The community metrics worth tracking are the ones that show whether members come back and take part: the active-member ratio, how many new members make a first post, how fast people get a reply, how many members are still around after 30, 60 and 90 days, and how much content members create compared with how much they only read. Total member count is the one number to demote. It only goes up, so it tells you how old your community is, not how healthy it is. Below is a working set of online community KPIs, how to calculate each, and how to avoid the ones that flatter you.
Third-party guidance in this post comes from FeverBee and CMX pages we fetched in October 2026. Where a metric is our own suggestion, we say so.
In this guide
- Why member count is a poor measure of community health
- The metrics that matter, in the order a member meets them
- Activation and new-member first-post rate
- Retention and the active-member ratio
- Response time
- Content created versus content consumed
- Member satisfaction
- Conversion to paid, for monetised communities
- How to avoid vanity metrics
- What BuddyNext documents for analytics
- A monthly review you can run in an hour
- Questions people ask
What are community metrics, and why is member count not enough?
Community metrics are numbers that describe how members behave on your site: who joins, who takes part, who stays and who leaves. Member count measures only the first of those.
Richard Millington of FeverBee makes the point bluntly in a post on bogus metrics. Registered members, he writes, is a count that only rises as long as accounts stay open, so it reflects the age of the community more than its success (FeverBee, Don’t Use Bogus Metrics). A site with 8,000 registered members and 120 people posting is a small community with a large mailing list.
The same post goes further. It argues that no single metric captures the value of a community and that every metric can be flawed or gamed. That is useful advice before you build a dashboard. You are not looking for the one number. You are looking for a short set of numbers that disagree with each other when something is wrong.
FeverBee’s guidance on whether a community is healthy asks for more than one data point and groups the numbers into three areas: membership, activity and sense of community (FeverBee, Is your community healthy?). That grouping is a good skeleton, and this guide follows it, adding the metrics a monetised community needs.
One more definition before we go on. A KPI (key performance indicator) is a metric you have attached a target and an owner to. Every KPI is a metric, but most metrics should never become KPIs. Pick a few.
Which community metrics should you track first?
Start with six: new-member first-post rate, active members in the last 30 days, the active-member ratio, 30-day retention, time to first reply, and content created versus consumed. Add satisfaction and paid conversion once those are stable.
Here is the whole set in one table, with how to calculate each and the mistake it most often hides. The formulas are our own working definitions, not an industry standard, so write down your own and keep them the same month to month.
| Metric | How to calculate it | What it tells you | Easy way to fool yourself |
|---|---|---|---|
| Activation (first-post rate) | New members who post, reply or react within their first 7 days, divided by new members that week | Whether the first visit leads anywhere | Counting a profile edit as activity |
| Active members | Distinct members with at least one action in the last 30 days | Direction of the community | Counting logins that do nothing |
| Active-member ratio | Active members divided by total registered members | How much of your list is alive | Reading a falling ratio as bad when you are growing fast |
| Retention | Share of a join-month group still active at 30, 60 and 90 days | Whether people stay | Averaging old and new members together |
| Response time | Median time from a post to its first reply | Whether a newcomer feels heard | Using the average, which one forgotten post ruins |
| Created versus consumed | Posts, replies and uploads per 100 active members, next to views per active member | Whether members take part or only watch | Celebrating views alone |
| Satisfaction | A short survey on a regular rhythm | How members feel, which logs never show | Mixing up product feedback with community feedback |
| Paid conversion | Free members who become paying members, divided by free members in the same period | Whether the community earns | Ignoring how many paying members cancel |
Notice the order. It follows what a member goes through: arrive, do something, come back, get answered, contribute, feel good about it, and perhaps pay. If a number late in that chain looks bad, the cause is usually earlier in the chain.
What is activation, and how do you measure the new-member first-post rate?
Activation is the first meaningful thing a new member does. In most communities that is a post, a reply or a reaction. The first-post rate is the share of new members who do it within a set window, for example their first seven days.
FeverBee’s list of core metrics includes the contribution rate, the share of members who post, and warns that a low rate may mean you are collecting lurkers. It also lists newcomer discussion starts, which show whether new members feel intimidated or unmotivated (FeverBee, measuring). Both are different ways of looking at activation.
The first-post rate is the single best early warning you have. Sign-ups can grow while it quietly falls, and that is the pattern of a community that is advertised well and welcomes badly.
To measure it:
- Choose the action that counts. Be strict. A reply or a first post counts. A profile photo does not.
- Choose the window. Seven days is a sensible start for a small community. Write it down and do not change it.
- Take everyone who joined in a given week, and count how many did the action inside the window.
- Divide, and record the number against the join week, not against the date you looked.
To confirm it worked, pick three members from that week by hand and check that your count would have included them.
If the rate is low, the fix is rarely more promotion. It is a better first step. Give new members one easy, named thing to do, such as an introductions thread (on a Reign site you can put that prompt on the first screen they see), and answer the first person who does it quickly. Our guide to launching a community and getting your first 100 members covers that opening stretch, and the metrics here are what you check while you do it.
How do you measure community retention?
Measure retention by cohort. A cohort is a group of members who joined in the same week or month. You check what share of each group is still active after 30, 60 and 90 days, and you compare groups against each other.
Why not just count active members overall? Because one large, loyal founding group can hide a leaky bucket. If your first 200 members stay forever and every group since has lost most of its people within a month, the overall number looks fine while the community is failing newcomers. Cohorts show it. The blended average does not.
FeverBee suggests a low-tech version: review about ten newcomers from three months earlier and look at where they dropped out (FeverBee, measuring). For a small community that hand review is often more useful than a chart, because it tells you why, not only how many.
CMX, the community-professional network, says something similar from a business angle. A summary of Evan Hamilton’s talk on its site describes retention as the most powerful thing a community can do for a business, and also the hardest to measure (CMX). Hard to measure is the reason to start early and keep the method stable.
Steps for a first retention table:
- List members by the month they joined.
- For each month, count how many did at least one action in days 25 to 35, days 55 to 65, and days 85 to 95.
- Divide each count by the size of that month’s group.
- Put the months in rows and the three checkpoints in columns.
To confirm it worked, read down one column. If the 30-day figure for recent months is lower than for early months, you have learned something real, and it is worth a conversation with a few of those members.
What is the active-member ratio, and how do you read it?
The active-member ratio is the number of members active in the last 30 days divided by your total registered members. It tells you what share of your list is alive. A single reading means little. The trend, read alongside registrations, means a lot.
FeverBee gives a worked example. A community saw its share of engaged members fall from 7.88 percent in March to 5.03 percent in June, and the article’s answer is that you cannot judge that from one number. If registrations grew over the period, the ratio can drop even while more people take part, which points to weaker conversion of newcomers into regulars or to members staying for a shorter time. They call both a concern but hardly fatal. If registrations did not grow much, a falling ratio is a major warning that members are leaving (FeverBee).
The same article makes another practical point. Track active members in the previous 30 days as an absolute number, not only as a percentage, because the absolute number is the key indicator of direction. Both numbers belong on the same page.
FeverBee’s measuring guide adds a warning in plain terms: when the active-member number starts to drop, you have a serious problem and a limited amount of time to correct course. It also notes that if active members fall while total contributions rise, a core group may be dominating the conversation (FeverBee, measuring). Hold that thought for the content section below.
Two traps to avoid:
- A loose definition of active. Logging in is not activity. Use an action: a post, reply, reaction, RSVP or message.
- Changing the definition mid-year. If you widen it to make the number look better, you have destroyed your own comparison. If you must change it, restate the earlier months too.
How fast should you reply to new posts?
As fast as you can sustain, and measured honestly. Response time is the gap between a post and the first reply to it. Use the median, not the average, and track the share of posts that get no reply at all.
FeverBee lists reply speed among its sampled measures and says faster replies indicate higher social presence and participation (FeverBee, measuring). We will not quote a target number, because no honest number fits every community. A busy product-support forum and a weekly book club have different rhythms. Set your own target from your first month of data and tighten it slowly.
What matters for a newcomer is not the number but what they feel. Someone posts their first question and hears nothing for two days. They do not post again. That one gap can cost you the activation you worked for.
Two things lower response time without hiring anyone:
- A named welcomer. One or two people whose job is to answer every first post. Rotate them so it does not burn out one volunteer.
- A view of unanswered posts. A page, or a daily email digest, that lists posts with zero replies. If you cannot see them, you cannot answer them.
For a support-style community, response time is also where moderation and metrics meet. Slow replies and unresolved reports tend to travel together. Our community moderation guide for WordPress covers reports, strikes and auto-hide rules, and our piece on building a customer support community shows where answer speed matters most.
How do you compare content created with content consumed?
Count what members make, such as posts, replies, uploads and event RSVPs, and set it beside what they only view. A healthy community has a visible share of members who create, and a larger group who read and learn from them.
The ratio of creators to readers will always lean heavily toward readers. That is normal and not a flaw. The question is whether the creating group is wide or narrow.
FeverBee suggests several numbers that get at this: contributions per active member, which shows engagement and whether a few members dominate; responses per discussion, which shows whether discussions interest members; and new discussions, which shows whether the community is stuck on a few topics or has fresh ground (FeverBee). They also note that a smaller group of participants can be fine if those members are more active.
Here is a simple way to run it. Each month, write down:
- Total posts and replies
- The number of distinct members who wrote at least one of them
- The share of all posts written by your top ten posters
- Views per active member
If total posts rise while the number of distinct writers falls, and the top ten write most of it, you have a core group dominating. FeverBee’s suggested response is a separate space for that group, or bringing newcomers into the conversation, rather than praising the rising total.
How do you measure member satisfaction?
Ask members, briefly and on a schedule, and keep the questions the same so you can compare answers over time. Behaviour data tells you what people did. Only a question tells you how they felt about it.
FeverBee’s measuring guide suggests surveying members yearly with a modified Sense of Community Index, which tries to capture whether people feel they belong (FeverBee, measuring). A yearly survey with that kind of question suits a large community. A small one can ask more often, with just two or three questions.
Good short questions, which are our suggestions rather than a published standard:
- How likely are you to recommend this community to a colleague or friend, from zero to ten?
- What is the main reason you come back?
- What nearly made you leave, or has made you stop coming?
Free-text answers to the second and third questions are worth more than the score. Read every one.
There is a caution. In the bogus metrics post, Millington argues that member satisfaction can mislead when members do not separate satisfaction with the product from satisfaction with the community, so a weak product pulls the score down even if the community is doing well (FeverBee). That applies if your community sits around a product. Ask about the community specifically, and keep product questions in a separate survey.
Which metrics matter if your community is paid?
For a paid community, add three: conversion to paid, paid retention (how many paying members renew), and the activity gap between paying and free members. Conversion alone is half a story. A community that converts well and loses payers each month is on a treadmill.
Conversion to paid is free members who became paying members, divided by free members, within the same window. If you have a free tier or a trial, define the window before you look at the data. If you have no free tier, the equivalent is visitors or applicants who became members.
Paid retention is the share of paying members who renew. Cohort it the same way as general retention: group payers by the month they started and check who is still paying at 30, 60 and 90 days.
The activity gap compares what paying members do with what free members do. If payers are quiet, they are likely to cancel. Looking at activity before the cancellation is more useful than asking afterwards.
We will not quote prices, fees or conversion targets here. They vary too much to be useful and any number we gave would be invented. Check your payment provider and your local rules for anything touching tax or consumer rights.
The mechanics of taking payment are covered in our guide to building a paid community on WordPress with Stripe. Come back to this section once the first few cohorts have gone through a full billing cycle.
How do you avoid vanity metrics?
A vanity metric is a number that rises easily, looks good in a report and does not change what you do next. Test every metric with one question: if this number moved by half, would I act differently? If not, it is decoration.
Millington’s list of bogus metrics is a useful checklist. In his post, each of these is flawed or easy to game (FeverBee):
- Registered members, because the count only rises as accounts stay open.
- Questions asked or active members, when they depend mainly on how many customers the brand attracts and how many problems the product has, which a community manager does not control.
- Questions with an accepted solution, because the rate stalls once the easy questions are answered, and some managers inflate it by deleting questions they cannot answer.
- Share of newcomers who become active, because newcomer quality tends to decline over time as the most valuable customers join early.
- Visitor numbers, which depend on factors outside the community’s control such as marketing spend.
- Member satisfaction, for the reason given earlier.
- Call or ticket deflection, because it is unclear whether a visitor found the answer or gave up.
Read that list twice. Several of the metrics in this very guide appear on it. That is not a contradiction. It means each one can mislead if you treat it as a score instead of a signal. The same article also flags metrics that depend on things a community manager does not control, which is a good test for any number you pick.
Habits that keep you honest:
- Pair every number with its opposite. Posts with distinct writers. Sign-ups with first-post rate. Conversion with cancellations.
- Choose a goal first. CMX quotes a Udemy case study with the advice to pick one community measurement goal and optimise toward it relentlessly (CMX). Tie that goal to what the site is for, whether that is retention, support or revenue.
- Set a baseline. The same CMX page advises keeping a baseline in mind so you can monitor progress over time.
- Give each metric an owner. CMX’s Moz case study recommends that every community project serve a function, with an owner for each function.
Gamified numbers need extra care. Points, badges and leaderboards can lift activity, and they can also inflate it with low-value posts. If you run them, watch the quality of what gets posted, not only the volume. Our comparison of the best WordPress gamification plugins covers the options, and the same caution applies to all of them.
What does BuddyNext document for community analytics?
BuddyNext documents an Insights area that shows daily, weekly and monthly active members (DAU, WAU and MAU), cohort views and funnels. These need a paid licence, so check the product page for what your licence covers. Wbcom makes BuddyNext, so weigh this section with that in mind.
Plain definitions first. DAU, WAU and MAU are daily, weekly and monthly active users: distinct members who did something in that day, week or month. A cohort is a group of members who joined in the same period. A funnel is a sequence of steps, such as sign-up, first post, first reply received, with the drop-off counted at each step.
According to Wbcom’s own write-up on community analytics, the Insights tab sits under Engagement in BuddyNext, depending on your licence (check the product page), and shows DAU, WAU and MAU with week-over-week change. It describes the data as tracked, dated and exportable. Its Cohorts view groups members by join week and shows what share of each group returned in later weeks. Its Funnel view tracks steps you choose, for example sign-up, first post, first reaction received and first follow, and shows where people drop off (Wbcom Designs, BuddyNext community analytics). The BuddyNext product page lists an analytics dashboard, cohorts and member insights among its paid features (buddynext.com).
Map that to the metrics in this guide:
| Metric in this guide | Where BuddyNext’s documented analytics help |
|---|---|
| Active members | DAU, WAU and MAU |
| Retention | Cohort view by join week |
| Activation | Funnel from sign-up to first post |
| Response time, satisfaction, paid conversion | Not covered by those views as described, so use your activity data, a survey and your payment records |
The same Wbcom write-up says WB Gamification, a separate Wbcom plugin, adds a dashboard with points awarded, active members and badges earned over the last 30 days, plus a churn-risk panel that flags members whose engagement has dropped, with a risk score and days since last seen. It says that panel needs about a week of activity before it is useful (source).
You do not need any of this to start. A spreadsheet and a monthly export will do for the first few months, as FeverBee’s measuring guide suggests: keep a spreadsheet and graph of your metrics and update it monthly. Move to built-in dashboards when the spreadsheet becomes the chore you skip. If you run BuddyNext under Reign, these views sit beside the design you already control. To see the product, there is a live BuddyNext demo, and the product site is buddynext.com.
If you are comparing platforms by what they let you measure and own, see the best self-hosted community platforms. Owning the site means you also own the raw data, which matters for cohort and retention work.
How do you run a monthly community review in an hour?
Keep it to one page and one hour. Pull the same eight numbers each month, write one sentence about what changed and one action you will take. A short review you actually run beats an elaborate dashboard that nobody opens.
- Pull the numbers. Active members in the last 30 days, total registered members, new members this month, first-post rate for last month’s joiners, 30-day retention for last month’s cohort, median time to first reply, posts per 100 active members, and the share of posts from your top ten posters. Add paid conversion and cancellations if you charge.
- Compare against last month and three months ago. FeverBee suggests the same comparison for new visitors: against last month and against six months earlier. To confirm it worked, you should be able to say for each number whether it is up, down or flat.
- Read ten real threads. Numbers tell you where to look. Threads tell you what is happening. Pick five that did well and five that got no reply.
- Find the earliest broken step. If retention fell, check activation. If activation fell, check response time. Fix the earliest one.
- Check the front door. Open your sign-up and first-screen pages as a new member on a phone. If the Reign header layout or the welcome page is the obstacle, change that before you change the community.
- Write one action. A single change, with an owner and a date. Next month, check whether the number it was meant to move did move.
If the numbers are consistently weak, it is worth asking whether the question is metrics or purpose. Our piece on why a community matters when AI can copy your content is a useful read on what members are actually coming for. For programme-style communities, where the arc of a cohort is the product, see how to run a cohort programme on your own site.
What does the site itself have to do with the numbers?
More than people expect. A cluttered sign-up, a slow first page or a confusing member directory all show up as a poor activation rate, and no amount of community management repairs a design problem.
Reign is the WordPress community theme from Wbcom Designs, and it is the design layer for a site like this. It works with BuddyNext, BuddyPress, BuddyBoss Platform, PeepSo and Wbcom’s own plugins, with five header layouts and colour tokens, so you can adjust the first screens a new member sees without writing code. Changing a layout is one way to test an activation problem. If you are on an older stack and want to keep your look, see keeping your Reign site and upgrading the community. A searchable, well-designed directory is also an activation tool, as covered in our guide to member directory plugins.
Questions people ask
What is a good active-member ratio for a community?
There is no honest universal number, and we will not invent one. FeverBee’s own worked example shows why: a falling ratio can be a minor concern in a growing community and a serious warning in a stagnant one. Track your own ratio monthly, track the absolute number of active members next to it, and judge the trend.
What is the single most important community health metric?
Millington’s view is that no single metric captures community value. If you must pick one to watch most closely, FeverBee’s guidance points to active members in the last 30 days, because when that number falls you have a serious problem and limited time to act. Pair it with the first-post rate for new members so you can see whether the problem is welcome or retention.
How often should I review community metrics?
Monthly is the right rhythm for most communities. FeverBee suggests keeping a spreadsheet and graph of core metrics and updating it each month. Check response time and unanswered posts more often, weekly or daily, since those you can fix quickly.
Do I need analytics software to measure community health?
Not at the start. A spreadsheet and an export of member activity cover everything in the table above. Software helps when you have enough members that cohorts and funnels are tedious to build by hand. BuddyNext documents DAU, WAU, MAU, cohort and funnel views in its paid Insights area, so check the product page to see whether that fits your plan.
How do I measure community ROI?
Start from the goal the community serves, then measure that. For a paid community the goal is clear: conversion, renewal and cancellations. For a support community it is closer to resolved questions and repeat contacts, with the caution that deflection is hard to prove.
Where to go from here
Pick the six starting metrics, write down how you calculate each, and run your first monthly review this week. If you are building the site that will produce these numbers, Reign gives you the design layer, and BuddyNext is the community engine that sits under it.