Fan Lifetime Value: A Cohort and Payback Model

TL;DR. Fan lifetime value (LTV) is the total revenue an average fan generates across the whole time they stay active, the number that tells you what a fan is worth rather than how many you have. You build it by grouping fans into cohorts by the month they joined, tracking how each cohort's revenue accumulates and decays, and reading off the average per-fan total once the curve flattens. Set that LTV against what it costs to acquire a fan and you get a payback period, the single most useful figure for deciding how much to spend, how to price, and which fans to chase. Across a roster the model turns "we got a lot of subscribers this month" into "here is what they will be worth and what we were allowed to pay for them," which is the difference between growth you can bank and growth you are guessing at.

Ask an agency owner how a creator is doing and the answer is almost always a count: subscribers this month, new fans this week, followers gained. Counts feel like progress because they only go up when things go right, but they hide the question that decides whether the account is a business or a bonfire, which is what each fan is worth over their whole life on the page. Two creators with identical subscriber numbers can be worlds apart: one full of $5 tourists who churn in a month, the other full of fans who spend for a year. Subscriber count cannot tell them apart; fan lifetime value can. This post is a working operator's guide to building that number properly across a roster: why revenue per fan is the metric that matters in 2026, the building blocks every LTV model needs, how to group fans into cohorts, how to turn churn into a defensible lifetime estimate, how to compute payback, how to use the model to decide, and where it breaks.

Why ARPU per fan matters more than subscriber count in 2026

Start with the shift reframing how serious agencies measure success. OnlyFans is still enormous and still growing: the platform reported $7.22 billion in gross fan spend for its fiscal 2024, up 9 percent, with 4.63 million creator accounts and 377.5 million fan accounts, and it has paid over $25 billion to creators since 2016. But the same disclosure carried the other half of the story. That 9 percent was the slowest annual gain since the platform's breakout, a long way down from the 118 percent spike of 2021, and fan accounts grew 24 percent while gross spend grew 9, so the average fan spent less than the year before. The pie is still getting bigger, but it is being cut into more, thinner slices.

Industry commentary through 2026 reads that trend the obvious way: as raw fan growth outpaces spend growth, the metric that separates winners stops being how many fans you have and becomes how much each is worth. Treat the specific "ARPU has overtaken subscriber count" framing as commentary rather than a hard platform statistic, since it is a read on the trend, not a number OnlyFans publishes. But the logic is hard to argue with, and the platform's own direction of travel points the same way. When Architect Capital took a 16 percent stake in OnlyFans' parent for $535 million in May 2026, valuing the company at about $3.15 billion, the stated purpose was to build financial services and products for creators. A platform investing in creator banking is betting that the money already flowing through it is the asset, not the raw headcount of fans, and that is the same bet you should make at the agency level.

Subscriber count is an input to revenue, not a measure of it, and it flatters you exactly when you are most at risk, because a page can add fans and lose money at once if those fans cost more to acquire than they ever spend. Fan lifetime value closes that gap. Our breakdown of OnlyFans revenue-per-fan benchmarks for agencies sits alongside this piece; here we build the model that turns benchmarks into decisions.

The building blocks: cohorts, retention curve, and average revenue per fan

Before any spreadsheet, get the three ingredients straight, because most homemade LTV numbers are wrong for the same reason: they mix fans of wildly different ages into one average and call it lifetime value. Lifetime value has three parts.

The first is the cohort, a group of fans defined by when they joined. It is the unit that makes lifetime measurable, because to know what a fan is worth over their life you have to watch the same group age together. You never average across the whole page at once, because a page mixing fans acquired last week and last year tells you nothing about how a fan behaves over time.

The second is the retention curve, how much of a cohort is still active and spending as the months pass. A cohort decays on a curve, steep at the start as tourists and mis-fits drop off, then flattening as the fans who were always going to stick settle into a habit. The shape of that curve is the biggest single driver of lifetime value, because a fan who stays twelve months is worth many times one who stays one. Our guide to subscriber retention and rebill for agencies covers the levers that bend this curve; here it is the raw material the model runs on.

The third is average revenue per fan, usually shortened to ARPU: how much an active fan spends in a given period, counting subscription plus pay-per-view plus tips plus everything else, divided across the active fans. ARPU is where the money actually lives, because on most well-run pages the subscription is the smaller part and pay-per-view and tips are the larger, so a page that counts only subscription revenue undercounts a fan by a wide margin.

Put the three together and lifetime value is almost a sentence: the average revenue a fan spends per period, summed across all the periods they stay active. Everything below is just doing that sum carefully, cohort by cohort.

Grouping fans into cohorts by acquisition month and channel

The cohort is where the model either becomes trustworthy or stays a fantasy, so build it deliberately. The default cut is by acquisition month: every fan who first subscribed to a creator in a given calendar month belongs to that month's cohort, tracked as a block from the month they joined onward. January's cohort has its own month-one, month-two, and month-three revenue, each measured as the group ages, never mixed with fans who arrived later. That keeps the number honest: average across all current fans instead, and a month where you acquire a flood of new fans drags the average down, not because those fans are bad but because they are young. Cohorts keep each group's clock separate, so you always compare month-three behavior to month-three behavior.

Once acquisition-month cohorts are in place, the second cut is by channel, where the fan came from, and this is where the model earns its keep as a decision tool. A fan from a warm, high-intent source behaves nothing like one from a cold blast of cheap traffic: the warm fan stays longer and spends more, the cold fan often churns in the first month having barely spent, and a blended LTV averages the two into a middle number that describes neither. Splitting cohorts by channel, so you can see the twelve-month value of fans from your best source against your worst, is what tells you which traffic is worth paying for and which is quietly losing money even though it "brings subscribers."

A note on granularity, because it is easy to over-build this. Monthly cohorts are the right default: weekly cohorts fragment the data into groups too small to be stable, and quarterly cohorts blur the early churn that matters most. Split by channel at the level you actually make spending decisions, usually a handful of buckets, not fifty campaigns.

Turning churn-by-tier into a lifetime-value estimate you can defend

Now the part that turns a retention curve into a single number: estimating how long the average fan stays and what they are worth, without waiting a year for the data to complete itself.

The clean way to think about lifetime is through churn, the rate at which active fans drop off each month. If a cohort loses a steady share of its remaining fans every month, average lifetime is roughly one divided by that monthly churn rate: a cohort churning 20 percent a month averages about five months, one churning 10 percent averages about ten. This is a model, not a measurement; it assumes churn is roughly steady after the early drop-off, which is why you apply it to the flattened part of the curve rather than the steep first month. But it lets you estimate lifetime from a few months of data instead of waiting for every fan to leave.

The refinement that makes the estimate honest is to split by tier, because a single blended churn rate hides the most important fact about the page: the fans are not one population. On almost every account a small group of high spenders churns slower and spends far more, while the mass of tourists churns fast and spends little. Blend them and your LTV describes nobody, dominated by the many cheap fans yet inflated by the few rich ones. Run the churn-and-ARPU math separately for a top tier and a base tier, weight by how many fans are in each, and the number holds up. This is the strategic heart of the whale strategy for agencies: the top tier is where most lifetime value concentrates.

Lifetime value then falls out of the two pieces you now have: average fan lifetime in months, multiplied by average revenue per fan per month, done per tier and summed. Discipline whatever number you land on two ways: state it as a range, because it is an estimate built on a churn assumption; and refresh it as real cohorts mature, so the modeled lifetime gets checked against fans who actually stayed or left.

Computing payback: acquisition cost against fan LTV

Lifetime value on its own is interesting. Lifetime value set against what a fan costs to acquire is decisive, because that ratio is what tells you whether the growth you are buying makes money, and how fast.

Start with the cost side, which agencies routinely underestimate. The cost to acquire a fan is everything you spent to get that fan divided by the number of fans it produced: for paid traffic, spend divided by new fans; for the promo and shoutout economy, placement cost divided by the fans it drove; for organic and in-house effort, fuzzier but still real, the share of a team member's paid time spent capturing fans. You need it in the right ballpark, because a fan acquisition cost you pretend is zero is how agencies convince themselves that unprofitable traffic is free.

With cost and LTV in hand, two numbers matter. The LTV-to-cost ratio is lifetime value divided by acquisition cost: a ratio below one means you lose money on every fan no matter how good the subscriber count looks, while a healthy ratio means each fan returns a comfortable multiple of what you paid. The more operationally useful figure is the payback period: how many months of a fan's spending it takes to earn back what you paid to acquire them. A fan who costs one month's spend to acquire pays back in month one and is pure upside after; a fan who takes eight months to pay back is one you are financing for most of a year, which is fine if your cash can carry it and dangerous if it cannot.

Payback is the figure to lead with across a roster because it speaks to cash, not just profit: an account can have a beautiful long-run LTV and still sink the agency if the payback period is longer than your runway. Our guide to the creator payback period and agency ROI works the same math one level up, at the creator rather than the fan, and the two nest neatly: fan-level payback tells you whether the traffic into an account pays, creator-level payback tells you whether the account itself pays, and both have to clear for the roster to be sound. Reckon in the platform's 20 percent fee throughout, because a fan's LTV to you is the after-fee figure, not the sticker spend.

Using the model to decide spend, pricing, and which fans to chase

A model that only produces a number is a trophy; one that changes what you do is a tool. Once you have per-cohort, per-channel, per-tier LTV and payback, three decisions get sharper.

How much you can spend to acquire a fan. This is the biggest unlock. Once you know a fan's after-fee lifetime value, you know the ceiling on what you can pay to acquire one and still make money, and the payback period at any given cost. That turns traffic buying from a gut call into a rule: spend up to where payback stays inside your cash comfort and the LTV-to-cost ratio stays healthy, and stop cold on any channel where after-fee LTV does not clear the cost. The channel-split cohorts make this real, so you fund the sources that produce fans worth more than they cost and cut the ones that produce cheap subscribers who never pay back.

How you price and structure the page. LTV is downstream of ARPU and retention, so the model tells you which lever to pull. If lifetime value is low because fans churn fast, the fix is retention and the welcome-to-rebill flow, not more traffic. If it is low because active fans barely spend beyond the subscription, the fix is the pay-per-view and upsell structure that lifts ARPU. If the top tier is thin, the fix is whale cultivation. Pricing changes stop being guesses when you watch their effect on later cohorts.

Which fans to chase, and which to let go. The model almost always shows that a small group of fans carries a large share of lifetime value while the marginal cheap fan is worth very little. That reshapes where effort goes: disproportionate attention to identifying and keeping high-value fans early, and a realistic, low-cost posture toward the mass of tourists. Chasing every fan equally is how you spend your best hours on your worst customers. Feeding these signals into your agency KPI and metrics dashboard turns a one-off analysis into a standing view.

Where the model breaks and how to keep the inputs honest

Every model is a set of assumptions wearing a number, and knowing where this one breaks keeps it from lying to you.

The most common failure is immature cohorts read as complete. A cohort three months old has not lived its full life, so any lifetime figure you extract from it is a projection leaning hard on your churn assumption, not a settled fact. Never make a large, irreversible spending decision on a number pulled from cohorts only a few months into a life you claim runs much longer. The fix is labeling: mark which numbers are modeled and which are observed, and let the observed ones correct the models as cohorts age.

The second is the whale distortion. Because a few fans carry so much of the value, an LTV average is fragile to a handful of big spenders; one creator's number can swing on whether two or three whales stayed or left this quarter, making the blended average jumpy and, on small pages, close to meaningless. The defenses are the tier split, which quarantines the volatile top from the stable base, and stating LTV as a range.

The third is dirty inputs. Garbage in produces confident garbage out, and the usual culprits are counting only subscription revenue while ignoring pay-per-view and tips, which understates ARPU; failing to reckon the platform's 20 percent fee, which overstates what a fan is worth to you; guessing acquisition cost or pretending it is zero, which flatters payback; and mixing test-period promo pricing into cohorts as if it were normal. Write your assumptions next to the number, the churn rate, the fee, the acquisition cost, so that when the output looks wrong you can find which input to fix instead of distrusting the whole thing.

The last caution is scope. Lifetime value is a planning instrument, not a promise, and it is most powerful used comparatively: which channel, which cohort, which creator, which pricing change produced better fan value. Use it to rank and set ceilings, not to forecast a creator's earnings to two decimal places. Standing up this model across a full roster, keeping the inputs clean and the cohorts current, is exactly the kind of behind-the-scenes analytics a white-label partner runs so owners can act on the output without building the machine. WhaleFinders operates as the marketing and operations arm for OnlyFans agencies, and modeling fan value across a fleet is part of that remit. If it is a lever you would rather delegate than build, the conversation starts on Telegram at t.me/whalefindersupport.

Frequently asked questions

What is fan lifetime value on OnlyFans, in plain terms?

It is the total revenue an average fan generates across the whole time they stay active on a creator's page, counting subscription, pay-per-view, and tips, measured after the platform's 20 percent fee. Where subscriber count only tells you how many fans you have right now, lifetime value tells you what each is worth over their whole life on the page, which is what lets you decide how much you can afford to spend to acquire one and still make money.

Why use cohorts instead of just averaging the whole page?

Because lifetime value is about how fans behave as they age, and you cannot see aging if you blend fans of different ages into one average. A page-wide average gets dragged down every time you add a batch of new fans, not because they are bad but because they are young. Grouping fans by the month they joined keeps each group's clock separate, so you compare month-three behavior to month-three behavior and the trend you read is real.

How do I estimate fan lifetime without waiting a whole year for the data?

Use churn. If a cohort loses a roughly steady share of its remaining fans each month after the early drop-off, average fan lifetime is approximately one divided by that monthly churn rate, so a cohort churning 10 percent a month averages about ten months. Apply it to the flattened part of the curve, split it by spending tier so whales and tourists are not blended, state the result as a range, and refresh it as real cohorts mature.

What is a fan payback period and why does it matter more than LTV alone?

Payback period is how many months of a fan's spending it takes to earn back what you paid to acquire them. It matters more than raw lifetime value because it speaks to cash, not just profit: an account can have a strong long-run LTV and still sink you if the payback runs longer than your runway. Leading with payback keeps you solvent while you grow, not just profitable on paper eventually.

How does the 20 percent platform fee change the model?

It changes what a fan is worth to you. Fans spend gross, but OnlyFans keeps 20 percent of every dollar before anything reaches the creator, so the lifetime value you build decisions on is the after-fee figure, not the sticker spend. Model LTV on gross spend and compare it to real acquisition costs and you overstate what every fan is worth by a fifth. Always deduct the fee before you set acquisition ceilings.

Does this model work for a creator with only a few hundred fans?

Yes, but read it more cautiously. On a small page a handful of high spenders can swing the average LTV from quarter to quarter, so the blended number is jumpy and closer to directional than precise. Lean on the tier split, state lifetime value as a range, and use the model comparatively rather than trusting an exact figure.