Clay Built a $3B Ecosystem by Not Building One. Here's the Framework.
A warm welcome to Lonelle, Patrick, Mark, Cam, Joaquin, Greg, Matt, Sarah, Ben and the 38 other new subscribers this week. I see you and I appreciate you!
Clay Clubs meet in 90 cities around the world and 20,000 Clay users contribute to its Slack community. It’s almost an understatement to say that the company’s $3B valuation grew from its ecosystem.
What it became is far different from how it started.
Clay didn’t start with a partner program. There were no tiers or certifications. For years, their entire community infrastructure was just a Slack channel they forced people to join.
The vibrant ecosystem sprouted and grew because of Clay’s approach. CEO Kareem Amin even told his team to engage with their product, customers and community like gardeners. “We’re nurturing growth, not directing it.”
I’ve spent dozens of hours studying Clay alongside Notion and Figma, two other brands with world-class customer communities. The big difference is that Notion and Figma are horizontal tools with massive addressable markets, and the lessons aren’t universally applicable.
Most of us are building brands that don’t fit that model. Clay’s playbook is more transferable because it built around a narrow ICP, and that constraint is actually what made it work.
At Recurrent, we think a lot about our customer community and I’d put it up against any other company our size. But ours looks like a seedling compared to Clay’s thriving forest.
After tons of research and years of lived experience, here’s the 4-phase sequence for knowing when and how to build an ecosystem for emerging products and services startups.
The Gardener’s Sequence
Most ecosystem case studies focus on the tangible outputs: Slack groups, ambassador programs, the templates. Less has been written about what had to be true about the product and the ICP before the community could form and compound.
That’s the gap this framework attempts to fill.
1. Prepare the Soil
Before Clay had an ecosystem, there was a constraint: a narrow user base with lots in common and incentive to learn together.
This wasn’t accidental. Clay kept a waitlist active even after scaling past millions in ARR. Co-founder Varun Anand said they dropped the waitlist at one point, but “one of the best decisions we made in the early days” turning the waitlist back on. It gave them the ability to be hyper selective about onboarding a narrow customer persona. Every customer had lots in common.
They also made a choice that seems borderline reckless: Clay stopped providing email support and web chat. The only way to get help was a public Slack channel. Every early customer was onboarded with a 30-minute Zoom call, and Varun wouldn’t let them off the call until they joined the channel.
This did three things at once:
It limited the frivolous support requests since people were surrounded by 200 peers.
It created a space for participation and collaboration among peers.
It helped customers feel seen because everyone in the channel had similar challenges and the same goals.
The Slack channel organically evolved from a support channel to a peer group. Customers were learning from each other.
Test it: Imagine a room filled with only your most engaged users.
How much do they have in common?
Do they share enough context to help each other without your involvement?
What would make them more likely to be collaborative vs competitive?
Lesson: A homogeneous early community is the soil that everything else grows from.
Common misstep: Don’t worry about community size here. Clay’s patience to stay narrow is what created the value that kept people coming back.
2. Spot the Organic Growth
Clay’s early ICP was agency owners. These people had clients, so they could pull through the product and Clay could grow with them. With their curated and narrow customer base, Clay was able to pay close attention to them and began noticing how they talked about the product:
Agency owners used LinkedIn to show off their creative Clay workflows and position themselves as experts in the emerging tech. Sharing about Clay helped the agency attract more customers, which helped Clay add more brands to its platform.
Power users went deep in private WhatsApp groups. GTM people wanted to learn with peers but did not want to share every trick publicly.
Slack users answered questions in the channel before the Clay team could respond.
People promoted themselves as “Clay experts.” If people could build a career on Clay expertise, those people would become motivated advocates with aligned incentives.
The infrastructure only works if organic energy already exists. The Slack threads, LinkedIn posts, WhatsApp groups, and self-appointed experts are only examples of signals that tell you the soil from Phase 1 is fertile.
Test it: Make a list of the places that people talk about your product or the problem that you solve. Social listening tools like Sparktoro or Google Alerts can be a good start, but they are not great at identifying uncommon or private channels. That’s where it’s helpful to talk to customers and query the collective wisdom of AI tools to flesh out your list.
Lesson: You can’t manufacture organic energy. But you can miss it. The signals are often in places that won’t show up in your dashboards.
Common misstep: Don’t interpret current customer silence as a signal of long term potential.
3. Feed What’s Growing
Clay founders (and an industry influencer early hire) started partnering with customers behind the scenes:
Telling them about new features
Sharing lessons to help them win new business
Creating personalized content for them to share
Clay realized UGC is most powerful when people are creating something, because then they can share what they made. Since Clay is a tool for building workflows, it gave their audience of builders something to show off. The content wasn’t testimonials or reviews. It was “look what I built.”.
Clay spent “a considerable amount of time supporting our experts with their own marketing campaigns, from promoting new business lines, to webinars, to one-off promotions,” the Clay team explained. They even use internal software to help creators design personalized videos about Clay features in their own voice. “At the end of the day, we want to help them look legit because if they grow, we grow.”
That last line is the governing principle of this phase. You’re investing more in your community’s success than you’re extracting from it.
Test it: If organic signals from the previous stage to find where you can enable those people without directing them. What would it look like to help your most active customers grow — their business, reputation, whatever?
Lesson: The key to success is aligned incentives, like we found in the Tesla referral program. You help your customers grow, which helps you grow.
Common misstep: Clay’s “gardener” philosophy nurtured authentic excitement at a time when structure and guidelines would have smothered it. Don’t try to over-engineer UGC.
4. Build the Trellis
Only after organic growth was undeniable did Clay formalize the ecosystem with the structure required to scale it.
Programmatic Community: Clay Clubs are events with a dedicated non-employee host. Clay helps them find a venue, covers the cost, and assists with promo and attendance.
Incentivized Content: Clay Creator programs gave customers formal reasons to post about Clay, but the real incentive was that the content helped customers grow their own businesses.
Credentialed Expertise: A partner program launched with escalating benefits and requirements, including affiliate rev share so partners earned from bringing on new Clay customers.
Official Role: Clay created a “GTM Engineer” job title, posted public job listings, and watched the market adopt it. Today there are hundreds of GTM Engineer postings at companies like Cursor, Webflow, and Notion.
Literal Alignment: Clay let its community invest at the same $1B valuation as top VC firms like Sequoia. Priority went to community members that reinforced the benefit of the ecosystem, both today and in the future.
Each of these moves would have totally bombed if they launched too early. The partner tiers work because there are enough successful partners to populate them. The expert directory works because the experts already had credibility. The equity offering works because the community already felt ownership.
The trellis simply supports growth that already exists.
Test it: It’s only after your customers have reached diminishing returns in Phase 3 that Phase 4 should be explored. Specifically, try to identify:
Where could program formalities help customers extract additional value?
How could structure deepen your relationship and further elevate with power users?
What would be the most minimum viable and most extreme version of each program? (Clay let people invest in the company.)
Lesson: The infrastructure should formalize what’s already happening.
Common misstep: In the Test It step above, I intentionally referenced your customer’s diminishing returns, not your company’s. The signal you’re looking for is not your own scale or efficiency or margin.
Gardening > Engineering
Most ecosystem programs fail because they try to start at Phase 4. Six months later, the partner program has 12 members and no momentum so it gets sunset.
Take inspiration from Clay that ecosystems are most vibrant when cultivated.
They constrained their community until it was dense enough to sustain itself. They watched for organic signals instead of manufacturing them. They invested in their users’ success before formalizing any program. And only when the growth was undeniable did they build the infrastructure to support it.
“How do I build a partner program?” ❌
“Is my product and ICP ready for one?” ✅
The companies that get this right won’t be the ones that copy Clay’s partner program. They’ll be the ones that copy Clay’s perspective.
Learned so much from this breakdown, their team seems so cool and it all seems so straightforward in hindsight. Love the forced joining of the Slack channel in the early days. And never knew they started the GTM Engineer, wonder if they thought it'd take off like this?