There’s this great saying about giving first, giving often and giving with no expectation of getting anything in return. Brad and the Techstars teams have done a tremendous job of evangelizing this around the world. While I agree with all of this I’ve struggled a lot with how to have confidence that I’m giving to the right things.
Near the end of last year, time felt as if it was slowing down a bit (it wasn’t) and I kept thinking more and more about this thought process and where it would lead me.
The question I found myself asking is, so what do I give to? Some things are really easy. Founders, for example, are really easy to want to help. Pi515, also a no brainer. But, what else?
That question led me on a interesting journey. I kept coming back to multiple questions:
- What do I believe in that I want to reinforce?
- What has impact to drive change in a way that matters more than a statement?
- If I can find things to support, that reinforce what I believe and impact change beyond just words, how do I contribute?
- How do I find these things if I’m not already connected to them?
The initial question was easy to answer. The second turned out to be self-reflective. I discovered it was more about what I’m not connected to that drives the type of change I want to be a part of. The third turned out to be obvious once the first two questions were answered. The fourth turned out to be most interesting.
To answer the fourth I had to start thinking about what a good way is to analyze things that aren’t in my social circle and their connected impact.
To find things I didn’t know, I ironically started the list with who I knew — and proceeded to map their networks through what I could find about them online. My primary data sources were:
- Conference notes
- Angel List
I ended up with a decent sized model. Once I passed ~10K elements it started to get slow enough that it made adding more elements orders of magnitude more difficult so I stopped contributing as aggressively.
Once I got 3 or more degrees of separation I started drilling down on individual cities. This is where it got interesting. I started learning all kinds of things about connectivity between Minneapolis, SF, Detroit, Indianapolis and the list goes on and on. One of the discoveries was how all encased the startup communities are in SF, Boulder and NY. The density of the ecosystems amongst each city is much stronger than it is in Des Moines. This might seem like an obvious thing to say but it’s another to see it reflected in data.
The companies that go public seem to have funding from seed to IPO that is largely coming locally through Series C or further. This keeps a large chunk of the returns in the community when there is a big event. It also appears founders with 1 or more large exits tend to restart companies in the same cities rather than leave, which is something I can’t really confirm as a trend anywhere but Boulder, NY and the bay. The numbers are just so small comparatively speaking everywhere else.
I could probably write another post about the difference between cities but because I was focusing on my contribution in Des Moines, I came back to it as the central focus.
Using the sources above I was able to pull together a decent data set to drive down on what has impact to drive change by focusing on who has the impact to drive change. Figuring out the who and the people surrounding them was next.
The risk I was concerned with is making a big decision about supporting something that was actually me creating a spurious correlation that didn’t exist. Making a decision based on a inferred connection rather than a direct or proven one.
The other problem is that when I started asking people what to support, the answers were often rife with their own bias. Their own bias, the more I started to dig in, was likely unconscious and driven on their own personal decisions. Had I focused only on feedback directly in my closest network, I wouldn’t have ended up where I did.
There’s no harm in that, I do the same thing, but it further concerned me about taking certain recommendations on where to spend time based on third parties qualitative assumptions on how things were connected. I took some of that verbal feedback and included it in my model when it created real correlations that weren’t available publicly, which is why I’m also not publishing the model as a part of this post.
What someone felt created change vs. what someone was able to articulate about how that change came about surfaced a lot of ungoogleable things.
How to get an answer to a question I can’t google caused me to recall a conversation I had years ago with a friend, Paddy, who was talking about eigenvector centrality and how it was driving some of his organization’s thinking. At the time I remember being interested but hadn’t really had a practical reason to put it to work. It seemed like a good approach to try.
What this left me with was a series of connection points and I decided I’d try to use Paddy’s approach to understanding connectivity and centers of influence in the technology community in Iowa.
My central concern was finding connectivity in new communities and organizations on a local level that I wasn’t connected to — physically and financially.
I started with my own network and starting mapping from memory some of my own connections. At it’s most basic level, this is what my 1st degree network looks like:
Disclaimer: Even though I’ve since departed Clay & Milk and left it in capable hands, that was not the case when I took this screenshot. Also, I added TAI as a connection even though it wasn’t one when I began.
Before too long, I’d invested more than a few hours scraping data and pulling down connections that stemmed from this starting point to the point that it grew into a more meaningful model that spanned globally.
I made a decision early on to connect people to companies and organizations but not each other unless they were connected through marriage. This decision could have skewed the model but helped drive priority to companies and organizations which was my intent.
Since I was focused on leveraging eigenvector centrality here’s an excerpt describing it:
… measures how well connected an element is to other well connected elements. In general, elements with high eigenvector centrality are the leaders of the network, though they may not have the strongest local influence.
Thankfully Kumu, the tool I was using, makes this simple. Once you build the model by deciding on the elements, attributes, and connection types, the actual software does the rest. There are two views I focused on looking at:
- What am I not connected to directly already that has high connectivity elsewhere?
- What am I not connected to that I’d get excited to contributed to?
I am already contributing to the first. The second organization on the list was Technology Association of Iowa. The third was another organization that I’m still working to get on the board of.
So when someone asks me “Why did you join the TAI board” I can say the following honestly:
- I was woefully unconnected from the TAI network of companies and associated organizations setting the policy and creating the framework that drives tech innovation in the state.
- It has core tenants I believe in:
- Develop & Recruit Talent
- Diversity & Inclusion
- Public Policy
I also like the people. It’s rare and a lot of fun when you find good people working on something you believe in where there is also a fit for you to contribute.
This is the first time I tried to analyze my contribution in the community using maths as Paddy would say.
Even though the answer, to join a board, might be boring to a lot of people I had fun pursuing the question of how to invest more in the community. And I’m happy to say, I’ve enjoyed contributing to TAI.