Method 4: Blueprinting
Rapid prototyping as exploration
Last week I walked through The Pattern Hunter—how to find meaning in qualitative data at scale. This week: what to do once you know what you want to build.
This one comes from my background in graphic design.
Early in my career, I worked in creative partnerships: an art director and a copywriter, the person doing the layout and the person writing the words. When you’re thinking about a campaign or a piece or whatever the project is, you’re simultaneously thinking about the concept, the creative, the idea, alongside how you’re going to express it. And the way you work through that is thumbnailing.
Thumbnails, if you haven’t seen them, are quick little sketches. Dozens and dozens of them. You’re exploring different ways to organize and shape your information, thinking about how the structure will impact how the content gets consumed, the experience you want to create for the person receiving it. When they open the first page, what’s the very first thing I want them to see? How am I setting the stage for what’s about to come? How can I use layout and design, type and space, to communicate something before the words even land?
The point of thumbnails is that they’re fast. Low fidelity. You’re not trying to create a finished piece—you’re trying to explore quickly. What if it’s laid out this way? What if this element is bigger? What if the story unfolds in a different order?
We used to do another exercise in creative work. Take the deliverable you’re producing—say, a report—and imagine it as something completely different. What if this report was a museum exhibit? What if it was an event, something spatial you could walk through and interact with? What does a report turned into an experience look like? This kind of thinking is why, over the years at Overlap, we’ve produced work that doesn’t look like traditional consulting deliverables. Instead of handing over a static document, we’ve created experiences that travel to different offices, interactive sessions that unfold over time.
AI lets you do this kind of exploration at a level of fidelity that wasn’t possible before. You can prototype multiple possibilities for what something could look like—not just outlines, but fully realized structural options—and evaluate them quickly. You’re thumbnailing, but at higher resolution and faster speed.
I am writing a book on strategy for nonprofit leaders. I had a full concept in my head: the audience, the core argument, the chapters, the case studies, the exercises. But I didn’t start writing Chapter One. Instead, I voice-dumped the entire concept into the AI: who this is for, what they need to understand, what each chapter should accomplish, how the pieces connect. Then I said: take all of this and build me the outline. Not the book. The outline. The AI came back with a chapter-by-chapter skeleton: titles, one-sentence descriptions, a logical sequence. From there I could see what was missing, what was redundant, where the argument didn’t build the way I wanted it to. I moved chapters around, combined sections, cut a chapter that was trying to be its own book. Only then did I start talking through the actual content, section by section, filling in a skeleton I trusted.
This way of working became so natural to me that I don’t think of it as a method anymore. It’s just how I approach anything with structure. A report, a presentation, a workshop agenda. Before I create the content, I sketch out how I want it to feel. I explore what done is going to look like.
Blueprinting is the discipline of building the skeleton before adding the flesh. You use the AI to generate structure first: outlines, frameworks, agendas, tables, sequences. You agree on the shape. You test the logic. And then, once you’re good with that, you start filling it in. Section by section. Not all at once.
This isn’t wildly different from writing a table of contents. It’s just that now I can create multiple tables of contents. I can explore whether the logic flows better if the information is presented this way or that way. I can think about how whatever I’m producing will be perceived, how it fits within a larger system of things I might be creating.
Maybe the real point is this: Blueprinting is an example of going deeper. Not just trying to get through the piece of work I have to get through, but actually spending time thinking about how whatever I’m producing will land, what the experience of it will be like. Because I’ve spent that time planning, the execution is faster. But the real value isn’t speed—it’s that I’ve more fully explored what I’m actually trying to do.
Why this matters.
There’s a cognitive tangle that happens when you’re producing something complex. You’re trying to figure out what to say—what’s the argument, what are the key points, what’s the evidence—while simultaneously figuring out how to organize it. What comes first? How do the sections connect? What’s the logical flow? Trying to do both at once is why complex documents feel overwhelming.
But we were never trying to do both at once in the art director/copywriter partnership. One person was thinking about the design, the structure, how the thing was going to work. The other was thinking about the words. Each person could give their full attention to their part.
Blueprinting gives you that same separation. Structure first, substance second. Each one gets your full attention.
And here’s where it connects to relational AI at its best: you’re using the AI as a collaborator. Here’s everything I’m thinking about. This is how I think conceptually it fits together. This is what I want the experience to be like. I’m open to suggestions from you (the AI)—what does that make you think about?
The moves.
In my practice, Blueprinting has three moves. (By this point there are more, but I’m just highlighting three each time. As always, take this as inspiration not prescription.)
The Skeleton Outline is the first move. You ask for the architecture before you worry about content. “This is where I think the book is going. This is how I think it conceptually fits together. This is how I think someone would use it. Clean this up to give me the outline of what I’ve just talked about so I can look at it and evaluate it.”
What makes this powerful is you can generate multiple options. Give me three different ways to structure this workshop—one that’s chronological, one that’s thematic, one that builds toward a decision. Then I choose, or combine.
Experience Translation is where it gets interesting. You take something that exists in one form and explore what it would look like in another.
Here’s a change management toolkit we’ve developed—it’s a report, a document. What if we thought about that content as a set of weekly emails drip-fed to the leadership team to support them in executing the ideas? What if we thought about it as a traveling experience to different offices? What if we expressed it as executive messaging, a one-hour introductory workshop, a training session, and a series of internal blog posts?
Suddenly you’re thinking about the system of execution around your content. You’re blueprinting not just one deliverable but the ecosystem of how that content gets into the world.
The Logic Flow Check tests whether your structure holds together before you invest in creating all the pieces. “I want to accomplish these things. Here’s the structure I’ve designed. Are we accomplishing what we said was important? What’s missing?”
The simplest version of this, and one I use more than any other, is: “What am I missing?” After laying out an annual planning process for a client, I asked: “What am I missing when thinking about the process of creating this organizational annual plan?” The AI came back with two things I hadn’t considered, two things that might be good builds. I’d likely have caught these eventually. Maybe I wouldn’t have. The point is I caught them all before the plan was built.
When to use it.
Before creating anything substantial. A quick email might not need a blueprint. A strategy document, a proposal, a workshop design, a program—those always do.
When you want to explore how something could exist in the world. Not just “what’s the outline” but “what are all the ways this content could reach the people it needs to reach.”
When you’re designing experiences that unfold over time. Workshops, courses, onboarding programs, multi-week projects, event agendas. Anything where the sequence matters and the pacing affects the outcome.
When you’re stuck. If you can’t start, you probably haven’t finished thinking about the shape of the thing. The blank page isn’t a content problem. It’s a design problem.
What’s underneath this.
When you face a complex project, the difficulty isn’t usually that you don’t know what to say. It’s that you can’t hold the whole thing in your head at once. The introduction depends on the conclusion. The recommendations depend on the analysis. Section three assumes something you haven’t established yet in section two. Everything is connected to everything.
Blueprinting breaks this tangle by making the connections visible and external. Once the structure exists on screen, your brain stops trying to hold it all internally. The overwhelming task of “produce this thing” becomes a series of manageable pieces.
And here’s where it connects back to Voice Dump Synthesis. Once I have the structure, I don’t sit down and type out each section. I record myself talking through each piece. I voice dump into the structure, section by section, thinking about what I want to say for each one. The blueprint tells me what to talk about. The voice dump gets the thinking out. That’s what happened with the book: the chapter outline existed, and then I talked through each chapter’s content into the skeleton, one at a time.
There’s also an evaluation loop that matters. When I have a blueprint, I can test it before I’ve invested in producing everything. “Here are the guidelines for this project. Here’s the structure I’ve designed. Are we accomplishing what we said was important? What’s missing?” The AI gives feedback on the structure against the criteria. It identifies gaps: “You’ve said engagement is a priority, but I don’t see where that’s happening in the design.” That kind of targeted feedback is possible when the structure is explicit and separate from the content.
This is really just good design. Thinking about who’s going to interact with whatever you’re creating. What they need. Getting feedback on ideas before you move them forward. You know, prototyping.
The difference is you can do it so much faster now. And because you can do it faster, you can do it more. You can explore possibilities you wouldn’t have had time for before. You can go deeper.
