I'm experimenting with ways to demonstrate the path from a raw idea to a well-shaped pitch. That is, how to go from "I think we should spend time on X" to "here's a specific concept for X that we're confident we can ship in six weeks."
To get through it, instead of just staring at a blank page, I opened up a Miro board. My idea was to externalize my thought process by making small steps in a horizontal chain.
This project is one of the final cycles of an app I'm building with Bob and Greg to analyze and cluster demand-side interview data. Here's a screenshot of the work in progress
All of these steps are built and working. The problem is that the clustering process in step four is confusing. It works, functionally, but it's very "power user" — you have to know a lot of things in your head and manipulate a lot of controls on the same screen that aren't well explained.
We agreed that if we could somehow "make clustering better" in six weeks, it would be time well spent. But that's too fuzzy to just start working on. I needed to shape that raw idea into some kind of specific project we could go do.
I didn't understand what "better" should actually look like.When this happens to me, sometimes I reach for a tool I'm calling TRACE. It's a technique to turn a fuzzy sense that something is wrong into specific focal points.First I observe somebody using the thing for a real purpose (or get a recording or interview them).Then I note every single step they took, from start to finish, including steps outside the tool. Eg. the timeline includes when they needed to switch to a different app, or Google a question, or apply a workaround.Any time they veer off the golden path and apply a compensating behavior, I flag that. This (a) indicates a problem, and (b) the compensating behavior shows what a solution could look like.I collect those flagged areas as the starting point for the design work.
As soon as I saw the specific problems, lots of different solution ideas started popping into my brain. I didn't want to slow down and start writing any one of them in detail because I was afraid I would lose the whole bunch. So I reached for a tool I affectionately call the DUMP.A dump is just a box where I tell myself "you can put anything here without worrying if it's right or not." Importantly, there is no structure and no relationships. Just one column, and things go in one after the other.
Whenever I have too many small things, I reach for AFFINITIZE. This is a chunking exercise that puts like with like, so you can go from many small things to a few big things. The secret to affinitizing is started with a fixed number of unnamed groups. Then, only after you've grouped things together, you look at the piles to understand what to call them.
I wasn't ready to write yet. But I could see all the parts of the pitch in my head. So instead of going down into the weeds of writing, I created another dump and put all of the ingredients of the pitch into it.
When I have a sequencing question, I reach for INTERRELATE. This tool draws out the causal dependencies between parts, so you see which ones should be upstream from the others. While it looks kind of technical, I think it captures what an expert does in their head when intuitively deciding which problems to solve first.INTERRELATE starts with taking a set of input elements and arranging them around a circle. Then you draw an arrow from one element to another if doing that thing will help you do the other. After all the arrows are drawn, you can count the inputs vs. outputs at each element to judge which things have are causes of the rest of the system versus which things are more effects of the rest of the system. Things with more outputs than inputs should happen earlier in development.