Reference card — paid community retention
Paid community member engagement rate
Decision tables for paid community operators tracking monthly member engagement rate: the exact formula and what counts as “engaged”; four engagement health tiers (Thriving / Healthy / Declining / Dormant) with 6-month renewal benchmarks per tier; per-tier operator interventions with timing and expected response rates; three calculation methods from manual Slack export to automated tooling; and full engagement rate benchmarks cross-tabulated by community size and price tier.
TL;DR
Monthly member engagement rate — the percentage of your total paying member base who posted at least one message in the last 30 days — is the primary ongoing health metric for a paid community, distinct from activation rate (a new-member, week-one metric) and from the individual member health score (a week-one diagnostic tool). A community in the Healthy engagement tier (35–54% monthly posting rate) renews its 6-month cohort at 60–74%; a Dormant community (<20%) renews at 20–38%. The ~50pp gap between Thriving and Dormant tiers is price-agnostic. The most common cause of a declining engagement rate in a community that launched well is the ghost member problem: long-tenured members who activated correctly in week one, contributed in months 1–2, and then silently disengaged between months 3–6 as the content became repetitive or the channel architecture stopped matching their evolving goals. The diagnostic intervention — a personal operator DM to the top-20 longest-tenured dormant members — surfaces the failure mode in 40–55% of cases and costs the operator approximately 30 minutes per month.
Engagement rate versus activation rate: two metrics, two different questions
Paid community operators commonly conflate two distinct metrics: activation rate and engagement rate. They are related but measure fundamentally different things at different time horizons, and the interventions they prescribe are different.
Activation rate answers one question: “What percentage of members who joined in the last 7 or 14 days completed the activation event?” The activation event is typically an introduction post in the designated introduction channel. Activation rate is a new-member, cohort-level, week-one metric. It diagnoses whether the onboarding system is converting new joiners into contributors in their first week. The operator monitoring activation rate is asking: “Is my Day 0 DM producing the first post?”
Engagement rate answers a different question: “What percentage of my entire paying member base is actively contributing right now?” The measurement window is 30 days, the denominator is all paying members (not just recent joiners), and the metric is updated monthly. Engagement rate is a whole-community, ongoing metric. It diagnoses whether the community as a product is delivering value across all tenure cohorts — not just to the newest 5–10% who joined last week. The operator monitoring engagement rate is asking: “Is the community alive?”
The operational reason to track both: a community can have excellent activation rate and declining engagement rate simultaneously, and the cause is not in the onboarding system. A community where 65% of new members post in week one but only 15% of total members posted last month is losing long-tenured members faster than new members are being added to the active base. Activation rate will not reveal this; engagement rate will. Conversely, a community with a 55% total engagement rate but a 25% new-member activation rate is sustaining high engagement from its long-tenured base while failing to integrate new cohorts. Engagement rate will not reveal that specific gap; activation rate will. Both metrics are required to understand the full picture.
The third metric — and why engagement rate is distinct from it: The individual member health score is a third related-but-different metric. The health score is calculated per member, in the first 7 days of membership, using three behavioral signals (intro post completion, peer interaction count, active channel engagement depth), and produces a four-tier assignment (Activated / Healthy / At-Risk / Churned) that determines the week-one intervention queue. Engagement rate, by contrast, is a community-level number calculated monthly for all members regardless of tenure. The health score tells you which individual members need which intervention in week one. Engagement rate tells you whether the community as a whole is sustaining contribution behavior beyond week one. A member who scored Activated in week one may show up as dormant in month 6 if the community’s content and programming have not evolved to match their changing goals. The health score is a first-week predictor; engagement rate is the ongoing measurement of whether the predictor’s promise was kept.
Table 1 — Engagement rate formula and measurement parameters
The engagement rate formula has four parameters that must be defined consistently for the metric to be comparable across months: the numerator definition, the denominator definition, the measurement window, and the exclusions. Each parameter has a correct specification and a common misspecification that inflates the apparent engagement rate without reflecting actual member contribution behavior.
| Parameter | Correct specification | Common misspecification | Why the misspecification inflates the metric | Practical measurement |
|---|---|---|---|---|
| Numerator Who counts as “engaged” |
Count of unique paying members who sent at least one message in any non-announcement, non-bot-only channel during the 30-day window. Each member is counted once regardless of how many messages they sent. | Count of total messages sent during the 30-day window; or count of unique users who opened the Slack workspace (login count). | Total message count inflates the numerator with high-volume contributors: a community where 10 members each sent 50 messages looks the same as one where 500 members each sent one message. Login count counts passive presence as engagement: a member who opens Slack daily to read announcements contributes 30 “logins” but zero messages, inflating the metric without producing the contribution behavior that predicts renewal. | Manual: Slack workspace export JSON → count distinct user IDs across all non-announcement channels in the 30-day window, filtered to paying member IDs. Semi-automated: Slack API conversations.history filtered by date range and paying member list. Automated: purpose-built tooling tracks message events per member continuously. |
| Denominator Who is in the base |
Total active paying members at the end of the 30-day window. A member is “paying” if their subscription is active (not cancelled, not in grace period, not on a comped trial) on the last day of the measurement window. | Total workspace members (including non-paying guests, alumni with retained access, or comped accounts); or total members who joined at any point in the community’s history including cancelled members. | Including non-paying workspace members inflates the denominator with accounts that were never part of the paying membership, depressing the apparent engagement rate below the true metric for the paying base. Including cancelled members makes a healthy community appear less engaged than it is. | Requires maintaining a paying member list updated monthly as members join and cancel. The source of truth is the payment platform (Stripe, Memberstack, Gumroad) cross-referenced with the Slack workspace member list. This cross-reference is the most operationally difficult part of the manual engagement rate calculation at scale. |
| Window The 30-day period |
Rolling 30-day window ending on the measurement date. Recalculate on the same calendar date each month (e.g., the 1st of each month) for month-over-month comparability. The window includes the full 30-day period regardless of when members joined: a member who joined 5 days ago counts if they posted in those 5 days. | Calendar month (January = 31 days, February = 28 days); or 7-day window; or 90-day window. | Calendar month length variation introduces noise: members who post once per week contribute 4 or 5 posts depending on the month length. A 7-day window captures weekly fluctuations rather than the monthly contribution pattern. A 90-day window is too forgiving: it counts members who posted once in the last 3 months as engaged, systematically overstating the active poster base. | Set a fixed recalculation date (typically the 1st of each month). Run the calculation on that date using a strict 30-calendar-day lookback. Document the date of each calculation so month-over-month comparisons are on the same cadence. |
| Exclusions What does not count as engagement |
Exclude: messages sent by bots or integrations (Zapier, Slack workflows, calendar bots, announcement bots); messages in announcement-only channels where members cannot post; reactions and emoji responses; messages from operator or community manager accounts. | Counting all messages including bot messages; counting reactions as posts; including operator or staff posts in the unique-member count. | Bot messages represent zero member engagement. Reactions are a weaker signal than posts and inflate the rate to include members who are present but not contributing content. Operator posts in the numerator measure staff activity, not member engagement. | In Slack API data, bot messages have subtype: bot_message or user IDs prefixed with B rather than U. Filter these before calculating the numerator. Maintain a separate list of staff Slack user IDs to exclude from the numerator. |
The formula with correct parameter specifications:
Engagement rate = (unique paying members with ≥1 non-bot message in any non-announcement channel in the last 30 days ÷ total active paying members on the measurement date) × 100
Example: a community with 180 paying members on July 31st, where 82 unique paying members sent at least one message between July 1st and July 31st, has an engagement rate of 82 ÷ 180 × 100 = 45.6% — in the Healthy tier (35–54%). If those 82 posts include 38 from the operator’s announcement account (excluded as staff), the real numerator is 55 unique paying member IDs, yielding 55 ÷ 180 = 30.6% — Declining tier. The parameter exclusions matter.
The 30-day window and the episodic poster: The 30-day window is intentionally generous because the goal is to capture value-exchange participants, not daily-habit posters. A community whose members post roughly once per month will see approximately 65–75% of members appear in any given 30-day window due to the random distribution of monthly posts across calendar days. The benchmark thresholds account for this: Healthy (35–54%) means that even accounting for the episodic-poster distribution, more than one in three paying members is actively contributing. Communities that fall below 20% (Dormant) are not missing episodic posters — they have a structural contribution problem where the majority of paying members have stopped generating content value in the community entirely.
Table 2 — Four engagement health tier thresholds and 6-month renewal benchmarks
The four tiers are calibrated against 6-month renewal rate benchmarks rather than 3-month benchmarks because engagement rate is an ongoing metric that reflects the community’s sustained value delivery. The 6-month mark is where the compounding effect of consistent contribution (or consistent absence of contribution) becomes fully visible in renewal data. At month 3, the engagement rate differential between Thriving and Dormant communities is present but smaller; by month 6, the 50pp gap in renewal rates is fully established.
| Tier | Engagement rate | What this means in practice | 6-month renewal benchmark | Primary operator risk | % of paid communities at tier |
|---|---|---|---|---|---|
| Thriving | ≥55% | More than half of all paying members posted at least once in the last 30 days. The community has an active contribution culture where posting is the norm rather than lurking. At 55%+ engagement rate, the community generates enough peer-to-peer value exchange that members who are not currently posting still experience value by reading active threads — which sustains their renewal decision even in months when they themselves do not post. | 74–88% (6-month renewal) |
Contribution inequality: at 55%+ engagement rate, there is a risk that the active-poster base is dominated by 15–20% of members who post very frequently, while the remaining 35–40% who post once per month are one negative experience away from going dormant. Monitor whether engagement rate is driven by many members posting moderately or a few posting heavily. Heavy-poster concentration creates a community that feels alive in aggregate but brittle if a small number of heavy contributors leave. | 12–18% of paid communities that have been running 6+ months with no structured onboarding; 28–40% of communities with structured onboarding (Day 0 DM + Day 3 nudge + weekly programming) |
| Healthy | 35–54% | Between one in three and one in two paying members posted in the last 30 days. The community has an established contribution culture, but a significant minority of paying members are in passive-consumption mode. In a community of 200 paying members at 40% engagement rate, 120 members have not posted in 30 days. Whether those 120 members renew depends on the value of the content produced by the 80 who are posting — and on whether lurking members have a clear mental model of where to re-engage when ready. | 60–74% (6-month renewal) |
Goal drift among the lurking 46–65%: members in passive-consumption mode are renewing because content from active posters is still valuable to them. The risk is that as content volume grows, lurking members lose their ability to find threads most relevant to their current goal. Monthly programming that surfaces goal-specific curated content (a weekly “best threads” roundup or topic-specific resource drops) is the primary intervention for sustaining renewal in the lurking segment. | 28–38% of paid communities at 6+ months with basic onboarding in place |
| Declining | 20–34% | Between one in five and one in three paying members posted in the last 30 days. The active-poster base is insufficient to generate consistent peer-to-peer value for the lurking majority. Members in the lurking segment who do not receive value from the passive content experience are making an active renewal decision: the community has not delivered visible value in the last 30 days. Communities in the Declining tier often have perfectly functional onboarding for new members; the problem is that long-tenured members who activated correctly in their first week are now in month 3–6 and have quietly stopped contributing. | 38–55% (6-month renewal) |
Silent churn acceleration: the Declining tier is where monthly churn rate starts compounding. Each month that engagement rate stays in the 20–34% range, a new set of previously-Healthy members crosses the 30-day posting threshold and enters the dormant category — which in turn reduces the active-poster pool that sustains remaining members, creating a negative feedback loop. Communities that stay in Declining for 2+ consecutive months without intervention typically reach Dormant within 3–4 months. | 22–30% of paid communities at 6+ months; the most common tier for communities that launched well but did not develop recurring programming |
| Dormant | <20% | Fewer than one in five paying members posted in the last 30 days. The community has crossed the threshold where passive-consumption value is no longer sufficient to sustain renewal decisions: members who are not posting are also not finding meaningful content from the small active-poster base. At <15% engagement rate, even the active-poster cohort is often posting without receiving replies — which accelerates their own transition to dormancy. | 20–38% (6-month renewal) |
Structural breakdown: the primary risk in the Dormant tier is not that individual members are about to cancel — it is that the community itself has lost the social proof that sustains renewal decisions for passive members. When a member opens Slack and sees that no one has posted in 3 days, the implicit signal is that the community is no longer active. This signal accelerates cancellation decisions independently of the individual member’s personal engagement level. Recovering from Dormant tier requires a structural intervention, not an incremental content adjustment. | 18–28% of paid communities at 12+ months without structured retention programming; 5–12% with structured onboarding and monthly programming |
The tier distribution percentages in the final column are the most useful diagnostic for benchmarking your community against comparable ones. A community that launched 12 months ago with structured onboarding and no ongoing programming typically sits in the Declining tier at month 12: the initial cohort of Activated members from the first 2–3 months has drifted toward dormancy as the content cycle became repetitive, while new member cohorts are activating at their normal first-week rates but the absolute number of active posters is not growing fast enough to sustain a Healthy tier engagement rate against the expanding total paying member base.
The ghost member problem: The most common cause of a declining engagement rate in a community that launched well is the “ghost member” problem: paying members who activated correctly in week one, contributed actively in months 1–2, and then silently stopped posting between months 3–6 without cancelling their subscription. Ghost members continue paying — they are in the denominator of the engagement rate formula but not the numerator — because the psychological cost of cancelling a subscription they have invested time in is higher than the cost of continuing to pay for a community they are not using. Ghost members inflate the denominator without contributing to the numerator, mechanically depressing the engagement rate. They also do not cancel immediately, which creates a false signal in monthly revenue numbers: the operator sees stable MRR while the engagement rate declines, and does not diagnose the ghost member problem until the renewal cohort begins cancelling 3–6 months after going dormant. The ghost member audit (described in Table 3) is the primary tool for diagnosing this pattern before it compounds into a cancellation wave. See the churn by tenure reference card for the full month-by-month churn rate distribution and the inflection point where ghost-member cancellations typically concentrate.
Table 3 — Per-tier operator intervention
Each engagement health tier produces a different diagnostic question and a different primary intervention. The interventions are sequenced by the tier’s underlying failure mode: Thriving communities need monitoring and celebration; Healthy communities need goal-specific content routing; Declining communities need a ghost member audit and personal outreach to the highest-tenured dormant members; Dormant communities need a structural audit of the entire community product before attempting any tactical intervention.
The interventions are designed around the principle that engagement rate is a lagging indicator: by the time the metric shows a tier decline, the failure mode has already been operating for 30–60 days. The interventions address the root cause that produced the decline, which will surface in the engagement rate 30–60 days after the intervention is implemented — not in the same measurement period.
| Tier | Primary intervention | Timing | Format | Expected engagement rate response | Common mistake |
|---|---|---|---|---|---|
| Thriving | Contribution recognition programming: monthly member spotlight, “top thread” curation, or cross-cohort introduction facilitation for highly active members | Monthly, on a fixed schedule (e.g., first Monday of each month) | A pinned channel post naming 3–5 members whose contributions in the last 30 days produced the most valuable peer interactions. Focus on members whose posts generated the most replies from other members, as this signals posts that produced the most peer-to-peer value exchange. Follow with a direct DM from the operator to each spotlighted member acknowledging the specific contribution. | Thriving communities sustain their engagement rate with recognition programming, not grow it. The goal is to reinforce the contribution culture by making visible the norm that posting produces recognition and reciprocal value — sustaining episodic posters in their posting pattern and signalling to dormant members that contribution is rewarded. Expected effect: 2–4pp sustained improvement in the episodic-poster segment from the social proof effect of visible recognition. | Focusing recognition on message volume rather than value produced. A member who sent 80 messages last month asking questions contributed less peer-to-peer value than a member who sent 10 messages that each generated 8-reply threads. Recognising volume trains the community that posting frequency is the currency; recognising replies-generated trains the community that contribution quality is the currency. |
| Healthy | Goal-specific content curation: weekly thread roundup or resource drop in 2–3 goal-relevant channels designed to give lurking members a frictionless re-entry point to posting | Weekly, on a consistent day (e.g., Thursday morning, when asynchronous community engagement is highest) | A brief post in 2–3 non-introduction, non-announcement channels surfacing one specific valuable thread from the last 7 days (with a direct link) and one open question inviting response. The open question converts the lurker from reader to poster with a single low-effort reply. Questions that produce the highest re-engagement are ones that require one sentence rather than one paragraph, reference something the lurking member likely has direct experience with, and do not require community-specific knowledge to answer. | 15–25% of members who see the weekly curation post click through to the linked thread; of those, 8–14% post a reply. Over 4 weeks of consistent weekly curation, Healthy-tier communities see engagement rate improvement of 4–8pp as dormant members who re-engage via curation posts return to more regular posting patterns. | Posting the weekly curation in #announcements or #general rather than in goal-specific channels. Members in the lurking segment are most likely to re-engage when they see content directly relevant to their stated goal. A resource drop in #growth-marketing will pull back growth marketer members who have been lurking; the same resource drop in #announcements will be seen by the whole community but acted on by very few, because it lacks the goal-signal that makes it personally relevant to any individual member. |
| Declining | Ghost member audit + personal DM to top-20 longest-tenured dormant members (members who have not posted in 30+ days, sorted by longest membership tenure) | Immediately upon identifying the Declining tier assignment; repeat monthly until tier improves | Step 1 — ghost member audit (20–30 min): pull a list of all paying members who have not posted in 30+ days, sorted by join date descending. The longest-tenured dormant members are the highest-priority diagnostic targets because their dormancy reveals what happens to a Healthy-tier member after 6–12 months — the failure mode today’s Healthy members will reach in 3–6 months. Step 2 — personal DM (30–45 min): send a direct Slack DM from the operator’s own account to the top-10–20 longest-tenured dormant members. One question only: “I noticed you’ve been quiet recently — is there something specific you were hoping to find here that you haven’t been finding?” No link, no CTA, no onboarding content. | 40–55% response rate from personal operator DMs to long-tenured dormant members; of respondents, 35–50% identify a specific content or channel gap; of gap-identifying respondents, 20–35% re-engage within 30 days when the operator addresses the identified gap. Engagement rate improvement from one ghost member audit cycle: 3–7pp, visible in the following month’s measurement. | Sending the ghost member outreach as a broadcast (e.g., a channel post asking “we miss you — what can we do better?”). Broadcast outreach to dormant members reads as a mass email. The personal DM from the operator’s account, referencing the specific member by name and asking a single question, produces the 40–55% response rate. Response rate on broadcast outreach to dormant members: 3–8%. |
| Dormant | Structural community audit: three-part assessment of (1) channel architecture, (2) content programming calendar, and (3) week-one activation rate for the last 3 new-member cohorts | Within 2 weeks of identifying Dormant tier; the audit is a prerequisite for any tactical intervention | Part 1 — channel audit (1–2 hours): list all active channels; calculate message volume per channel in the last 30 days excluding bot/announcement messages; identify channels with zero or <3 member messages. A Dormant community typically has 3–5 channels that account for 85%+ of all member messages, and 10–20 channels where no member has posted in 30+ days. Archive dead channels. Part 2 — programming calendar audit (30 min): review the last 12 weeks of programming events; calculate average member posts in the 48 hours after each event vs. before. If programming events are not producing a measurable posting spike, the programming format is not creating the re-entry opportunity it needs to. Part 3 — new member activation review (20 min): pull activation rate for the last 3 new-member cohorts. If activation rate is below 35%, fix onboarding before trying to fix programming. | Dormant tier communities that complete the three-part audit and act on its findings typically see engagement rate improvement of 8–15pp over 60–90 days, returning to Declining or Healthy tier. Communities that skip the audit and attempt tactical fixes (more content, more events, broadcast outreach) in Dormant tier see 0–3pp improvement because the tactical fix is not addressing the structural cause. | Treating a Dormant engagement rate as a content problem. The instinct when engagement drops is to produce more content. In a Dormant community, the content problem is often the opposite: there is too much low-value content (announcements, reposted links, operator monologue posts) and not enough peer-to-peer content (member questions, member-to-member recommendations, collaborative threads). More content in this environment accelerates dormancy by worsening the signal-to-noise ratio problem that is already making the community feel overwhelming to lurking members. |
The intervention table is designed for monthly use: at the end of each 30-day measurement period, the operator calculates the engagement rate, assigns the tier, and implements the corresponding intervention. The intervention will be visible in the next month’s engagement rate — not the same measurement period. This 30–60 day lag between intervention and measurement is the primary reason engagement rate is difficult to manage reactively: operators who implement an intervention in month 1 and expect to see improvement in month 1’s measurement are measuring before the intervention has had time to work.
Why engagement rate drops in months 3–6: The most predictable engagement rate decline pattern occurs between months 3–6 of the community’s operation. The cause is a structural mismatch between the content needs of members at different tenure stages. Month-0–2 members need foundational content: introductions, basic resources, orientation to community norms. Month-4–6 members need advanced or evolving content: current debates, emerging practices, peer-to-peer collaboration on specific problems. If the community’s programming calendar was designed for new members — which is common in communities that prioritize onboarding correctly — month-4–6 members start finding the content repetitive and begin lurking. Their lurking reduces the peer-to-peer thread activity that new members find valuable, which also reduces new member activation over time. The solution is a tiered content calendar: foundational content for new members in specific channels, advanced content in separate channels that month-4+ members can self-select into, and programming events that serve multiple tenure stages simultaneously. See the member retention reference card for the monthly cohort review framework that identifies this pattern before it reaches Declining tier.
Table 4 — Engagement rate calculation method by community size and tooling
Engagement rate can be calculated from Slack data using three methods with different setup requirements, time costs, and accuracy levels. Unlike the individual member health score (calculated once per member at Day 7), the engagement rate must be calculated monthly for every member in the paying base — which makes data volume grow linearly with community size and makes the correct method choice more consequential as the community grows.
| Method | Setup time | Monthly calculation time | Data required | Accuracy | Best for |
|---|---|---|---|---|---|
| Manual Slack export | 0 setup — available immediately to any workspace admin | 45–90 min per monthly calculation for communities under 100 paying members. Steps: (1) download Slack workspace export from Workspace Settings → Import/Export Data (JSON); (2) build a list of paying member Slack user IDs from the payment platform; (3) search channel JSON files for unique user IDs that posted in the 30-day window, excluding bot IDs and announcement channels; (4) divide by total paying members and multiply by 100. | Slack workspace export (JSON; requires Slack Pro/Business+ plan); a list of paying member Slack user IDs from the payment platform; a spreadsheet to tally unique poster IDs and calculate the rate. | Moderate. Manual review introduces human error in the user ID cross-reference and the bot-message exclusion step. At under 100 members the error rate is low enough that manual calculation is reliable; above 100 members the probability of miscounting grows significantly. | Best for communities with <100 paying members. At 100 members the manual export review takes 60–90 minutes; the calculation accuracy risk starts to exceed the time saved vs. the semi-automated approach. At 200+ members, manual calculation is not recommended. |
| Semi-automated (Slack API + spreadsheet) |
3–6 hours of initial setup. Requires: (1) creating a Slack API key from api.slack.com with channels:read, channels:history, and users:read scopes; (2) building a spreadsheet or script that pulls unique message senders per channel for a 30-day window; (3) cross-referencing with the paying member list from the payment platform. Can be templated and reused monthly after initial setup. |
15–30 min per monthly calculation after initial setup. The spreadsheet auto-populates the unique poster count from the API pull; the operator updates the paying member denominator and reviews the tier assignment. | Slack API key with appropriate scopes; a spreadsheet or script that automates the API calls and unique-poster-count calculation; a current paying member list from the payment platform (updated monthly). | High. API data is more reliable than manual export review for catching bot messages (the API returns subtype fields that are easier to filter programmatically), and the spreadsheet formula prevents human error in the tallying step. |
Best for communities with 100–500 paying members. Above 500 members, the Slack API rate limits on message history retrieval may require batching across multiple days, which introduces a timing inconsistency in the 30-day window calculation. |
| Automated (purpose-built tooling) |
One-time Slack OAuth install (<5 minutes). Signal capture begins at the next member join or message event after install. The engagement rate is calculated continuously and displayed in a dashboard. | <5 min per monthly review. The engagement rate dashboard shows the current month’s tier assignment, the active-poster count, and the dormant member list sorted by tenure; the operator reviews the dashboard and queues the tier-appropriate intervention. | Slack OAuth install with message.channels and member_joined_channel scopes; integration with the payment platform to maintain the paying member list automatically. No workspace export or manual API key management required. |
Highest accuracy. Event-driven message capture eliminates batch timing inconsistency; automatic bot-message exclusion; automatic paying-member denominator updates when members join or cancel via the payment platform integration. | Appropriate for any community size above 50 paying members. The monthly review time is consistent regardless of community size because the dashboard presents pre-calculated tier assignments rather than requiring the operator to process raw data. |
The transition from manual to semi-automated is appropriate at approximately 100 paying members — when the monthly manual export review crosses 60 minutes and the human error risk in the user ID cross-reference becomes meaningful. The transition from semi-automated to purpose-built tooling is appropriate at approximately 300–500 members, where API rate limit batching and paying-member denominator management become the primary accuracy risks. For communities tracking both individual member health scores and community-level engagement rate, purpose-built tooling that handles both calculations from a single OAuth install is almost always more efficient than maintaining two separate manual or semi-automated workflows.
The paying member denominator management problem: The most operationally difficult part of the engagement rate calculation is not the numerator (pulling the unique poster count from Slack data) but the denominator (maintaining an accurate count of total paying members updated when members join or cancel). Payment platforms (Stripe, Memberstack, Gumroad) do not automatically remove cancelled members from Slack workspaces; Slack does not automatically sync with payment platform subscription status. Communities managing the engagement rate calculation manually must maintain a separate “paying member Slack IDs” list updated every time a member joins (adding their Slack ID) or cancels (removing it). At 200+ paying members with typical monthly churn of 5–8% and new-member join rates of 10–15%, this list requires 30–50 updates per month to remain accurate. An inaccurate denominator is the most common source of engagement rate calculation error in manually-managed communities — typically an inflated denominator (including cancelled members who still have Slack access) that depresses the apparent engagement rate below the true number.
Table 5 — Engagement rate benchmarks by community size and price tier
Engagement rate benchmarks vary by community size more than by price tier. The primary driver is operator-bandwidth: smaller communities (under 100 members) can sustain higher engagement rates because the operator has personal bandwidth for individualised outreach to dormant members and active thread participation as a conversation catalyst. Larger communities (300+ members) typically see lower engagement rates because the operator cannot personally DM dormant members at scale, and the community must sustain contribution culture through programming and channel architecture rather than direct operator outreach. Price tier affects the absolute renewal rates associated with each engagement tier (higher-priced communities sustain higher renewal rates at the same engagement tier due to stronger self-selection motivation) but does not materially change the engagement rate benchmarks themselves.
| Community size (paying members) |
Typical engagement rate range | Thriving threshold (achievable without tooling) |
Primary engagement driver | Primary dormancy driver | 6-month renewal at Healthy tier (by price tier) |
|---|---|---|---|---|---|
| Under 100 members (early-stage community) |
30–65% (wide range; highly operator-dependent) |
Achievable manually. Operators of sub-100-member communities who run active weekly programming (one AMA or office hours per week) and personal outreach to dormant members typically sustain 55–65% engagement rate with 5–8 hours per week of operator time. | Operator direct participation: in communities under 100 members, the operator’s own posting activity accounts for a significant share of total thread activity. Operator DMs to dormant members produce 55–70% response rates (vs. 40–55% in larger communities) because the personal relationship between operator and member is stronger at this size. The operator is the primary engagement catalyst. | Operator bandwidth depletion: as the community grows from 30 to 100 members, the operator’s weekly time cost for maintaining personal outreach and active participation grows linearly. Communities that do not systemize engagement programming before reaching 60–80 members typically experience engagement rate decline at 80–100 members when operator bandwidth for personal outreach is fully consumed. |
Under $50/mo: 58–72% $50–$149/mo: 64–78% $150–$299/mo: 68–82% $300+/mo: 72–86% |
| 100–300 members (growth-stage community) |
25–50% (structured programming required to sustain Healthy tier) |
Achievable with systematic programming. Communities at this size sustain Thriving tier (≥55%) only with a documented weekly programming calendar, channel curation (archiving dead channels, creating goal-specific channels based on member stated goals), and a semi-automated engagement rate calculation that surfaces dormant members monthly for personal outreach. | Programming cadence: at 100–300 members, the operator can no longer sustain engagement through personal outreach alone. The primary engagement driver shifts from operator-direct participation to recurring programming events (weekly office hours, bi-weekly AMAs, monthly resource drops) that give the full member base a predictable re-entry point for posting each week. | Content repetition: at this size, the community typically has 6–18 months of posting history, and topics that generated high engagement in months 1–3 have been revisited enough that they no longer generate the same reply rates. Long-tenured members who remember the original thread often lurk rather than repost. A searchable content archive and a curated “best threads” channel prevent the repeated-topic problem from becoming the primary dormancy driver. |
Under $50/mo: 52–66% $50–$149/mo: 58–72% $150–$299/mo: 64–78% $300+/mo: 68–82% |
| 300–1,000 members (scale-stage community) |
18–40% (structured sub-communities required to sustain Healthy tier) |
Difficult without structural intervention. Communities above 300 members that sustain Thriving tier typically do so through sub-community architecture: not a single large community but a set of smaller, goal-specific channels (each with 30–80 active members) that sustain their own engagement culture independently. The engagement rate of the full community is the average of its sub-community engagement rates; a community with five 60-member sub-communities each at 55% engagement will report 55% overall even though no individual member engages with the full 300-member community. | Sub-community self-organization: at 300–1,000 members, the channels that sustain the highest engagement rates are typically the ones that coalesced around a specific goal or interest group and developed their own norms and regulars. The operator’s primary engagement driver is identifying these sub-communities and providing them with resources (dedicated channels, facilitated events, member spotlights) that reinforce their self-organization. | Audience mismatch: at scale, the community’s founding member persona and current new-member persona are often different. A community that launched targeting senior product managers may attract more junior members as it grows, and the content quality and peer interactions that senior members joined for are diluted by a growing volume of more basic questions. This audience mismatch is the primary driver of long-tenured member dormancy at 300+ members. See the retention reference card for the cohort review framework that surfaces this pattern before it compounds. |
Under $50/mo: 45–60% $50–$149/mo: 52–68% $150–$299/mo: 60–76% $300+/mo: 65–80% |
| 1,000+ members (mature community) |
12–28% (sub-communities and structured programming required; direct operator engagement not scalable) |
Requires purpose-built tooling and sub-community architecture. Communities above 1,000 paying members that sustain engagement rates above 35% are structured as federations of sub-communities with dedicated facilitators or community managers, a content programming team, and automated engagement rate tracking that surfaces dormant sub-communities before they drag down the community-wide metric. | Facilitated peer connections: at 1,000+ members, the random-encounter probability that any two members with compatible goals end up in the same thread drops significantly. Communities that sustain high engagement at this scale invest in facilitated peer connection — structured introductions, working groups, cohort programs — that create intentional peer-to-peer connections rather than relying on ambient channel posting. | Community bureaucracy: as a community scales past 1,000 members, the time required to get approval for programming, respond to member requests, and manage channel governance often slows the operator’s responsiveness to the engagement patterns the monthly metric reveals. The primary dormancy driver at this scale is organizational latency: the gap between identifying that a sub-community is declining and implementing the intervention is long enough that the decline compounds into cancellation before the fix is deployed. |
Under $50/mo: 38–54% $50–$149/mo: 45–62% $150–$299/mo: 55–72% $300+/mo: 62–78% |
The benchmark ranges in Table 5 are wide at each community size because operator execution quality — the quality of weekly programming, consistency of ghost member outreach, channel architecture decisions — accounts for a larger share of the engagement rate variance than community size or price tier alone. A 200-member community with an operator who runs a consistent weekly programming calendar and monthly ghost member audit will sustain 45–55% engagement rate; the same-sized community with no programming cadence will settle at 18–28%. The 20–30pp range within each community size tier is the operator’s execution premium — the share of engagement rate determined by how the operator runs the community, not by its structural characteristics.
The engagement rate and the three-touch welcome sequence: Engagement rate and the onboarding welcome sequence (Day 0 DM, Day 3 conditional nudge, Day 7 scorecard) operate at different time horizons but are causally linked. The welcome sequence determines activation rate (the percentage of new members who post in week one); activation rate determines how many active posters each new cohort contributes to the monthly denominator; those contributions either sustain or degrade the month-over-month engagement rate. A community that improves its new-member activation rate from 35% to 55% by implementing a structured Day 0 DM and Day 3 conditional nudge will see its engagement rate improve by approximately 4–8pp over the next 3–4 months as the higher-activation cohorts accumulate in the active-poster base. The reverse is also true: a community whose onboarding breaks will see its engagement rate decline 3–4 months after the onboarding break, when the cohorts that failed to activate enter the dormant segment of the denominator without contributing to the numerator.
Putting the five tables together: the monthly engagement rate review routine
The five decision tables in this reference card are designed to be used together as a monthly review routine. On a fixed date each month (typically the first business day), the operator runs the engagement rate calculation using the method appropriate for the community’s size (Table 4), assigns the community to a tier using the threshold table (Table 2), and implements the corresponding intervention (Table 3). The formula parameters (Table 1) are applied consistently each month to ensure the engagement rate is comparable across periods. The benchmark table (Table 5) provides context for evaluating whether the current engagement rate is above or below what comparable communities at the same size and price tier typically achieve.
The full monthly review for a community of 150 paying members using the semi-automated API method takes approximately 30–45 minutes: 15 minutes to run the API pull and update the paying-member denominator, 5 minutes to calculate the engagement rate and assign the tier, and 15–25 minutes to implement the tier-appropriate intervention. At $99/mo, the expected return on a 30–45 minute monthly review — from the engagement rate improvement that sustains 6-month renewal rates above the Healthy tier threshold — is approximately $1,200–$2,400 per month in retained subscription revenue that would otherwise cancel at month 6 without the engagement intervention.
The engagement rate metric and the individual member health score are complementary diagnostic tools that together give the operator complete visibility into both the onboarding system (health score, week-one) and the sustained community product (engagement rate, monthly). The health score tells you which new members to intervene with in week one; the engagement rate tells you whether the community is delivering the value those week-one activations promised. Neither metric alone is sufficient: the health score without engagement rate monitoring leaves the operator blind to month-3–6 dormancy; the engagement rate without health score monitoring leaves the operator without the week-one intervention data that determines which new cohorts will sustain the engagement rate 3–4 months later.