One of my first insights into this kind of project is that 'network mapping' or 'social network analysis' or any other kind of labeling means NOTHING to most people. And even if you show them an example, a simple generic map with a few nodes and some names of fictional persons - it still means NOTHING. It's a big 'So What?', and an even bigger 'Why would we want that?'.

What engages people is thinking about the connections in THEIR OWN network. And even then, the visualization light only turns on when there is something for them to LOOK AT. When we first started mapping, most people were introducing the subject with yarn and post-its - which still seems like the best place to begin, because it's tactile, hands-on, happens in real space, and emerges from collective activities and conversations.

My frustration with post-its and yarn was that a) it could get really messy, esp. when you have a wall that tape & post-its have a hard time sticking to b) it had to get taken down, it was a brief, one-time thing that usually couldn't be returned to. It got across the network concept, but wasn't much of a tool beyond that.

And then there was this huge gap from yarn & post-its to SNA tools - where there was nothing to see until the whole project was done. Even if the decision-makers were all-in with only yarn & post-its and some generic example pictures, most of the network didn't get it & was often suspicious of yet another request for them to generate data. It was an uphill slog to get people to participate in the mapping exercise.

So right away, I started mapping whoever was interested and didn't worry about the rest of the group. I figured, if we got a handful of people (and they're often, naturally, the better connected among the network, because the topic interests them) to be on a map, and if we then made that map in Kumu, and enabled those who were ON the map to share it with their peers - there would be a natural gravity to the process. Others would want to show up, too - they'd want to show how they were connected in meaningful ways as well.

But of course, that meant you had to have tools that made that easy. It wouldn't work if the data-gathering to mapping routine was too big a hassle - because no-one would be willing to repeat it several times if it was.

So Tim made an easily-repeatable data-wrangling process for all our projects and we found that approach worked great. We'd just start was with whoever was willing and able to start, and let the rest of the network come along at their own pace. And once we could see how key that was, we designed sumApp to make our approach easy.