheet. I think this matters more now. People can dump meeting notes into AI and get a polished deck back in minutes. It looks finished, so nobody asks whether it says anything. A ghost deck forces you to turn that pretty deck into something that is more structured and clear AND more importantly, aligned with the work you'll actually do. I think with AI you can actually do this as you go, and "solve" it by doing many more iterations. But if you are new to this, I highly recommend trying to print out some blank slides and actually trying to write out titles for your points. I wrote up how to build one from an issue tree: Do you sketch the deck before you start the analysis, or after? Hit reply; I read all of them. |
Freelancer, creator & writer
I ran 4-5 client workshops over the last few months, and all of them ended with one question: what does working with AI mean for your team? This is THE question right now Here's what I'm seeing inside organizations: Each person has their own approach; no one is sharing recipes Confusion about which models or platforms to use Top-down dictates without the leaders even knowing what they want. Lots of overconfidence with junior people shipping slop One prompt turns into a 30-page deck, and then...
In my workshops, I often use this slide: I'll then ask people to drop in the chat: "What percentile do you think 'high-performance' stands for if 100% is best? Answers flow in: 85% 95% 70% 60% 88% 99% It's almost always like this. I do something like this to show people that while we often have an exact number, it rarely matches what people are thinking or perceiving. This phenomenon is usually referred to as the illusion of transparency. If you can get more specific and exact, you will...
The latest version of Claude can create PowerPoint slides. So I gave it a challenge. Turn this "bad example" I use in my workshops into a good slide based on the principles I share on strategyU.co I gave it this simple prompt that took me no longer than ten seconds to write: Take this slide and take the lessons on slide design on strategyu.co and turn this into a compelling slide, improve the fonts, get to one single insight, match the title and content, and have the title be a clear takeaway...