Foothold

Metrics & engagement

I fixed activation. My monthly engagement rate was still 38%. When I ranked every member by their individual contribution rate, I found 68 people who had activated and gone quiet by month three.

Yuki Tanaka ran The Independent Consultant Network at 5.8% monthly churn despite 62% week-one activation. She had already done the activation audit, built the Day-0/3/7 sequence, and watched activation improve. When she stopped looking at her community’s engagement rate as a single aggregate number and started looking at it as 200 individual scores, she found a segment her activation metrics had been hiding from her the whole time.

Published 22 October 2026 · 12 min read · Reference: Slack community member engagement rate — formula, benchmarks by community type, and what to do when it’s low →

The community: The Independent Consultant Network

Yuki Tanaka launched The Independent Consultant Network in February 2025. The positioning: a paid Slack community for independent management consultants and strategy consultants who operate outside the Big Four and MBB, $149 per month. The specific gap she had felt herself after going independent three years earlier: there is no room where independent consultants can talk candidly about the operational layer of their practice — pricing models, contract structures, client pipeline management, the difference between what a Big Four alumnus is expected to know and what nobody actually teaches you about running a solo or small-team practice.

The initial membership came from her personal network on LinkedIn, where she had been writing about independent consulting for two years. She had 52 founding members in February 2025 and was adding 18 to 22 new members per month through a combination of organic referrals and her LinkedIn audience.

Month eight: the activation fix that didn’t fix the churn

By October 2025, the community had grown to 200 members at $149 per month: $29,800 MRR. Monthly churn: 5.8%.

She had diagnosed and treated the activation problem in April 2025. An audit of her Slack member list had found that 58 of 130 members at that point had never posted a single message — 45% of the total membership. She had run a 20-member manual DM test using intake form data from her checkout process (primary consulting focus area, biggest current practice challenge, community goal), achieved 75% posting within 72 hours against a 34% baseline, and built the full Day-0/3/7 sequence. Week-one activation had climbed from 34% to 62% over the six months since the sequence launched.

Monthly churn had dropped from 8.9% to 5.8% in the same period. The activation work had been worth it. But 5.8% still meant she was losing 11 to 12 members per month. Net growth continued, but slowly. And she could not find an obvious next lever.

She had been following the conventional diagnosis order for low engagement: fix activation first, then look at content. She had fixed activation. Her content programming was reasonable — a weekly “what are you working on” thread, a monthly deal-structuring roundtable, a #proposals channel where members shared anonymised proposal structures and pricing decks. She was posting three to four operator-initiated threads per week. None of it was moving churn below 5.5%.

The aggregate that hid the distribution

In late October 2025, she used the formula from the Slack community member engagement rate reference card to calculate her monthly engagement rate for the first time:

Monthly engagement rate = members who posted or reacted at least once in the last 30 days ÷ total paying members × 100.

From Slack analytics (Admin → Member analytics → Active this month): 76 members had been active in the last 30 days. Subtract 4 staff and test accounts: 72 paying members. Total paying members from Stripe: 200. Engagement rate: 36%.

The reference card’s benchmark for a knowledge and peer-learning community at her price point: 45 to 55%. She was below the lower end of healthy.

She had run the three-stage diagnostic from the reference card. Stage 1 (activation rate): above 50%, so not the root cause. Stage 2 (operator content cadence): three to four operator threads per week, more than two-thirds conversational. Cleared. Stage 3 (channel structure and topic relevance): she had nine channels, well below the sprawl threshold; the content had remained focused on the core independent consulting use cases since launch. No obvious structural failure.

The diagnostic said her engagement problem was not activation, content cadence, or channel structure. But the number was still 36%.

She put the reference card down and looked at the aggregate rate differently. 200 members. 72 active in the last 30 days. That meant 128 members had not posted or reacted in the last 30 days. A hundred and twenty-eight people paying $149 per month who were not engaging.

She had been thinking about that 128 as a percentage of the total. She had not been thinking about it as 128 individual people, each with a different history in the community and a different distance from their renewal decision.

Building the per-member engagement ranking

She downloaded the Slack workspace export and built a simple per-member engagement score for all 200 members.

The score used three inputs, all pulled from the workspace export JSON files and her Slack member activity CSV:

Maximum score: unlimited in theory (for highly active members), but practically 90% of scores fell between 0 and 30. She ranked all 200 members from highest to lowest and divided into quartiles: top 25% (50 members, scores ≥12), second quartile (50 members, scores 5–11), third quartile (50 members, scores 1–4), bottom quartile (50 members, scores 0).

The distribution was steeper than she expected.

Fifteen percent of her membership was generating nearly three-quarters of the community’s visible activity. The aggregate engagement rate of 36% was not evenly distributed — it was being held up by a relatively small number of highly active members while the majority participated at low levels or not at all.

This was not itself surprising. Heavy-user distributions are common in online communities. What concerned her was what she found when she looked at the bottom quartile by tenure.

The activation shadow

She added a tenure column to her spreadsheet: how many months had each member been in the community? Then she filtered the bottom-quartile list (50 members with zero posts in the last 30 days) by tenure.

The filter result:

She cross-referenced those 36 members against her activation records from the Day-0/3/7 sequence. Of the 36, 29 had posted an introduction in their first 7 days — they were recorded as activated. Thirteen had received peer replies on their intro posts. Some had been visibly active in channels for their first few weeks. All had gone silent somewhere between weeks 4 and 20.

She had also checked the third-quartile list (50 members with 1 to 4 posts in the last 30 days) and found a similar pattern: 32 of the 50 were in the 2-to-9-month tenure range, and 27 of those 32 had posted an intro in week one.

In total: 68 members with 2 to 9 months of tenure who had activated in week one and whose engagement had faded to near-zero by the time she was doing this analysis. She named this segment the activation shadow: members who had crossed the week-one activation threshold but had never developed an ongoing engagement pattern, and were now paying for a community they were not participating in.

These 68 members were invisible to her activation tracking. Her activation rate measured week-one behavior. All 68 had passed. They were invisible to her new-member cohort analysis. Most had been in the community for more than 90 days; they had already cleared the month-3 renewal at whatever rate their week-one tier had predicted. They had slipped into the passive-paying category without triggering any alert in her existing monitoring.

She was about to find out what their renewal rate actually was.

What the renewal data showed

Of the 68 activation-shadow members, 40 had upcoming renewal dates within the next 90 days. She tracked those 40 members forward.

At the 90-day mark: 11 of 40 renewed. 29 cancelled. 28% renewal rate.

For comparison, she checked the top-quartile members with upcoming renewals in the same 90-day window: 39 of 44 renewed. 89% renewal rate.

The bottom quartile of her per-member engagement ranking was renewing at one-third the rate of the top quartile. A 61-percentage-point gap in renewal probability, explained entirely by their current engagement score, not by any property of when they had joined or how they had been onboarded.

She had been losing 11 to 12 members per month. Looking at the math backward: if the activation shadow was renewing at 28%, and there were roughly 68 of them in the community at any given time with an average tenure of 5 months, she was getting about 14 cancellations per quarter from this segment alone — roughly 4 to 5 per month. That was 35 to 45% of her total monthly churn from a segment she had not previously identified.

The other 55 to 65% of churn was coming from members in the bottom quartile who had never activated, from top-quartile members who had departed for non-engagement reasons (budget cuts, career changes, company paid for it and then stopped), and from sporadic mid-tenure exits across all quartiles. The activation-shadow segment was not the only churn source, but it was the largest single identifiable category and the one with the clearest intervention path.

The intervention: re-routing, not win-back

Her first instinct was a win-back message: acknowledge the silence, offer something specific, invite them back. She tested this on 8 activation-shadow members in November 2025. The message structure: “I noticed you haven’t had a chance to be active lately — is there something specific you were hoping to find in the community that you haven’t come across?”

Three of 8 responded (38% response rate). One returned to active posting. Two had work reasons for the silence and said they intended to re-engage but didn’t in the 60-day follow-up window. Five didn’t respond.

A response rate of 38% was not bad. A 12% re-engagement rate from the message was not good enough to justify the approach at scale.

She thought about what the activation-shadow members were experiencing. They had activated. They had found the community relevant enough to post at least once. Something had happened between week 1 and month 3 that had reduced the community’s relevance to their day-to-day work — not catastrophically, just below the threshold where opening Slack felt worth the friction. The most common version of this, in her mental model: their immediate consulting project when they joined had ended or evolved, the channels they had been active in when that project was live were now less directly useful, and they had not re-discovered a new entry point that matched their current situation.

Win-back messages acknowledge the gap. They don’t solve the entry-point problem.

She tried a different approach: re-routing. Same personal DM format, but instead of acknowledging silence, she treated the member as currently active and pointed them to a specific current conversation.

The re-routing message structure:

She tested this structure on 12 activation-shadow members in November, pulling their original intake form responses (she had kept them all in a spreadsheet since launch) and matching each to the most active current thread in their closest channel.

Results: 7 of 12 posted within 72 hours. 58% re-engagement from a single message. Three posted in the specific thread she had linked. Two posted in the same channel on a different thread. Two posted in a different channel entirely, apparently triggered by re-opening Slack and finding something else active. Four of the 7 who re-engaged were still posting 30 days after the re-routing DM.

The win-back format had produced 12% re-engagement. The re-routing format produced 58%. The difference was the frame: win-back said “you’ve been absent,” which put the member in a defensive posture about their own behaviour. Re-routing said “here is something specific that is relevant to you right now,” which gave the member a concrete next action that required no acknowledgment of a gap.

The quarterly re-engagement routine

She systematised the per-member engagement scoring into a quarterly review that ran alongside her monthly activation work.

The routine, run every three months:

Step 1: Per-member engagement score calculation (30 minutes). Download the Slack member activity CSV export. For each member, record their 30-day message count and distinct channels active. Pull the workspace export JSON and check peer reply counts for each member’s messages in the 30-day window. Apply the scoring formula (messages + 2×peer-replied posts + 3×channels, capped). Rank all members from highest to lowest. Divide into quartiles. Record the score in her spreadsheet alongside tenure.

Step 2: Activation shadow identification (15 minutes). Filter the bottom two quartiles (members with low engagement scores) by tenure: remove anyone with less than 60 days tenure (they are in or near the activation window and handled by the Day-0/3/7 sequence). Of the remaining low-engagement members with 60+ days tenure, cross-reference against activation records to identify those who activated in week one. This is the activation-shadow list for the quarter.

Step 3: Re-routing DM queue (45 minutes to 1 hour). For each activation-shadow member, open their original intake form response. Pull up the current most-active thread in their matched channel (she had five channels: #proposal-frameworks, #client-acquisition, #pricing-and-scope, #operations-and-tools, #career-and-transitions). Write a personalised re-routing DM using the structure above. She batched these in groups of ten, writing ten DMs per session, aiming to send all activation-shadow DMs within one week of the quarterly review.

Step 4: Follow-up tracking (10 minutes, 30 days later). Note which activation-shadow members re-engaged (posted within 30 days of the DM). Update their tier in the spreadsheet. Members who did not re-engage after a re-routing DM were flagged for the Churned-tier personal DM from the member health score framework — a different message acknowledging the gap and explicitly surfacing whether something had changed since they joined.

The full quarterly routine took 90 minutes to two hours, plus the DM-writing time. For her 68-member activation-shadow pool at the start of the system, the DM-writing time was 3 to 4 hours spread over a week. For ongoing quarters, with 10 to 15 new activation-shadow members per quarter (the ones who activated in the prior quarter and faded over the next 90 days), the DM-writing time was 45 to 60 minutes per quarter.

Three-month results

She ran the full re-routing system for the November 2025 through January 2026 activation-shadow cohort (the 68 original members plus new quarterly additions). Results as of February 2026:

What the engagement rate aggregate had been hiding

Looking back at the eight months before she built the per-member ranking, Yuki could see that her community-level engagement rate had been reporting a correct but misleading number. 36% was technically accurate: of 200 members, 72 were active. But the 36% number suggested a problem that was uniformly distributed across the membership — a community where everyone was somewhat under-engaged. The per-member ranking showed a different reality: a community with a small, highly active core and a large passive periphery, within which a specific identifiable segment (the activation shadow) was both the highest-ROI intervention target and the hardest to see with community-level metrics.

The three-stage diagnostic she had run before building the per-member ranking — check activation, check content cadence, check channel structure — had correctly cleared all three stages. But the diagnostic assumes that below-benchmark engagement is explained by one of those three root causes. The activation shadow is a fourth root cause the diagnostic doesn’t cover: members whose activation succeeded but whose ongoing engagement was never maintained because the community’s programming didn’t specifically surface new entry points as their professional situations evolved past the goals they had at the time they joined.

The activation sequence she had built was excellent at getting members from zero to one post in week one. It was not designed for members who had been in the community for five months and needed to get from zero posts this month back to one post this week. Those members needed something different: a specific, current, personally relevant reason to open Slack that was rooted in their present situation, not their original onboarding frame.

Re-routing was that something different. The quarterly review and per-member ranking was the system that identified which members needed it and when.

Frequently asked questions

What is the activation shadow in a paid Slack community?

The activation shadow is the segment of paid community members who completed week-one activation — posting at least once in their first seven days — but whose engagement faded between months 2 and 5, leaving them as passive-paying subscribers by month 3 or 4. These members are invisible to week-one activation metrics (which only measure first-week behaviour) and to monthly new-member cohort analyses. They appear in the community-level engagement rate as part of the non-posting majority, but the community-level rate does not distinguish between members who never activated (and need onboarding interventions) and members who activated and faded (who need re-routing to current relevant content). Per-member engagement scoring, applied quarterly to the full membership, is the only method that reliably surfaces the activation shadow as a distinct, identifiable segment with its own renewal profile and its own intervention path.

How is per-member engagement rate different from community-level engagement rate?

Community-level engagement rate is a single aggregate number: the percentage of all paying members who posted or reacted at least once in the last 30 days. It answers whether the community overall is in a healthy range — 45 to 55% for knowledge and peer-learning communities, 35 to 45% for professional networking communities — but it hides the distribution of engagement across individual members. A 40% community-level engagement rate looks the same whether it comes from 200 members all contributing at a moderate level or from 80 highly active members and 120 completely passive members. Per-member engagement scoring assigns each member an individual score based on their recent activity: messages sent, peer replies received, and distinct channels active in a rolling 30-day window. Ranking all members by this score reveals the distribution — specifically, whether the aggregate rate is being held up by a small core while a large passive group approaches their renewal decision. See the Slack community member engagement rate guide for the formula and benchmark ranges by community type.

How do you build a per-member engagement score for a paid Slack community?

Download the Slack member activity CSV from Workspace Admin → Analytics → Members. For each member, record messages sent in the last 30 days and distinct non-introduction channels active. For peer replies, use the workspace export JSON (available from Workspace Settings → Import/Export Data on paid Slack plans) and check the reply_count field on each message per member. A simple scoring formula: 1 point per message sent, 2 additional points for each message that received at least one peer reply, 3 points for each distinct non-introduction channel active (cap at 3 channels, 9 points maximum). Rank all members from highest to lowest raw score. Divide into quartiles. For the full five-table reference — including signal weights, tier thresholds, per-tier interventions, calculation methods, and renewal benchmarks by health score tier and community price tier — see the paid community member health score reference card.

What renewal rate should I expect for bottom-quartile engagement members?

For paid Slack communities in the $100 to $199 per month range, members in the bottom quartile of per-member engagement scoring — those with zero posts in the last 30 days and minimal peer interaction history — renew at month 3 at 22 to 35%, regardless of their week-one activation status. A member who posted an intro in week one but has been silent for 60 days renews at approximately the same rate as a member who never activated: the week-one activation signal weakens as a predictor as the length of the silence increases. By month 3 with no posts in the last 45 days, the current per-member engagement score is the relevant predictor, not the historical activation event. The re-routing intervention for activation-shadow members — a goal-relevant DM connecting them to a current active thread, without acknowledging the silence — converts 40 to 50% to posting within 72 hours and improves month-3 renewal from around 28% to 50 to 55% for re-engaged members.

How is re-routing different from a win-back message for inactive community members?

A win-back message acknowledges that a member has been absent and invites them to return. Its implicit frame is backward-looking: “I noticed you haven’t been active.” Re-routing treats the member as currently engaged and points them forward to a specific current conversation: “there’s a thread happening right now that’s directly relevant to what you said you were working on when you joined.” Win-back messages for activation-shadow members produce 8 to 15% re-engagement. Re-routing messages that name a specific current thread in the member’s matched channel and connect it to their original stated goal produce 40 to 50% posting within 72 hours. The difference is not tone — it is what the message implies about the reason for the silence. Win-back implies disengagement; re-routing implies the member is still relevant and the community has content for them right now. Activation-shadow members typically faded because they ran out of entry points that felt applicable to their current situation, not because they lost interest. Re-routing solves the entry-point problem directly. Win-back does not.

Related: Slack community member engagement rate — formula, benchmarks by community type, and what to do when it’s low · Slack community engagement rate — the denominator trap case study (Priya Sharma, The Developer Relations Collective) · Paid community member health score — why login frequency predicts churn wrong and the three behavioral signals that actually predict renewal · Slack community health metrics — six numbers and what to do when each one drops.