Membership Metrics: What to Track When Revenue Repeats (MRR, Churn, LTV)

Published July 26, 2026

Part of: Digital Products — our full guide on this topic.

Disclosure: Some links below are affiliate links. If you sign up through them we may earn a commission at no extra cost to you. We only recommend tools we'd genuinely suggest to a friend. See our full disclosure.

There is a guide on this site about which business metrics to track, and its central idea is a good one: map your numbers to your funnel. Traffic, signups, engagement, sales. Each stage asks one question, and where the number drops is where you are losing people.

That model works right up until your revenue repeats.

A funnel is a one-way path that ends at the sale. For a one-off product, that is the whole story — the sale happens, the money arrives, and the next sale is a fresh trip through the same funnel. For a membership, a subscription product or a recurring commission, the sale is where the measuring starts. Most of this month’s money came from people who bought months ago and simply have not left. No stage of a funnel measures that, because a funnel has no concept of revenue you already had.

This guide covers the numbers that do.

Honest disclosure: some links below are affiliate links. If you sign up through one I may earn a commission at no extra cost to you. Everything here is my genuine assessment.

The quick version

• Track what you kept, not just what you sold. Last month’s members are the biggest input to this month’s revenue.
• A flat revenue line is two completely different businesses. Break the month into movements or you cannot tell which one you are in.
• Growth here is a subtraction. New minus lost. No funnel metric has a minus sign in it.
• Averages lie; cohorts don’t. “Of everyone who joined in March, how many are still here?” is the only version of retention that lets you compare before and after a change.
• LTV is useful and over-claimed. Know what its formula assumes before you plan around it.
• Annual plans break every number above unless you deliberately separate cash from revenue.
• Never borrow a churn benchmark. Your own trend is the only comparison that means anything.

Start with the number the funnel cannot show you

MRR — monthly recurring revenue — is the total your currently active subscribers pay you in a normal month. Twelve people on a $9 plan is $108 MRR. That is it. There is no clever part.

It is worth having a name for it because it answers a question a sales figure cannot: how much of next month do I already have? A one-off business starts every month at zero and re-earns the whole thing. A subscription business starts at last month’s MRR minus whoever leaves, and everything you do is either adding to that base or defending it.

Two mistakes are worth heading off immediately, because both quietly corrupt every number downstream:

• Do not put one-off sales in it. If you sold three copies of an ebook this month, that is real money and it does not belong in MRR, because MRR is specifically a measure of what repeats. Track one-off income as its own line. Mixing them produces a number that looks like a subscription base and behaves like a lottery.
• Do not confuse MRR with cash. Money arrives in lumps, on payout delays, net of fees, with annual plans landing all at once. MRR is the smooth underlying figure; your bank balance is what actually happened. You need both, and you need to know which one you are looking at.

A flat revenue line is two completely different businesses

Here is the specific failure that makes a funnel-shaped dashboard dangerous for a subscription.

Your revenue is $600 this month. It was $600 last month. On the chart, a flat line.

That flat line is either:

• Nobody joined and nobody left. Your product is holding. You have a traffic problem — a real one, but the calmest kind, because everything you already built is still working.
• Twenty people joined and twenty people left. Your marketing is working hard, and something after the sale is undoing all of it. You are on a treadmill, paying (in time, content and attention) to stand still, and the moment your promotion slows the whole thing falls.

Same chart. Same total. Opposite futures, opposite fixes. If you only track revenue, you cannot distinguish them, and the second one will feel fine for months before it doesn’t.

The fix is to stop measuring the level and start measuring the movement.

The four movements

Any change in your MRR from one month to the next is made of at most four things:

1. New — MRR from members who joined this month.
2. Expansion — existing members paying you more: an upgrade to a higher tier, a switch to a bigger plan.
3. Contraction — existing members paying you less: a downgrade.
4. Churned — MRR from members who left or whose payment stopped.

And the arithmetic that a funnel dashboard has no equivalent of:

Net change = New + Expansion − Contraction − Churned

That minus sign is the entire difference between the two kinds of business. In a one-off business, a good month is additive: you sold, you earned, and nothing that happened this month can take away what you earned last month. In a subscription, you can have your best acquisition month ever and go backwards.

If you run one flat-priced tier — which most solo memberships do, and sensibly — then expansion and contraction are zero and you only need two columns: what came in, what went out. Do not manufacture tiers just to have more metrics. Add the middle two the day you actually have something to upgrade to.

These four movements are also the only instrument you have for reading whether a change you made was a good one — which is the practical reason for making changes one at a time and letting a full billing cycle run before making the next. Move three things in one month and the churn figure tells you something happened without telling you what, and a number you cannot attribute is a number you cannot act on.

Once you can see the movements, the diagnosis in handling cancellations and failed payments becomes usable, because the churned figure splits again — into people who decided to leave and people whose card simply stopped working. Those are different problems and only one of them is about your product.

Why the average member lies, and what to use instead

The obvious way to measure loyalty is to average how long your members stay. It is also the way that produces the most confidently wrong conclusions.

An average across all your members mixes someone who joined last week with someone who has been there two years, and it is dragged around by whichever month you happened to grow fastest. Run a successful launch and your average tenure falls, because you just added a pile of brand-new members — which looks exactly like your retention getting worse at the moment it got better.

The honest question is a cohort question: of everyone who joined in March, how many were still here in month two? Month three? Month six? Then the same for April’s joiners, and May’s.

That is a handful of rows in a spreadsheet, and it does something no other metric on this page can do: it lets you tell whether a change you made actually worked. If you rewrote your onboarding in April, the April cohort’s month-two number against March’s is your answer. A blended, whole-business churn rate can never tell you that, because it mixes the people who experienced the change with the people who didn’t.

This is a different job from asking when people leave — that question diagnoses a single exit, and the cancellations guide works through what an early exit means versus a late one. A cohort table compares versions of your business against each other over time. It is the closest a solo operator gets to a controlled experiment.

Two practical notes. Cohorts only become readable at a certain size — with four joiners in a month, one person leaving is a 25% swing and means nothing. If your numbers are small, group by quarter instead of by month and be patient. And pick one definition of “still here” (paid this month) and never change it, because a metric redefined halfway through is worse than no metric.

LTV: genuinely useful, routinely over-claimed

Lifetime value is what a member is worth to you across their whole stay. The standard shortcut:

LTV ≈ average revenue per member per month ÷ monthly churn rate

At $9 a month and 5% monthly churn: $9 ÷ 0.05 = $180. The division works because a constant 5% monthly loss implies an average stay of 1 ÷ 0.05 = 20 months.

It is a useful planning number — mainly for answering “how much time or money can I justify spending to get one member?” — and it is the number most often quoted as though it were a fact. Four honest caveats:

• It assumes churn is constant, and it usually is not. In most memberships the first month or two churns far harder than later months: people who were never a fit leave quickly, and the ones who stay past that tend to keep staying. A single blended rate flattens that curve and gets both ends wrong.
• You need a reliable churn rate to use it, and you do not have one at three months old. Computing LTV from your first quarter mostly measures your launch, not your business.
• It is revenue, not profit. Platform fees, payment processing and anything you pay to run the thing come out of that $180 before it is yours. Platform fees compared covers what the common ones actually take.
• It is an average, and averages hide the shape. A membership where most people leave in month two and a few stay for years can produce the same LTV as one where everybody stays ten months, and those are not the same business to run.

Use it to size decisions, not to justify them. And if the number you get feels implausibly high, check whether your churn rate is quietly a launch artefact.

Annual plans break all of this (and how to keep it honest)

Annual billing is one of the better levers a small membership has — the membership pricing guide makes the case for it — and it distorts every metric on this page unless you handle it deliberately.

A $90 annual plan is not a $90 month. For MRR it counts as $7.50 a month, because that is what it is: twelve months bought at once. Count it at face value and you will record a spectacular month, then eleven quiet ones, and conclude something has gone wrong when nothing has.

Two specific traps:

• Cash you have been paid is not yet revenue you have earned. The money is in your account in January; the obligation to deliver runs to December. Sell a batch of annual plans, treat the whole lump as this month’s income, and you can spend a year’s revenue in a quarter while still owing eleven months of work.
• Annual churn is invisible for eleven months. A monthly member who is unhappy shows up in your numbers in weeks. An annual member who stopped opening your emails in March shows up as a non-renewal in December, by which time the cause is nine months cold. If you sell annual plans, watch engagement, not just billing — a member who has not logged in for two months has effectively already left; the invoice just hasn’t noticed yet.

There is a third trap that only appears once, at the end. An annual plan is a promise stretching up to twelve months into the future, which means it is also the thing that decides how long a wind-down takes if you ever stop: closing a membership explains why the closing date falls out of your longest outstanding term, and why one annual member who renewed last week is the difference between a month’s notice and a year’s.

To be clear about scope: this is about reading your own dashboard, not about how to record any of it in your accounts. How prepaid income should be treated in your books depends on where you are and how you file. Basic bookkeeping for a solo business covers the practical habits, and an accountant covers the rest — nothing here is accounting or tax advice.

Judge your traffic by retention, not conversion

This one changes what you do with the rest of the site.

Every traffic and conversion guide — driving traffic, opt-in pages, improving your conversion rate — implicitly ranks channels by how well they convert. For a one-off product that is the right ranking, because the sale is the end of the story.

For a subscription it isn’t. A channel that converts at 4% and whose members are gone by month two is worth less than a channel converting at 1% whose members stay a year. The high-converting channel wins on the dashboard the funnel gives you and loses badly on the money.

So record where each member came from, and read your churn by source. It is one extra column, usually a signup question or whatever your checkout already captures. What it typically reveals is that the audiences that took longest to persuade are the ones that stay — which is an argument for the slow channels, email and search, over whichever burst of attention converted best last month.

The smallest version that actually works

You do not need analytics software for any of this. One spreadsheet, one row per month, five columns:

(Illustrative rows on a $9 plan — March lost 3 of 40, a 7.5% churn rate; April lost 5 of 46, 10.9%. Two good-looking months of growth, and the number that matters got worse.)

Fill it in once a month, on a fixed date, from your billing dashboard. That is a fifteen-minute job and it gives you everything above: the movements, a churn rate (left ÷ members at start), the trend, and — after a few months of keeping a second tab with one row per joining cohort — a retention curve.

Then use the recurring revenue projector for the part measurement cannot give you: the forward view. Feed it your real churn number rather than a hopeful one and it will show you the ceiling your current pace settles at — new signups ÷ churn rate — which is the single most clarifying number in a subscription business. At 5% monthly churn, about 46% of any group of joiners is gone within a year, and fifteen signups a month settles at 300 members and stops climbing. Seeing where your own pace flattens out tends to reorder your priorities faster than any advice does.

What not to track

• Someone else’s churn benchmark. It varies so widely by price, audience and product type that it carries almost no information about you. Your own trend, measured the same way each month, is the only comparison worth making.
• Total member count as a headline. It goes up when you add cheap members and down when you prune, and neither movement means what it looks like. MRR and retention already tell you the story.
• Anything daily. These are monthly quantities made of small events; a day’s worth is noise. The one exception is failed payments, which decay — a card you chase this week is often recoverable and the same card next month usually isn’t.
• Metrics you will not act on. The test from the original metrics guide applies unchanged here: if a change in the number would not change what you do, it is decoration.

Where this fits

These numbers only exist if your platform can tell you them, which is worth checking before you build. What you need is a billing system that distinguishes a cancellation from a failed charge, tells you when each member joined, and lets you export the list — and that is easier when the checkout, the membership area and the email list know about each other rather than being stitched together.

Systeme.io is the one I point people to for a small membership for that reason: recurring billing, the member area and the email list sit on one plan, with a free tier you can run a real membership on while you find out whether it works at all. (That is an affiliate link — if you upgrade to a paid plan through it I may earn a commission, at no extra cost to you; see the full affiliate disclosure.) Features and free tiers change, so confirm the current details on the provider’s own site. If you have not built the thing yet, the free membership site guide covers the setup, pricing a membership covers the number you charge, and launching one covers getting the first members in.

The bottom line

A funnel dashboard answers did I make a sale? A subscription needs to answer what did I keep? — and no stage of a funnel measures the members who did nothing this month except not leave, which is where most of your money comes from.

So track four things and ignore the rest. MRR, so you know what next month starts at. The movements behind it, so a flat line resolves into either a stable business with a traffic problem or a treadmill with a retention one. Cohort retention, because it is the only version of the number that can tell you whether something you changed worked. And LTV, held loosely, as a sizing tool rather than a fact.

Do it in a spreadsheet, once a month, on a fixed date. Keep cash and revenue in separate columns so annual plans cannot flatter you. Read churn by source before you decide which traffic channel deserves next quarter. And resist every published benchmark you meet, because the only meaningful comparison in a recurring business is you against yourself, three months ago.