Retention & Health Metrics
Paid community member engagement rate: why the ghost member problem grows invisibly until month 6, and how monthly tracking prevents the Dormant spiral
The community launched well. Week-one activation was strong — 60% of new members posted an intro, 45% formed at least one peer connection by day 14, and the operator had a clear sense that the onboarding sequence was working. Nine months later, the same community sits at 17% monthly engagement rate. More than four in five paying members have not posted anything in the past 30 days. The operator is not sure when it happened. The WAU dashboard showed a gradual decline, but no single week looked alarming enough to trigger intervention. By the time the engagement rate fell below 20% and crossed into the Dormant tier, roughly 40–50% of the non-engaging members were already in the pre-cancellation evaluation window — making cost-versus-value comparisons against what the subscription now meant to their daily professional life. This pattern — strong launch, invisible decline, Dormant-tier arrival — is not a failure of community quality. It is a structural blind spot in how paid community operators measure ongoing health: they track new-member activation, but they do not systematically track the disengagement of members who activated in months one through four. Those members — long-tenured, once-activated, now silently absent — are the ghost member problem. The monthly engagement rate is the instrument that makes ghost members visible. But the engagement rate only becomes a retention tool when you understand the mechanism behind how ghost members accumulate, how to detect them before the rate falls to Dormant, and what the structural audit looks like for a community that has already crossed the threshold.
What the engagement rate measures and the ghost member blind spot inside the formula
The monthly engagement rate formula is simple: divide the number of paying members who posted at least one message in the past 30 days by the total number of paying members, and multiply by 100. A community with 300 paying members where 120 posted something last month has a 40% engagement rate, which places it solidly in the Healthy tier. A community where 54 of 300 members posted has an 18% rate, which is Dormant. For the exact formula parameters — what counts as “posted,” which members to include in the denominator, which channels to count for numerator eligibility, and how to handle trial members and grace-period accounts — the paid community member engagement rate reference card covers the decision tables in full. The formula itself is not what creates the blind spot. The blind spot lives in how the denominator is constructed.
When you calculate monthly engagement rate, every paying member — regardless of whether they joined six days ago or eighteen months ago, regardless of whether they posted every week for the first six months or never posted at all, regardless of whether they are actively deciding to cancel right now or deeply embedded in three ongoing threads — appears once in your denominator. The formula treats all non-posters as equivalent absences. A member who joined three weeks ago and has not yet posted their intro is identical in the formula to a member who joined fourteen months ago, posted actively for eight months, formed twenty peer connections, ran a thread that became the community’s most-linked resource, and has not logged in for the past sixty days. Both are in the denominator. Neither is in the numerator. Both reduce the engagement rate by the same amount per member slot.
This equivalence is the ghost member blind spot. The long-tenured formerly-active member is a ghost member: a paying subscriber whose behavioral relationship with the community ended months ago, whose investment in the subscription is currently zero, and who is in the pre-cancellation window without any external signal that would surface them to the operator before they cancel. The new member who has not yet posted is an onboarding case — a member whose relationship with the community is forming, not ending. These two situations require completely different operator responses. The engagement rate formula cannot distinguish them because the formula does not track history; it measures a 30-day window. Operators who rely solely on the engagement rate number see their metric declining and correctly diagnose that there are too many non-posting members, but they cannot determine whether the problem is in onboarding failure, ghost member accumulation, or some combination of both — and each problem has a different intervention.
Understanding the ghost member mechanism is what transforms the engagement rate from a diagnostic number into a retention management tool.
The ghost member pattern: how long-tenured members disengage silently
Ghost members did not start as ghost members. Almost universally, the members who become ghosts in months four through eight had strong behavioral signals in months one and two. They posted in #intros. They replied to threads. They formed peer connections. In the language of the individual member health score, they were Activated or Healthy in their first sixty days. The ghost member is not a member who joined reluctantly and never committed — they are a former participant who drifted away from a community they initially valued.
The disengagement mechanism follows a consistent pattern, and it is slow enough at each step to feel invisible. In month two or three, the member’s posting frequency begins to decrease. This is partly natural — the novelty of a new subscription drives higher engagement in months one and two than any subsequent month — and partly a signal of the first structural problem: the member is running out of threads that feel worth their specific contribution. They are no longer in the phase where everything in the community is new and potentially interesting. They are now looking for the specific conversations that match where they currently are in their work, and if those conversations are not reliably present in the channels they have bookmarked, they begin to substitute other sources: newsletters, podcasts, direct peer conversations outside the community. The community becomes one of several professional inputs rather than the primary one, and their posting frequency reflects that shift.
In month three or four, the member’s peer connections inside the community begin to atrophy. This is the most important structural step in the ghost member mechanism. The peer connections formed in months one and two were anchored in the threads that were active during that period. When those threads stop being active — when the peers the member connected with shift to different topic areas, or when the member’s own professional context changes enough that their prior peer connections are less relevant — the anchor points that made the community feel indispensable begin to loosen. A member who once had six reliable peers whose posts they looked for specifically now has three — and then two — as the individual-level peer interaction count falls. Each step in this decline corresponds to one fewer specific reason to open the workspace. The member does not decide to disengage. They simply log in less often, reply less often, and eventually find that the workspace has become a place they think about opening rather than a place they open habitually.
By month five or six, the member has crossed from infrequent engagement into the functional equivalent of ghost-member status: they are paying, they have not cancelled, and they have not posted in 45 or more days. At this point, the pre-cancellation psychology is well underway. The member is aware that they are not using the subscription. They are aware of the monthly charge on their card. They are constructing a narrative about the community that justifies either returning (which requires finding a compelling reason) or cancelling (which requires convincing themselves that the gap cannot be filled). The operator has a closing window to intervene — roughly 30 to 60 days after the 45-day silence threshold — before the cancellation narrative solidifies. After that window, the re-engagement conversion rate drops from 18–28% to 8–12%, and the win-back economics become significantly worse.
What makes this pattern structurally dangerous is that ghost members continue to pay during the disengagement period. A member who began disengaging at month three and cancelled at month eight paid five months of subscription fees without receiving meaningful value. This is not just a retention statistic — it is a compounding problem. Ghost members who eventually cancel represent the worst possible outcome: a member who experienced value erosion, formed a negative mental model of the community during the disengagement period, and left without any win-back intervention that might have rescued the relationship. The monthly engagement rate surfaces this problem at the community level. The individual member health score — specifically the peer interaction count and active channel engagement depth signals — surfaces it at the member level, typically 45 to 90 days before the ghost member would show up in the engagement rate trend. Using both metrics in parallel is how operators close the detection gap.
Why communities that launched well are particularly vulnerable to ghost member accumulation
The communities that suffer most severely from ghost member accumulation are frequently the ones that had the strongest launches. This appears paradoxical until you understand the mechanism: a strong launch produces a high first-month activation cohort, which creates a large population of Activated and Healthy members at the six-to-twelve-month mark. The stronger the launch, the more members enter the community with high behavioral engagement in months one and two — and the more members are available to transition into ghost member status in months four through seven if the community’s programming does not evolve to meet their needs at that stage of membership tenure.
Communities with weak launches have low first-month activation rates. Many new members never post an intro, never form peer connections, and churn early — before the three-month mark. This is clearly bad, and it produces high early-stage churn. But it also means the community never accumulates a large ghost member population, because the members who would have become ghosts churned before they could disengage. The weak-launch community has a different problem profile: high early churn, low absolute membership count, but a relatively high ongoing engagement rate because the members who stayed past month three are self-selected for engagement.
The strong-launch community retains more members through month three because the activation experience was better. A higher proportion of new members posted intros, formed peer connections, and found channels worth contributing to. Those members paid month two and month three and month four. By month five, the community has a large base of long-tenured members — and a significant number of them are in the early stages of ghost-member disengagement. By month eight, the operator who is tracking the monthly engagement rate sees the rate declining from a 47% Healthy baseline toward 35%, then 28%, then 22%. Each individual step is small enough to look like natural variance. The cumulative trajectory is a Dormant-tier arrival.
The week-one activation work — the three-touch onboarding sequence that produces intro post completion, the Day 3 peer-routing intervention, the Day 7 health check for operators — is documented in the paid community welcome sequence guide and is the foundation for avoiding early churn. But that work produces Activated members, not permanently retained members. Activation is the beginning of the retention challenge, not the end of it. The community that does activation well and then stops measuring behavioral engagement at the community level has invested in the inputs and stopped tracking the outputs.
The four engagement tiers and what is actually driving each one
The four engagement tiers — Thriving (≥55%), Healthy (35–54%), Declining (20–34%), and Dormant (<20%) — describe the same community at different stages of a ghost member accumulation cycle. Understanding what is driving the rate at each tier is essential for choosing the right intervention.
Thriving (≥55%): the ghost member population is small and manageable
A Thriving-tier community has a six-month renewal benchmark of 82–94%, the highest of any tier. The underlying reason is not simply that more than half of members are posting — it is that the ratio of active members to ghost members is still low enough that active members do most of the community’s social work by default. New members join and encounter active threads almost immediately. Existing members who start to drift toward disengagement encounter reply activity that pulls them back. The community’s social density is self-reinforcing.
Operators in the Thriving tier tend to underestimate the value of the current state and under-invest in the proactive interventions that would maintain it. The most common Thriving-tier error is treating high engagement rate as a signal that no intervention is needed, when the actual requirement is content programming to sustain activity density and proactive peer deepening among top contributors to prevent burnout. A Thriving community is not a set-and-forget state — it is a state that requires active maintenance to avoid the slow drift toward Healthy that precedes the drift toward Declining.
Healthy (35–54%): the ghost population is growing but not yet dominant
A Healthy-tier community has a six-month renewal benchmark of 61–76%. The gap between the Thriving and Healthy tiers in renewal rate — roughly 15–20 percentage points — is explained almost entirely by the difference in ghost member population size. A community at 45% engagement has a ghost member population of roughly 55% of paying members. Most of those 55% are distributed across three categories: never-activated members who never posted (early churners who are still paying), recently disengaged members (months 3–4 since last post, in the early ghost phase), and confirmed ghost members (5+ months of silence while still paying).
The Healthy tier is where the engagement rate trend direction matters more than the absolute rate. A community at 48% that was at 55% three months ago is trending toward Declining and should be treated with the urgency appropriate to its trajectory, not its current position. A community at 36% that was at 28% three months ago is recovering from a prior intervention and should be evaluated against that baseline. Operators who read the engagement rate as a static number rather than a directional signal consistently misdiagnose their position on the accumulation cycle.
For the individual-member counterpart to the community-level engagement rate — the three behavioral signals (intro post completion, peer interaction count, active channel engagement depth) that predict individual renewal at 73% accuracy — the paid community member health score reference card covers the per-member scoring and tier assignment. The health score is the instrument for identifying which specific members in the Healthy-tier ghost population are earliest in the disengagement cycle and therefore the highest-priority for intervention.
Declining (20–34%): the ghost population is now large enough to affect the new-member experience
A Declining-tier community has a six-month renewal benchmark of 38–54%. At this rate, fewer than one in three paying members is contributing to community threads in any given month. The ghost member population is now large enough to create a structural experience problem for new members: when a new member joins a Declining-tier community and browses the channels they bookmarked from the onboarding sequence, they encounter threads where replies took days rather than hours, where the last post in some channels was two or three weeks ago, and where the sense of a live, responsive community that they purchased is not what they find in the workspace. The Declining community’s new-member experience problem feeds directly back into its ghost member problem: new members who activate poorly because the community feels quiet begin the ghost-member disengagement pattern earlier than they would in a Thriving or Healthy community, which accelerates the rate at which the total ghost population grows.
Interventions in the Declining tier are still effective but must address both the existing ghost population and the new-member activation pipeline simultaneously. Programming interventions that create fresh activity — expert Q&A sessions, operator-seeded discussion threads, structured peer matching events — can temporarily increase the numerator of the engagement rate formula, which improves the new-member experience and slows the ghost-member accumulation rate for recent members. But they do not address the confirmed ghost members who have been silent for five or more months and whose pre-cancellation evaluation is already in progress. For those members, the direct ghost member re-engagement intervention described below is required in parallel.
Dormant (<20%): the ghost population is dominant and the community is in a negative engagement loop
A Dormant-tier community has a six-month renewal benchmark of 19–35%. Fewer than one in five paying members contributed anything in the past 30 days. At this rate, 40–60% of the total paying member roster is typically in confirmed ghost-member status (90+ days of silence while still paying), and an additional 15–25% is in the early ghost phase (30–89 days of silence). The remaining 15–20% of active members are carrying the entire content weight of the community — a situation that consistently produces active-member burnout, which further reduces the numerator, which accelerates the Dormant spiral.
The Dormant tier is characterized by a negative engagement loop: low activity makes the community feel less valuable to existing members, which reduces their posting frequency, which makes the community feel less valuable to new members, which reduces new-member activation rates, which means fewer new voices enter the active-member pool, which further reduces activity. Operators who try to break this loop with content programming alone — more expert sessions, more structured discussions, more curated threads — frequently find that attendance at these events is low because the members who would attend are in ghost-member status and not monitoring the workspace closely enough to see the announcement. The Dormant tier requires a structural intervention that addresses the ghost member population directly, not just the content programming.
The tenure distribution of a Dormant-tier community’s ghost members reveals something operators find counterintuitive: a disproportionate share of Dormant-tier ghost members are long-tenured members who were once Activated. The strong-launch communities described earlier are disproportionately represented in the Dormant tier specifically because of this: they launched well, retained a large early cohort, and then watched that cohort disengage gradually across months three through eight without a systematic intervention to catch them at the disengagement inflection point. The member churn by tenure reference card documents the three tenure-specific cancellation patterns (the month-3 evaluation cliff, the six-month value-stagnation window, and the twelve-month “what have I actually done here” review) that explain when in the ghost-member disengagement cycle cancellations are most likely to cluster.
How to run a ghost member audit before the engagement rate falls to Dormant
The ghost member audit is a member-by-member tenure analysis that identifies which members are in the early ghost phase (30–89 days since last post) before they become confirmed ghosts (90+ days of silence), and which confirmed ghosts have a re-engagement window that has not yet closed. The audit should run monthly, immediately after calculating the engagement rate for the month, using the same Slack member activity export data. The timing matters: running the audit monthly alongside the engagement rate calculation ensures that early ghost members are identified with 30–60 days of re-engagement runway remaining, rather than being identified after the window has closed.
Step 1 — Build the ghost member matrix. Open the engagement rate spreadsheet you built for this month’s calculation (or create one from the Slack member activity CSV export). For every member who did not appear in the numerator — every member with zero messages in the past 30 days — add two columns: their join date, and their last-message date. The join date comes from your billing system export. The last-message date is the most recent date in the Slack activity export where their message count was greater than zero.
Step 2 — Segment the non-posting members into five categories. Category 1 — New members in onboarding window: joined within the past 14 days. Exclude from the ghost member analysis; they are not yet eligible for the engagement rate denominator. Category 2 — Never-activated members: joined more than 14 days ago, last-message date is null (never posted anything in any public channel). These are members who failed the week-one activation barrier. They need the welcome sequence audit, not the ghost member re-engagement intervention. Category 3 — Early ghost members: joined more than 90 days ago, last-message date is 30–89 days ago. These members were active but are now entering the disengagement pattern. The re-engagement window is open; intervention converts at 24–34% within 30 days. Category 4 — Confirmed ghost members: joined more than 90 days ago, last-message date is 90–180 days ago. These members are in the pre-cancellation window. The re-engagement window is closing; intervention converts at 18–28%. Category 5 — Deep ghost members: joined more than 90 days ago, last-message date is more than 180 days ago or null for members with prior activity before a cutoff. These members have very likely already made a mental decision to cancel and are in a passive subscription state. Win-back converts at 8–12%.
Step 3 — Build the intervention queue for Categories 3 and 4. For each Early ghost member (Category 3), record three things: their last active thread or channel, their stated goal from onboarding (if you collected this), and one Activated or Healthy member in the community with adjacent context who could be peer-routed to their last active thread. For each Confirmed ghost member (Category 4), record the same information plus a note on whether they had any direct interaction with the operator or community team at any point in their tenure. Members with prior direct interaction have higher re-engagement conversion rates for personal DMs than members who were always self-directed. The intervention queue separates members by category so the operator can prioritize Category 3 contacts first (higher conversion, wider window), then Category 4, then Category 5 if time allows. Deep ghost members (Category 5) are candidates for a win-back email at the next billing date rather than a personal DM.
Step 4 — Record the ghost member ratio for trend tracking. Divide the total count of Category 3 + Category 4 members by the total paying member count to get the ghost member ratio. This is a separate metric from the engagement rate: a community can have a 35% engagement rate (Healthy tier) with a 28% ghost member ratio, which signals a growing problem that will push the rate into Declining within two to three months if not addressed. Track the ghost member ratio monthly alongside the engagement rate. If the ghost member ratio is increasing faster than the engagement rate is decreasing — which happens when long-tenured members are accumulating in ghost status faster than they are cancelling — the operator is building up a cancellation wave that will arrive in a concentrated window when the six-month pre-cancellation evaluation period ends for a cohort of ghost members simultaneously.
The three-part structural audit for Dormant-tier communities
A community that has already crossed into the Dormant tier (<20% engagement rate) requires a structural intervention, not an incremental content push. The three parts of the Dormant-tier audit are sequenced intentionally: Part 1 is the ghost member triage (understand the population before acting), Part 2 is the direct ghost member re-engagement sequence (address the confirmed ghost population with personal contact), and Part 3 is the channel architecture audit (address the structural reason new members are not activating and why re-engaged ghost members may disengage again without a channel environment worth returning to).
Part 1 — Ghost member triage: understand the population before acting
The first step in the Dormant-tier audit is to run the full ghost member audit described above and build the five-category breakdown of the non-posting member population. This breakdown tells the operator what the Dormant rate is actually composed of, which determines which interventions are worth deploying and in what order. A Dormant-tier community where 60% of ghost members are Category 2 Never-activated members has a fundamentally different problem from one where 60% are Category 3 and 4 early and confirmed ghost members. The first community has an onboarding failure as the primary driver of low engagement; the second has a mid-tenure disengagement failure. The interventions for these problems are different, and deploying the wrong one wastes the operator’s limited time and the member’s attention budget on an outreach that does not address the root cause of their absence.
A practical benchmark for Dormant-tier communities: the ghost member population (Categories 3 and 4 combined) typically represents 40–60% of total paying members. Of that ghost population, roughly 55–65% are Confirmed ghost members (Category 4, 90–180 days of silence) and 20–30% are Early ghost members (Category 3, 30–89 days of silence). The remaining 10–20% are Deep ghost members (Category 5, 180+ days of silence). The operator should expect to spend the most effort on Category 3 and 4 re-engagement, and should not invest significant effort in Category 5 win-back during the initial Dormant-tier recovery — the conversion economics do not justify the time relative to the higher-ROI intervention available in Category 3.
Part 2 — Direct ghost member re-engagement: personal, specific, and within 14 days
Ghost member re-engagement that converts follows the same principles as the At-Risk individual member intervention described in the member health score post — personal, specific, and tied to the member’s stated goal — but applied at the community level across a larger population and with adapted timing. The personal DM for a Confirmed ghost member (Category 4) differs from the At-Risk DM in one key structural element: it must acknowledge the gap rather than pretend the absence did not happen. A Category 4 member who has been silent for 90–180 days is aware of the gap. An outreach that references a “great thread this week” as if they have been following the workspace all along signals to the member that the operator either does not know how long they have been absent or is pretending not to know. Neither interpretation is conducive to re-engagement.
The DM structure that produces the highest re-engagement conversion for Category 4 Confirmed ghost members is: (1) acknowledge the gap without apology or accusation (“I know the past few months have probably been busy”); (2) name something specific that has happened in the community since their last activity that is directly relevant to the goal they stated when they joined; (3) offer a single concrete low-barrier re-entry point that does not require them to have been following recent activity (“There’s a thread [peer name] started in [specific channel] about [topic adjacent to member’s goal] that I thought was exactly the kind of thing you were working on when you first joined — worth dropping a reply if you’re back and have a few minutes”); (4) optionally, offer a direct re-orientation call for members with whom the operator has had prior direct interaction (“If it would be useful to catch up on what the community has been working on since you were last here, I’m happy to do a quick call to point you to the right corners”). This structure converts Confirmed ghost members at 18–28% back to active participation within 30 days. Generic check-ins convert at 4–8%.
For Early ghost members (Category 3), the DM is nearly identical in structure but references a thread they can still participate in because it is recent: (“I noticed the discussion in [specific channel] this week about [specific topic] — given what you were working on in [their last active thread], I thought it would be worth a reply”). The early ghost re-engagement converts at 24–34% because the gap is short enough that the member is not deeply embedded in the absence narrative and because the specific thread gives them an immediately actionable re-entry point without requiring them to browse a workspace that felt overwhelming or quiet at the time of their last visit.
In parallel with the personal DM, peer-routing increases the re-engagement conversion rate by creating an independent pull factor. Identify one Activated or Healthy member for each ghost member outreach who has adjacent context — working on a related problem, active in the channels the ghost member was previously engaged with — and route that peer to the ghost member’s last active thread or their original intro post. The goal is not to manufacture a false interaction; it is to ensure that when the ghost member receives the operator DM and returns to the workspace to look at the thread they were pointed to, they see a response that demonstrates the community is still paying attention to the problem space they originally joined for. Operator DM alone produces 18–28% conversion for Category 4. Operator DM combined with peer-routing produces 22–35%. For the full retention management framework that connects the individual health score, the community engagement rate, and the per-tier intervention rhythms into a weekly operator ritual, the paid community member retention reference card covers the complete system.
Part 3 — Channel architecture audit: address the structural cause of re-disengagement
The third part of the Dormant-tier audit is the one most frequently omitted, and it is the reason many Dormant-tier communities that successfully run ghost member re-engagement sequences find themselves back at Dormant rates within four months. The ghost member re-engagement sequence can return formerly-active members to the workspace. It cannot make the workspace worth staying in if the channel architecture does not give re-engaged members an ongoing reason to post.
Dormant-tier communities almost universally have a channel architecture problem that predates and partly drives the ghost member accumulation. The pattern is consistent: when the community launched, the operator created channels around anticipated discussion topics. Those channels accumulated early-stage posts from the initial cohort. As the community grew and membership diversified, new channels were created to address new interests. At month twelve, the workspace has 30 to 50 public channels, of which typically 60–70% have had zero messages in the past 30 days. A member returning to the workspace after two months of absence opens the sidebar and sees a list of channels where the most recent posts are from weeks or months ago. The silence is visual and immediate. It reinforces the pre-cancellation narrative rather than contradicting it.
The channel architecture audit for a Dormant-tier community has three steps. First, export the channel activity data from the Slack admin panel and sort by last-message date. Any channel with zero messages in the past 90 days should be archived. Archive, not deleted — the channel history is preserved for members who want to reference it, but the channel disappears from the sidebar, which reduces the visual footprint of inactivity. For a 30-channel workspace, this typically results in archiving 15 to 25 channels and leaving 5 to 15. Second, review the remaining active channels and evaluate whether each one is organized around a member job-to-be-done or around a topic category. Channels organized by job-to-be-done (“scaling past $500k revenue,” “building my content marketing function”) give members with that specific goal an obvious recurring destination. Channels organized by topic (“#marketing,” “#operations”) are too broad to feel like a personal space and too narrow in another dimension — any member can claim relevance but few feel it is specifically theirs. The channels that consistently generate the highest engagement depth and the lowest ghost-member rates are the ones that members can name their job in. Third, for each retained active channel, identify whether the operator is seeding at least one discussion thread per week or whether activity is entirely member-generated. In Dormant-tier communities with only 15–20% active membership, entirely member-generated activity is insufficient to keep channels feeling live. The operator must seed threads in the highest-priority channels weekly for the duration of the re-engagement period — typically the 60–90 days following the ghost member re-engagement sequence — until the active member population grows large enough to sustain organic thread density.
When the channel architecture audit is completed alongside the ghost member re-engagement sequence, the combined effect on engagement rate recovery is measurable within 60 to 90 days: communities that run both interventions together see engagement rate recovery of 8–15 percentage points in the two months following the audit, compared to 3–6 points for communities that run only the re-engagement sequence without the channel consolidation. The arithmetic is straightforward: ghost member re-engagement increases the numerator by bringing formerly-active members back to posting. Channel consolidation increases the probability that any given re-engaged member posts again the following month, which sustains the numerator increase rather than producing a one-time bump followed by re-attrition.
The month-6 cancellation cliff and what the engagement rate shows in advance
The month-6 cancellation cliff is the tenure point at which cancellations are most concentrated in Declining and Dormant-tier communities. The mechanism is predictable. A member who became a ghost in month 3 or 4 has been in the pre-cancellation evaluation period for two to three months by the time month 6 arrives. The month-6 billing renewal is the decision point: the member sees the charge, explicitly evaluates whether they got value from the subscription in the past 30 days (they did not), and cancels at a rate of 45–65% in Declining and Dormant-tier communities, compared to 8–16% in Healthy and Thriving-tier communities for the same tenure cohort.
What makes this cliff predictable in the engagement rate data is the following pattern: communities that will hit a concentrated month-6 cancellation wave typically show a consistent downward engagement rate trend that began three to four months before the cliff. A community at 42% in month 1, 36% in month 2, 29% in month 3, and 23% in month 4 is four months into a consistent 6–8 point monthly decline. The members who will cancel at month 6 are largely identifiable at month 3 as Category 3 early ghost members: they were active in months 1–2, fell silent in month 3, and will cross the 90-day silence threshold by month 4. The window to intervene with the early ghost re-engagement sequence closes between months 3 and 4. Operators who track the engagement rate monthly and run the ghost member audit alongside it have a roughly 30–45 day window to act on the early ghost population before that window closes and the Confirmed ghost DM becomes necessary. Operators who check the engagement rate quarterly and notice the Dormant-tier number in month 7 have already lost that window for most of the month-6 cancellation cohort.
The engagement rate is not a lagging indicator in the sense that it tells you what happened. It is a near-real-time indicator of the current ratio of active to non-active paying members. What makes it feel like a lagging indicator is the absence of the ghost member audit alongside it: without knowing which non-posting members are in which phase of the disengagement cycle, the operator cannot determine how much of the rate decline is preventable and how much has already converted to pre-cancellation narrative. The combination of monthly engagement rate tracking and the five-category ghost member audit converts the engagement rate from a health report into a retention operating system. The 2-minute Onboarding Health Check can identify whether your community’s primary engagement problem is in the activation phase (week-one onboarding) or the retention phase (mid-tenure ghost member accumulation) — the intervention path is different, and diagnosing the primary driver correctly determines whether your next session is spent on the welcome sequence or the ghost member audit.
Frequently asked questions
What is paid community member engagement rate?
Paid community member engagement rate is the percentage of paying members who posted at least one message in the past 30 days: (members who posted ≥1 message in the past 30 days) ÷ (total paying members at the start of the window) × 100. The formula requires three parameter decisions: what counts as “posted” (public channel messages and replies; not emoji reactions or DMs), which members to include in the denominator (paying members active for more than 14 days, excluding free-tier and trial accounts), and which channels count for numerator eligibility (all public channels, excluding bot-only and announcement-only channels). The four tiers are Thriving (≥55%), Healthy (35–54%), Declining (20–34%), and Dormant (<20%), with six-month renewal benchmarks of 82–94%, 61–76%, 38–54%, and 19–35% respectively. For the full decision tables, calculation methods by community scale, and per-tier interventions with expected response rates, see the paid community member engagement rate reference card.
What is the ghost member problem in paid communities?
The ghost member problem is the accumulation of long-tenured paying members who activated in week one — posted an intro, formed peer connections, contributed to channels early in their membership — and then silently disengaged in months 3 through 6 without cancelling. Ghost members continue to pay throughout the disengagement period, appear in the denominator of the monthly engagement rate formula, and drag the community-level rate toward the Dormant tier with no external signal to the operator until the rate decline becomes visible. The mechanism follows a consistent pattern: novelty-driven engagement drops in month two or three, peer connection anchors begin to atrophy as prior threads become inactive in month four, and by month five or six the member is in the pre-cancellation evaluation window. Ghost members are detectable 45–90 days before they show up in the engagement rate trend through individual-level peer interaction count and active channel engagement depth signals — the signals tracked by the individual member health score — and through the five-category ghost member audit described above.
What are the four paid community engagement tiers?
Thriving (≥55%): more than half of paying members posted in the past 30 days; six-month renewal 82–94%. Ghost member population is small; social density is self-reinforcing for new members. Operator focus is maintenance and preventing active-member burnout. Healthy (35–54%): roughly one in three to one in two paying members posted; six-month renewal 61–76%. Ghost member population is growing; trend direction matters more than absolute rate. Declining (20–34%): fewer than one in three paying members posted; six-month renewal 38–54%. Ghost member population large enough to affect new-member experience; negative engagement loop beginning. Interventions must address both existing ghosts and new-member activation pipeline simultaneously. Dormant (<20%): fewer than one in five paying members posted; six-month renewal 19–35%. Ghost member population is 40–60% of total paying members; active members are carrying disproportionate content weight and at burnout risk. Requires three-part structural audit: ghost member triage, direct re-engagement sequence, and channel architecture consolidation.
How do you calculate paid community member engagement rate manually?
The manual calculation requires two data exports and approximately 30–45 minutes per monthly run. Step 1: export paying members from your billing system; exclude members who joined in the past 14 days (onboarding window exclusion). Step 2: export the Slack member activity CSV from workspace admin (Admin → Analytics → Members, 30-day window). Count the paying members from Step 1 who appear in the Slack export with at least one message in the 30-day window, excluding bot-only and announcement channels. Step 3: divide the Step 2 count by the Step 1 total and multiply by 100. Step 4: record the rate alongside the prior two months; the trend direction matters more than any single month’s absolute rate. Step 5 (ghost member audit extension): for every non-posting member in the paying roster, record their join date and last-message date from the activity export; segment into the five categories (new in onboarding window, never-activated, early ghost, confirmed ghost, deep ghost) to build the intervention queue for the following week.
What should a paid community operator do when engagement rate drops below 20% (Dormant tier)?
The Dormant tier requires a three-part structural intervention, not an incremental content push. Part 1 — Ghost member triage: run the full five-category ghost member audit to understand whether the low rate is driven by never-activated members (onboarding problem) or confirmed ghost members (mid-tenure disengagement problem); the interventions are different. Part 2 — Direct re-engagement sequence: deploy personal operator DMs to Category 3 early ghost members (30–89 days silent) and Category 4 confirmed ghost members (90–180 days silent), each referencing a specific thread and a named peer; personal and specific DMs convert at 18–28% for Category 4 and 24–34% for Category 3; generic check-ins convert at 4–8%. Run peer-routing in parallel to create an independent pull factor; operator DM combined with peer-routing produces 22–35% re-engagement for Category 4. Part 3 — Channel architecture audit: export channel activity data; archive all channels with zero messages in 90 days; consolidate the remaining active channels around member job-to-be-done rather than topic categories; seed at least one operator-initiated thread per high-priority channel per week for 60–90 days until organic thread density is restored. Communities that run all three parts together see 8–15 point engagement rate recovery within 60–90 days, compared to 3–6 points for re-engagement sequence alone.