Running a cohort program is easy. Running one that actually moves the needle for entrepreneurs, filling seats and producing outcomes you can prove, is a different job entirely.

Karen Collins knows the difference. She's the Director of Programs at LAUNCH Chattanooga, running multiple cohort programs including the Startup Matrix Accelerator and the Kitchen Incubator of Chattanooga, the only freestanding kitchen incubator in Southeast Tennessee. Before joining LAUNCH, she spent years in city government, ran the Homeless Coalition, wrote federal grants, and went through LAUNCH's own program herself as a participant before joining the team.

We brought Karen together with EIC's David Oliver and Taylor Sherbine for a conversation on what separates high-impact cohort programs from ones that just fill seats. Here's what she shared.

Start with community need, not program design

The most successful cohort programs are built in response to something real happening in the community. Before designing a single session, Karen's advice is to stop and look outward.

"Before you start building it, look in your community to see what the needs are. Look to see what is already being done. Don't reinvent the wheel. Try to fill the gap of what's not being served, and build your program that way."

For LAUNCH Chattanooga, that gap revealed itself during the pandemic. The organization had been running its Startup Matrix Accelerator for years, a general entrepreneurship cohort. But conversations on the ground made clear that food entrepreneurs had specific needs no one was addressing at scale. The result was the Kitchen Incubator of Chattanooga. It wasn't an easy call. One board member actively advised against it. He later admitted he was wrong.

Three questions to answer before designing any cohort:

1. What are entrepreneurs in your community struggling with that existing programs don't address?
2. What are policymakers and funders asking for?
3. What's already being served well, and what gap are you uniquely positioned to fill?

How to build a cohort that sets participants up to succeed

Who you put in a cohort shapes every outcome that follows. The 'just put butts in seats' approach creates noise that drowns out real signal. Karen's approach to participant selection is methodical, and it starts long before anyone sits in a room together.

Design your application around your funder's requirements. If you've committed to serving low-to-moderate income entrepreneurs, minority-owned businesses, or food and beverage founders specifically, your application needs to capture those data points at intake. "Make sure those folks kind of fall into those buckets," Karen says. "That is not to exclude anybody, but that's just to make sure you have the things you need."

Use a rubric or structured interview to evaluate strength. When you have more applicants than seats, a scoring rubric or set of readiness-focused interview questions makes the process fair and defensible.

Set expectations clearly before anyone accepts. Time commitment, attendance requirements, what participants will and won't walk away with: communicate all of it upfront. Many of your applicants are still working full-time while trying to launch a business. Make the ask explicit early.

How to design a data strategy before your first cohort begins

The programs that can prove their impact to funders five years from now are the ones that designed their data strategy from the start, not the ones trying to reverse-engineer outcomes after the fact.

Karen has been on both sides of this. She's judged grant applications as a funder. She's written federal grants. And she's been the program director walking into an organization with historical knowledge spread across three different platforms, with no clear way to connect the dots.

"A lot of times what happens when you're new to an organization is that historical knowledge leaves when that person leaves. You're like: who was that person you worked with five years ago? And nobody knows."

Three questions to answer before your first cohort launches:

1. What metrics has your funder committed you to tracking?
2. What does success look like at one year out? Three years? Five?
3. What do you actually have the capacity to track consistently over time?

LAUNCH Chattanooga uses Catalyzer to consolidate program data, combining intake applications, business coaching notes, cohort attendance, and longitudinal client tracking into one place. One of Karen's custom fields: What brings you joy? A reminder that the data you collect can reflect the humanity of the work you're doing, not just the outputs you owe a funder.

What to do when participants drop out of a cohort

Dropping out of a cohort doesn't mean a founder's entrepreneurial journey is over. The organizations that treat it that way lose the chance to re-engage some of their most motivated future participants.

Karen plans for one or two participants to exit every cohort. After someone leaves, her protocol includes:

• A check-in at four weeks: not just to follow up on program logistics, but to make sure they're okay as a person
• Keeping their record active: they stay in the system so re-engagement doesn't mean starting over
• A waitlist and re-engagement pathway, so when the next cohort opens, they're already in the pipeline, not cold
• Alumni events, where former participants, including those who didn't finish, are invited back into the community

"Never close the door on their dream of entrepreneurship because it didn't work in your particular organization," Karen says. "If your front door is not the door they should be walking through, give them a referral to another program that may be a better fit."

One of her participants from the last cohort who dropped out reached back out and is now enrolled in the current one.

How to use AI in your cohort curriculum

AI tools can genuinely help your facilitators and your founders, but they can't replace community knowledge.

Karen's take: "If you're going to use AI to strictly create your curriculum, go back over it with a fine-tooth comb. That curriculum may not be what your community needs. It may not cover the needs of that community."

Where she does see clear value: AI for transcription, note-taking during coaching sessions, and directing training within cohorts. "Those kinds of things are really, really helpful for our entrepreneurs. But you have to understand: AI is a tool. It is not a be-all and an end-all."

The risk with AI-generated curriculum is generic output. If your program is built around the specific needs of food entrepreneurs in Southeast Tennessee, or retiring corporate executives in suburban Ohio, or immigrant-owned businesses in South Florida, no general-purpose AI model is going to know what your community actually needs. Someone has to bring that knowledge. The AI can help you move faster once you know what you're building. It can't tell you what to build.

How Catalyzer supports cohort management end to end

The operational challenge Karen describes: managing applicants, running intentional selection, tracking session engagement, following up with dropouts, and reporting outcomes to funders over five years. That's exactly what Catalyzer's cohort management feature was built for.

Intake and application: Program-specific application forms link directly to client profiles in Catalyzer. When a returning participant applies again after two years, their record updates automatically, so you're tracking them longitudinally, not starting over with a blank profile.

Participant selection: Review applications inside Catalyzer, accept participants directly into a cohort, and move waitlisted applicants into a holding queue with easy email communication to keep them engaged.

Session attendance: Take attendance manually or let participants self-check-in via a link or QR code. Facilitators can track who's present and identify patterns of disengagement before a participant goes quiet entirely.

Dropout protocol: Keep inactive participants in a waitlist cohort with a note on their record and a clear re-engagement pathway for the next cycle. No one falls through the cracks because they're still in the system.

Impact reporting: All cohort activity rolls up into program-level goals. When a funder asks what percentage of your cohort was minority-owned, or what the average revenue change was at 12 months, the data is there, because you collected it intentionally at intake.

If you're currently managing cohort applications in a form tool, tracking attendance in a spreadsheet, and sending follow-up emails out of your personal inbox, Catalyzer gives you one place to do all of it. For free.

The bottom line

Running a high-impact cohort program isn't about having the best curriculum or the most weekly sessions. It's about knowing your community, selecting participants with intention, building a data strategy that survives staff turnover, and staying connected to every founder, even the ones who don't make it through.

Karen Collins has been on every side of this work: funder, grant writer, participant, and program director. Her message is consistent: do the discovery first. Build for the gap. Serve the whole person. And make sure your systems can back up every outcome you claim.

EIC hosts regular webinars with practitioners like Karen. Subscribe to our newsletter to stay in the loop, or start your free Catalyzer account and see how it fits into your existing workflow.

‍