Experiments and lessons learned from small town ecosystems

In the military world in which I spent most of my professional life as a spouse prior to Michigan, networks are distributed, mobile, and digital-first by necessity. You could be living in Japan, Florida, or Kansas in four months, so you build connections that can transcend distance. This turned out to be helpful working in Michigan’s startup ecosystem while living away from the central hubs.

Living in the town my dad’s family has called home for generations means my life is fundamentally regional. Once or twice a week, I commute two hours to Detroit at 6 a.m. And once or twice a week, I’m driving an hour to Ann Arbor, MI or Toledo, OH for events, grocery shopping, or personal errands I can’t find close by. Working regionally and crossing these geographic lines regularly is a rhythm that’s affected how I think about startup ecosystem work, and the opportunities hidden in plain sight. 

For those of us living in more rural areas, the time, energy, and resources to stay right in the middle of startup ecosystem conversations are not always possible. So this is where online tools and structure can bridge the gap. 

Here are two experiments I ran that showed me what that takes.

‍

What 524 people can tell you about where to start

I've had informal conversations with local entrepreneurship leaders about what it would take to help more people in our county get real access to AI and emerging technologies. The early version of that conversation was the version everyone has: how do we teach everyone here about AI. Broad and somewhat fuzzy in terms of execution. It’s a big task!

Perusing open source data one day, I found the Workforce Intelligence Network employment data on my county, things became more actionable. What I found: roughly 524 people in our county currently work in IT positions. And that represented an increase of 32% over a five-year period. What did this mean?

In conversations in the larger metro areas, talk about AI agents, OpenClaw, and frontier models is commonplace. But where I live, I couldn’t seem to find a place to have those conversations. In day-to-day life, they don’t seem necessary with how things operate here. The WIN data, however, showed me that at least 524 people within this geographic radius are likely to be asking these questions or building with these tools in some capacity. So this shifts the conversation from how do we get technology to this community? To: how do we create structure for the people already building to find one another and collaborate?

One day per week, I walk to a coffee shop downtown, drink a black coffee and clack away on my laptop. This past visit, I looked up and saw about five others, heads down in their laptops, focused and working. This is our pop-up, smaller-town version of a coworking space. 

I tested my assumptions and set up a coffee gathering for people who are actively building with AI locally. Through this, I discovered an international athletic professional development training company building their own servers and doing highly sophisticated AI translation that cut their production time from months to minutes with high accuracy — about two miles from my house. 

That's the actual model, and it travels: the knowledge and people are there, in some form. They’re just hidden in plain sight. Find the existing concentration of expertise in your area — however small it looks next to a hub's numbers — and build the connective structure for that group first. 

‍

Six lessons learned from five distributed communities in action

Physical proximity still has its limits. This is where distributed, online communities can close the gap. Over the past six months, I did a Slack-specific experiment to see distributed communities in action. I joined five different active Slack communities to experience what types of things work in keeping communities engaged in these spaces. Two were related to the startup ecosystem where I’m in grad school, one was for revenue operators, one was for community operators, and one was for nonprofit professionals. Same tool. Five very different vibes.

Here are a few takeaways about what works and what stalls:

• Purposeful friction can create a better experience: One community provides immediate entry with one click. This was easy to access, but it also meant the sheer number of people joining the community made it hard to decipher where and how to find value. In contrast, the communities with slightly more friction or delays in the process, such as one human-augmented check or even 24 hours lag time, helped filter out who was in the digital room. This filter helped maintain the value.

• Onboarding and retention automation: One on-brand community for revenue operators leveraged onboarding automation, so that as soon as I landed in the community, I knew exactly what my next steps were, what I can and should post, and how to participate. When I fell inactive after a period of time, the founder of the group sent a direct message to ask about the experience and what would add value. I assume this was automated, but the direct 1:1 outreach cut through and brought the group top of mind. 

• Bridging to real-life: A Slack for community operators showed me the power of leveraging in-person encounters between a few to bring a sense of energy and connectedness to the whole group. As Slack members participated in the annual summit, they posted pictures, takeaways, and shared resources. So even those who could not be physically present in the room could still have visibility into the conversations happening even states away. Photos shared turned the group from a series of usernames to real people with whom to engage.

• Leveraging human hosts: The same community that built out many automations to help moderate the group leveraged human-hosts. These hosts would help curate the week’s posts into roundups (probably AI augmented), and facilitate active, ongoing conversations between members, tagging them, sharing additional links and resources, and elevating important conversations. And on a much smaller-community scale, one other group was small enough to not require automations. The primary host of this group messaged me when I joined, letting me know the etiquette for introductions and the benefit of reviewing past posts to see how conversations happen there. This insider-wisdom helped me know how to avoid missteps and show up well.

• Crowdsourcing situational wisdom: The most valuable element of each of these groups was less sharing links and events, and more the hard-won industry lessons you won’t ever see on a LinkedIn post or discussed at a networking event. Someone can post a specific, complex challenge, and get quick, real-time insights from peers or others further down the field about what they’ve actually tried in real life. I think the challenge lies in how to moderate other types of content to ensure these conversations take center stage. 

• Is Slack the tool? I’m not 100% convinced, but it is tool. I think I realized posting in a Slack community is like standing at the mic in front of an auditorium of people you don’t know, with professional stakes. The more clear it is how you fit in that room, the more confidence you’ll have in speaking. So the larger the number of people in the room, the more structure is needed to keep it valuable. 

This experiment showed me that purposeful friction to entry, moderated volume of participants, clear rules of engagement, and retention check-ins can create the structure that makes these spaces more valuable. 

There are a lot of these communities out there, but it looks like many tend to be city-based. But the state-wide community has precedent: Maybe I can convince the Ohio Startup Network to let a Michigander like me into their new community to see how it’s working.

‍

Why it matters

I didn't expect that the same instinct which helped me move without knowing where I would live, would be the same instinct needed to engage in a statewide ecosystem from a rural home. But it is. More tools exist than ever to close geographic gaps. Remote-first tools can create opportunities for two-way conversations between major hubs and towns far-removed from them. Whether in a coffee shop, thanks to open source data, or in companies I hadn’t noticed down the street; the closer I look, the more value becomes visible in the small town where I live. I’d love it if the major hubs could get a glimpse of the good here, too.

‍