Foothold

Revenue models & community economics

I had 380 paid members and a 23% activation rate. Four content interventions later, nothing moved. The problem was my revenue model.

By Foothold · October 2026 · See the five revenue model patterns and their activation stakes →

Marcus Chen ran The CMO Collective — a paid Slack community for Chief Marketing Officers and VP Marketing leaders at B2B SaaS companies, $49 per month, open enrollment, launched January 2025. He built the community the same way he built his newsletter: make it accessible, write consistently, trust that good content brings good people. Six months in he had 380 members, $18,620 in MRR, and a week-one activation rate that had not moved in four months of concerted effort to improve it.

The number was 23%. He had run four separate interventions to change it. None of them had worked. He had been treating the problem as a content and onboarding problem for months. The problem was his revenue model.

The four interventions that didn’t move the number

Marcus had been tracking week-one activation rate since month two — the percentage of new members who posted at least once in their first seven days. It started at 21%, which he attributed to the community being new. By month four it was 23%, which he attributed to being in a growth phase. By month six it was still 23%, and he started running experiments.

Intervention one: Biweekly “CMO Spotlight” AMAs. He reached into his network and booked a different CMO or VP Marketing every two weeks for a 45-minute live Q&A session in #ama-live. Attendance on the first one: 52 members. By the fourth one: 43 members. Good numbers, he thought. He cross-referenced the attendance list against his Slack admin export: 48 of the 52 first-session attendees were from the 23% who were already posting in the workspace. The AMAs were reaching his already-engaged members. They were not reaching the 77% who had never opened Slack that week.

Intervention two: A #wins channel with a weekly prompt. Every Monday at 9 AM a bot message posted to #wins: “Share your biggest win from last week — a campaign result, a hire, a number you’re proud of.” Week one of the channel: 18 replies. Week two: 14. Week three: 9. By week six it had settled at 4–6 replies per week, all from the same rotating set of 12–15 already-active members. The 77% who had not yet posted were not opening Slack to see the Monday prompt.

Intervention three: A revised welcome email. He rewrote the automated welcome email from a generic “Welcome to The CMO Collective!” with a channel list to a personal note about his own CMO journey, why he had built the community, and a single ask: Would you DM me one thing you’re working on right now that you’d want a CMO peer’s perspective on? He tracked replies over 60 days. Reply rate: 11%. Of those replies, he converted 8 into first posts in the workspace by personally routing each person to a specific channel with a specific thread reference. 8 activations from 200+ welcome emails was not the lever he needed.

Intervention four: A monthly CMO Toolkit resources digest. A curated email with five links, tool recommendations, and a short operator note from Marcus. Open rate: 31%. Click rate: 6.3%. He looked at who was clicking: 94% of the clicks in the first three months came from members who had already posted in the workspace. The digest was a retention tool for the already-engaged. It was not an activation tool for the never-activated.

After five months of effort, activation rate: 24%. One percentage point above where he had started.

What the Slack admin sort showed

He had been reluctant to run the admin sort because he suspected it would show something uncomfortable. In month six he ran it anyway.

Of 380 members, 294 had never sent a single message in the Slack workspace. Zero messages, ever. These were paying members — $49 per month, on average for 3.1 months each — who had joined the workspace, read the welcome post, and closed Slack. Some had opened it once more to look around. Most had not returned after the first visit.

He cross-referenced the Slack export with his Stripe data by member join date and email address: this took about two hours. Then he ran the cancellation data against activation status for the prior three months: of 68 cancellations, 63 were from the 294 never-activated members. The five activated-member cancellations were all cited reasons — two career changes, one budget cut, two who said the community had been good but they were stepping back from all paid subscriptions for a quarter.

Ninety-three percent of his churn was coming from 77% of his membership who had never posted.

The math of the problem was now clear. He had $14,406 in MRR from 294 members who had never used the product they were paying for. They were paying for two or three months on average, then cancelling. The activated 86 members were staying at a much higher rate and generating significantly more LTV. But there was something else in the data that he did not notice yet: even his activated members were churning at a higher rate than he expected for a community with good content and engaged discussion. The average tenure for activated members was 6.8 months — shorter than he thought it should be for people who had evidently found value.

The cancellation interviews

He sent a personal DM to 15 cancelled members. Not a survey link — a direct message: Hi [name], I noticed you cancelled The CMO Collective last month. I’m working on improving the community and I’d really value one honest sentence about why you left. No wrong answers, totally off the record.

Nine of the 15 replied. He read the responses twice before he understood what he was looking at.

None of the nine responses mentioned his content. None mentioned the AMAs or the resources digest or the welcome email. They were not talking about what he was producing. They were talking about who else was in the room.

The composition problem

He went back to the Slack admin export and sorted by member-listed job title. Of 380 members:

His stated community was “The CMO Collective — a peer network for marketing leaders.” His actual membership was 16% CMOs and VPs and 84% everyone else who called themselves a marketer and found $49/month a reasonable price to try a marketing community.

This was not an onboarding mechanics problem. His CMO members were opening the workspace, reading the #intros posts, scanning the threads in #demand-gen and #brand-and-messaging and #go-to-market, and concluding that the community was not their peer group. No day-0 DM was going to change that conclusion. The composition they encountered was a function of the membership selection mechanism — a $49/month open-enrollment model that signalled “anyone who calls themselves a marketer is welcome” rather than “this is a curated room of CMO-level practitioners.”

He had been trying to fix a composition problem with content interventions. The content interventions were reaching the 23% of members who had activated, nearly all of whom were his CMO and VP Marketing ICP. They were producing good engagement from that 23% precisely because those members had found their peers in the workspace despite the diluted composition. But the 77% of members who were CMOs and VPs who had joined expecting a different community, or who were directors and growth marketers who found the content too senior for their stage, were not a fixable population with better AMAs or a more personalised welcome email.

The revenue model diagnosis

The root cause was the revenue model. Marcus had built a high-volume low-price model in a niche that demands curation. The economic logic of high-volume low-price is volume: maximise the number of paying members, rely on peer content at scale to generate value, keep the price accessible to the full addressable market. That logic works when the value proposition is “access to a broad peer community of practitioners.” It breaks when the value proposition is “a curated room of C-level peers who can speak to your specific strategic challenges.”

The price signal was the problem. $49/month says “this is accessible to any marketer who wants to try a community.” His ICP — sitting CMOs at B2B SaaS companies paying $49/month for their own software tools, managing seven-figure marketing budgets — read $49/month as a signal about who else was in the room. A CMO peer network that charges $49/month is not curating. A CMO peer network that charges $299/month with an application is making a different claim about composition. The price point communicates the admission standard.

The low-volume high-price model would have aligned the signal. At $299/month with a screening application, the CMO considering joining would ask themselves: “If they are charging this much and screening applicants, the people inside probably match my level.” At $49/month open enrollment, the same CMO asks: “Let me look inside and see who’s actually here.” When they looked, they saw the composition data Marcus had been avoiding — and they cancelled.

The LTV math that made the switch obvious

He ran the LTV numbers for both models.

Current model ($49/month, 23% activation, 380 members):

Target model ($299/month, curated admission, projected 70% activation):

The current model produced $10,867 in cohort LTV per month from 55 new members. The target model would produce $49,636 in cohort LTV per month from 12 new members — nearly five times the LTV value from one-fifth the acquisition volume. Even accounting for the higher friction of a curated application process, the economics of the low-volume high-price model were not a close comparison.

The MRR picture was the same. 380 members at $49: $18,620. 92 curated members at $299: $27,508. The 52% MRR increase came from having one-quarter the members. That freed him from the growth treadmill — having to acquire 50+ new members per month just to stay flat against the 77% non-activation churn.

The eight-week transition

He drafted an email to all 380 current members explaining the change: The CMO Collective was moving to an application-based $299/month model in eight weeks. The reason, stated plainly: he wanted the community to be a genuine CMO peer network, which required admitting members with a consistent ICP profile. Current members were invited to apply for re-admission at no charge. Those who were not re-admitted, or who preferred not to apply, would receive a two-month refund.

He expected resistance. He received almost none. Responses from current members were mostly one of two types: CMOs and VPs who were relieved (“This is exactly what I was hoping the community would become — I’ll apply immediately”) and directors and growth marketers who were gracious (“I understand, good luck with it”). Three members sent emails that were mildly frustrated. None of the nine cancelled members he had interviewed sent anything negative.

Over eight weeks he ran 86 screening calls with applicants from the existing membership. His acceptance criteria: active CMO or VP Marketing role, B2B company at Series A or later, specific stated challenge or goal in the application that corresponded to one of his core channels, and a LinkedIn profile that matched the application claims. He accepted 92 members, all from the existing base. The remaining 288 received refunds.

New MRR on the first day of the new model: 92 × $299 = $27,508. That was 48% higher than the $18,620 MRR he had been running with 380 members.

Building the Day-0/3/7 sequence for the curated model

Switching revenue models improved the composition problem — but curated members still needed a structured first week. Even a CMO who had passed a screening call and paid $299 had the same blank-workspace problem as any new member: they joined the Slack workspace, saw the channel list, and had to figure out where to introduce themselves and what to do first. The application had captured their stated goals; the sequence needed to use that data.

He added four questions to the application form that fed directly into the Day-0/3/7 messages:

  1. What’s your current CMO priority? (Brand & positioning / Demand generation efficiency / GTM team structure / Board & investor relations)
  2. What stage is your company? (Series A / Series B / Series C+ / Post-IPO or PE-backed)
  3. What’s the specific challenge you’re working through right now? (Open text, 2–3 sentences)
  4. What would make this community worth $299/month to you? (Peer sounding board for specific decisions / Frameworks and resources from operators at similar stages / Introductions to CMOs in adjacent markets / All three)

He mapped each priority answer to a primary channel route: #brand-and-positioning, #demand-gen-efficiency, #gtm-team-structure, #board-and-investor-comms. A member who listed “GTM team structure” and described their challenge as “deciding whether to split growth and brand into separate functions as we scale to Series C” received a day-0 DM that cited the most recent thread in #gtm-team-structure about function-splitting decisions and asked one specific question keyed to their stated challenge: There’s a thread in #gtm-team-structure right now about separating brand from growth at Series B/C that maps closely to what you described — what’s your current thinking on the timing of the split?

The day-3 nudge went only to members who had not yet posted, and referenced the same channel with a lighter ask: Hope the first few days in the community have been useful. If you haven’t had a chance to check in on #gtm-team-structure yet, there’s a thread there right now that I think you’d have a perspective worth sharing.

The day-7 operator scorecard gave him a short list: three members who had not yet posted, their application stated challenge, their last-login date from Slack admin, and a suggested message tailored to their stated goal. He spent about 15 minutes on this each Friday morning, sending two or three personal follow-up messages.

Before fully automating the sequence, he ran a manual test with 30 of the newly re-admitted members whose applications had been flagged as high-ICP-fit but who had not yet posted in the first week. He sent personalised DMs using the application answers. Twenty-four of the 30 replied within 48 hours. Twenty-three posted within 72 hours. A 77% activation rate from the manual test.

What changed in five months

He tracked four numbers monthly: week-one activation rate, month-two renewal rate, weekly active poster rate, and MRR.

By month five:

The membership composition shift was the most striking change. CMOs and VP Marketing went from 16% of membership to 74%. When he opened the workspace now, the threads were different — about budget allocation, board narratives, the decision to hire a VP of Demand Gen versus a VP of Brand, whether to in-house performance marketing at Series B. These were the conversations he had intended to build the community around when he launched. They were not happening at $49/month open enrollment with 380 members.

The activated members who had stayed through the model transition were explicit about the difference. One CMO who had been in the community since the early days sent him a message: “Whatever you changed, it worked. The last three weeks have had more useful conversations for me than the prior six months combined. The room finally matches what you said it was.”

The two-layer activation problem

The concept worth naming from Marcus’s experience is the two-layer activation problem: most activation failures exist at one of two layers, and the fix depends on which layer you have.

Layer 1 is the composition layer — the revenue model problem. A community’s revenue model (who it admits and at what price) determines the composition of the room. When the composition is mismatched to the ICP’s peer expectations — when a curated-expectation niche is running on an open-enrollment price point — members open the workspace, see the composition, conclude this is not their peer group, and do not activate. No Day-0/3/7 sequence can change that conclusion. The sequence is sending a personalised message to someone who has already made a judgment about the room. The judgment stands.

Layer 2 is the mechanics layer — the onboarding problem. When the composition is right and members are genuinely in their expected peer group, a member who opens the workspace and sees relevant conversations may still fail to activate if they do not know where to start. The blank-workspace problem: 20 channels, a welcome post, and no clear first action. This is the problem a Day-0/3/7 sequence solves directly. A personalised day-0 DM that says here is the specific thread in the specific channel that maps to your stated challenge removes the blank-workspace friction and produces a 50–78% activation rate in communities with good composition.

Most paid community operators treat every activation problem as Layer 2 and invest in content and onboarding mechanics that reach only the already-activated minority. The four interventions Marcus ran — AMAs, the wins channel, the welcome email revision, the resources digest — were all Layer 2 solutions to a Layer 1 problem. They were well-executed. They worked for the 23% who were already activated and whose composition match meant they would have found their way anyway. They had no path to reach the 77% who had closed Slack after their first visit because the room did not match their expectation.

The diagnostic test that distinguishes the layers: run 20 personal DMs to non-activated members using the highest-personalization approach you can manage — goals cited from application data or LinkedIn research, specific live thread references, direct questions tied to their stated challenge. If that test produces below 30% activation, you are likely looking at a Layer 1 problem: those members have already concluded the room is not right and are staying inertially until their next billing date. If the test produces 50%+ activation, the problem is Layer 2, and a Day-0/3/7 sequence built on intake form data will replicate that result at scale.

For operators whose niche demands curation — executive peer networks, high-ticket mastermind communities, communities where the value proposition is explicitly “be in a room with the right peers” — the low-volume high-price model is not just an option. It is a precondition for the community ever working. The pricing strategy decisions that follow — what price to charge, how to structure trials, annual versus monthly billing — layer on top of a composition decision that the revenue model makes first.

Marcus ran four good content interventions and learned, five months later, that the lever was something none of those interventions could move. The activation number did not change because the composition number had not changed. Both changed only when the revenue model changed.

Frequently asked questions

How do you know if your activation problem is a revenue model problem versus an onboarding mechanics problem?

Four signals point to a revenue model (Layer 1) problem: (1) activation rate stuck below 25% despite multiple content interventions that are known to work; (2) cancellation interviews citing “not my peer group” rather than “couldn’t find value”; (3) activated members churning at months 5–8 even when early engagement was strong; (4) a price point accessible to a wider audience than your ICP. The fastest diagnostic: send 20 highly personalised DMs to non-activated members. If activation from those 20 is below 30%, the problem is likely composition (Layer 1). If it jumps to 50%+, the problem is mechanics (Layer 2) and a Day-0/3/7 sequence will fix it at scale.

What does the transition from high-volume low-price to low-volume high-price look like operationally?

The transition has five steps: (1) Announce the change 6–8 weeks ahead with a transparent explanation of why. (2) Offer all existing members the option to apply for re-admission at no charge. (3) Set objective ICP criteria and run screening calls or application reviews. (4) Accept the members who qualify; refund the rest a defined period (Marcus offered two months). (5) Relaunch at the new price with an application form on the landing page. Expect to retain 15–25% of current membership — but MRR often increases if the per-member price increase is large enough. 92 members at $299 generated more MRR than 380 members at $49.

Why do high-volume low-price communities with acceptable activation still see elevated churn at months 5–8?

Even when 30–40% of members activate in week one, if the community composition is mismatched to the ICP’s peer expectations, those activated members encounter a relevance cliff at months 3–6 as the network effects fail to compound. In a CMO-focused community where only 16% of members are actually CMOs, an activated CMO spends their first month engaging — but over months 3–6 finds that the peer relationships they hoped to build are thin relative to the general-marketer majority. The month-5 to month-8 churn that results is a composition problem, not a content problem. In properly curated communities, activated members tend to have 85–92% month-two renewal rates; in high-volume communities with composition drift, that number settles at 55–68% even after fixing the week-one onboarding mechanics.

Can a Day-0/3/7 onboarding sequence fix a revenue model mismatch?

No — but running the sequence as a manual test is the fastest way to determine which problem you have. Pick 20 non-activated members, send highly personalised messages citing their specific stated interest and a relevant live thread. If activation from those 20 is below 30%, the members have likely already decided the community is not their peer group; the sequence cannot change that conclusion. If activation jumps to 50%+, the composition is fine and a structured sequence will replicate that result at scale. A Day-0/3/7 sequence applied to a composition-mismatch community will move activation from 23% to perhaps 30–35% — a real improvement, but not the 65–78% rates operators achieve when the composition is right and the sequence removes only the blank-workspace friction.

Related: Paid community revenue models — the five economics patterns and activation stakes for each · Paid Slack community member LTV — the segmentation test that changed Priya’s investment priorities · 18 months without measuring activation — and what Hiroshi found when he finally did · Paid community pricing strategy — what to charge and how to structure trials.