Dating app unit economics, the operator's guide

Most dating apps do not die from a bad product. They die from one math problem: they spend more to acquire a paying subscriber than that subscriber will ever return. The product can be polished and the reviews can be warm, and the company still quietly runs out of cash. The honest way to know whether your model works is to run the numbers, not the narrative.

This guide walks through the unit economics of a dating business the way an operator actually uses them. It defines the six numbers that decide whether you can afford to grow, shows where dating breaks the normal software playbook, and gives you a calculator to pressure-test your own model in a couple of minutes.

What unit economics actually means in dating

Unit economics is the per-user version of your profit and loss. Instead of asking whether the whole company makes money, it asks whether one more paying user makes money, and how quickly. Get that right and growth compounds in your favor. Get it wrong and every dollar of marketing spend digs the hole deeper.

Dating has its own version of this math, and it is harsher than most categories. Three things make it harder. Paying lifetimes are short, because a product that works removes the reason to keep paying. Monetization is concentrated, because a small slice of users carries the revenue. And acquisition is lopsided, because you are rarely buying a user, you are buying one side of a marketplace that only works when both sides are in balance.

The only six numbers that matter

You can drown in metrics. For deciding whether to spend the next dollar, six numbers do almost all the work.

ARPU, and why it can mislead you

ARPU is average revenue per user, taken across everyone who uses the app, including the large majority who never pay. It is a useful number for company-level forecasting and for comparing scale between products. It is a dangerous number for setting acquisition budgets, because it blends payers and non-payers into a single average that describes no real person. In dating, where most installs never spend a cent, ARPU is small and easy to misread.

ARPPU, the number the business runs on

ARPPU is average revenue per paying user. This is the number that actually drives a dating business, because dating revenue comes from the minority who subscribe, buy boosts, or spend on credits. When you size the value of acquiring a customer, you size it on ARPPU, not ARPU. Confusing the two is the single most expensive mistake in this category, and we come back to it below with real money attached.

Average paying lifetime

Lifetime is how long a paying user keeps paying before they cancel. The quick estimate is one divided by your monthly churn rate. At 20 percent monthly churn, the average paying lifetime is five months. At 10 percent it is ten months. Small changes in churn swing lifetime hard, which is why churn deserves more attention than almost any other lever in the model.

LTV

LTV, or customer lifetime value, is the total gross revenue you expect from a paying user before they leave. At its simplest it is ARPPU multiplied by average paying lifetime. It is gross, which means it is a story about revenue, not profit. Before you celebrate an LTV number, remember that refunds, chargebacks, app store commissions, and payment fees all sit between that figure and your bank account.

CAC payback

CAC is the blended cost to acquire one paying subscriber. CAC payback is how many months of ARPPU it takes to earn that cost back. This is the most underrated number in dating, because it governs cash, and cash is what kills companies. A business with a healthy long-term LTV to CAC ratio can still fail if it takes nine months to recover CAC and it only has six months of runway.

LTV to CAC ratio

This is the headline. Divide LTV by CAC and you get the multiple of value you create for every dollar of acquisition. In paid dating, the working floor is around three to one on a gross basis. Below that, spend stops being efficient once the real deductions land. Far above it, say beyond five to one, you usually have unused growth capacity and are underspending.

Run your numbers

The calculator below starts with sensible dating defaults and updates as you type. Change the inputs to match your own app and watch the outputs move. Treat the result as a pressure test, not a forecast.

Why dating breaks the standard software playbook

Most software financial models assume monthly churn somewhere between three and five percent. Plug that into the lifetime math and a paying user sticks around for two or three years. That assumption is why so much general SaaS advice does not transfer to dating.

In paid dating, monthly churn commonly sits between 15 and 25 percent on a one-month plan. The reasons are structural. The product is designed to end its own usefulness, because a user who meets someone has every reason to cancel. Intent is spiky, so people subscribe for a hard month of looking and then stop. And the category is crowded, so users keep several apps installed and move between them.

That single churn difference compresses LTV by an order of magnitude compared with a typical SaaS model. If your projections show a heroic LTV to CAC ratio, the first place to look for the error is the churn assumption. Honest churn is the fastest way to turn a fantasy model into a real one.

The ARPU and ARPPU trap, with real money attached

Here is the mistake in numbers. Imagine an app with a 19.99 dollar monthly plan and a 3 percent free-to-paid conversion rate. ARPU is roughly 60 cents, because almost everyone is free. ARPPU is 19.99, because that is what a payer pays.

Now imagine you set acquisition budgets off ARPU. You look at 60 cents, decide users are nearly worthless, and either underspend into stagnation or, worse, you compute LTV on a blended figure that quietly folds in non-payers and overstates how many paying months you are really buying. Operators who plan on ARPPU, and who track conversion as its own separate lever, make far cleaner decisions. They know what a payer is worth, they know what share of installs become payers, and they keep the two numbers apart.

The dating-specific CAC problem: you are buying a ratio, not a user

In most categories an acquired user stands on their own. In dating, a user is only valuable if there are enough of the people they want to meet. Acquire a wave of men into a market with too few women, or the reverse, and your new users churn fast because the core experience is empty. Your nominal CAC looks fine and your real CAC, the cost to acquire a user who actually stays, is far higher.

This is why blended CAC in dating hides so much. The honest version segments by the side of the marketplace and by geography, because the cost and value of each side differ sharply, and because liquidity is local. A balanced book in one city tells you nothing about an unbalanced one next door. When you read your CAC, ask which side you bought and whether the other side was there to meet them.

Gross is a story, net is the truth

Every number above is gross until you subtract the deductions, and in dating the deductions are large. App store commissions take a meaningful cut on mobile subscriptions. Payment processing carries fees. Refunds happen. And disputes, where a customer asks their bank to reverse a charge, are both a direct loss and a threat to the whole business.

Card networks treat dating as a high-risk category and watch dispute rates closely. Sustained rates above roughly one percent can put a merchant account at risk, which is an existential problem, not a line item. Tight operations keep disputes well below that. As a reference point, on the platform High Intent runs, disputes currently sit under 0.3 percent, and the discipline that gets you there, clear billing descriptors, honest paywalls, fast cancellation, and active fraud screening, is its own body of work. The point for your model is simple: take your gross LTV, then haircut it for fees, refunds, and disputes before you compare it to CAC. The honest LTV to CAC floor of three to one is a net floor, not a gross one.

Subscriptions versus credits: same revenue, different economics

Two apps can earn the same ARPPU and have completely different unit economics depending on how they package the money. A subscription bills on a cycle and gives you predictable, recurring revenue, but it also hands the user a recurring decision to cancel, which is where dating churn bites. An a la carte or credits model, where users buy boosts, super likes, or message packs, converts intent into revenue in the moment and is far less exposed to monthly cancellation, but it is lumpier and depends on frequent high-intent sessions.

Neither is better in the abstract. What matters for your numbers is that they change the shape of LTV. Subscription LTV is churn-driven and smooth. Credit LTV is frequency-driven and spiky. If you run a hybrid, model the two streams separately and then combine them, rather than blending everything into one average that hides which engine is actually working.

A worked example: does this app deserve more spend?

Take a one-month plan at 24.99 dollars, a 4 percent free-to-paid conversion, 20 percent monthly churn, and a blended CAC of 35 dollars per paying user.

Average paying lifetime is one divided by 0.20, or five months. Gross LTV is 24.99 multiplied by five, about 125 dollars. LTV to CAC is roughly 125 over 35, about 3.6 to one on a gross basis. CAC payback is 35 divided by 24.99, about 1.4 months.

On the surface this is a fundable model: payback well under three months and a ratio above the three to one floor. But apply the haircut. Take 30 percent off LTV for store commissions, fees, refunds, and disputes, and gross LTV of 125 becomes a net 87. Net LTV to CAC is now about 2.5 to one, under the floor. The same app that looked healthy gross looks marginal net. This is exactly the gap that sinks operators who never run the second calculation.

Operator benchmarks for dating unit economics

Use these as orientation, not gospel. They are operator estimates drawn from how paid dating books tend to behave, and your category, geography, and plan mix will move them.

• Monthly churn on a one-month plan: roughly 15 to 25 percent.
• Free-to-paid conversion: roughly 2 to 6 percent, higher in high-intent and niche products.
• CAC payback target: under three months, ideally under two.
• LTV to CAC floor: about three to one net, with healthy books running four to one and up.
• Dispute rate ceiling before account risk: keep it well under one percent.

If your model sits far outside these, that is not necessarily wrong, but it is a flag to check your assumptions before you pour money into growth.

How to move each number

Churn is usually the highest-leverage lever, because lifetime is so sensitive to it. Better onboarding, faster time to a first quality match, and win-back flows for lapsed payers all extend lifetime. Conversion responds to paywall placement and to packaging that matches intent, not to raising the price alone. CAC responds to creative, to channel mix, and most of all to marketplace balance, because acquiring the scarce side cheaply is worth more than another wave of the abundant side. And payback responds to plan mix, since annual and quarterly plans pull cash forward even when underlying retention behavior is similar.

The conversion lever most operators leave alone

Conversion from free to paid is where many dating apps leave the most money, because it is easier to obsess over CAC than to fix the paywall. The levers are specific. Where you place the paywall matters more than its price: gating the moment of highest intent, such as seeing who liked you or sending the message that matters, converts better than a generic upsell. Packaging matters, because a single confusing plan converts worse than a clear good, better, best ladder. And timing matters, because intent in dating spikes early in a session and early in a user's life on the app, then fades. A one-point move in conversion changes ARPU and the number of paying months you can buy for the same CAC, which is why conversion belongs next to churn at the top of your list, not buried in a growth backlog.

Read cohorts, not averages

Every number in this guide is easier to fool yourself with when you read it as a single blended average. A blended LTV mixes your best month with your worst. A blended CAC mixes a cheap organic week with an expensive paid push. The operators who get this right read by cohort: they group users by the month they joined and follow each group's retention and revenue over time. Cohorts show you whether retention is improving or quietly decaying, whether a pricing change actually held, and whether last quarter's growth was real or just pulled forward. If you only ever look at the company-wide average, you will miss the moment the model turns, because the average lags the cohort by months. Build the cohort view once and check it every month.

What a buyer or investor will test first

If you ever raise or sell, these are the numbers that get tested before anyone reads your vision deck. A buyer will rebuild your LTV from raw payment data, not from your slide. They will check whether churn is calculated on payers or on all users, whether CAC includes the incentives and the attribution you quietly dropped, and whether your dispute and refund rates sit inside what processors tolerate. They will look hard for the gap between gross and net that we walked through above, because that gap is where optimistic models hide. The cleaner and more honest your unit economics, the faster a process moves and the less the price gets chipped away in diligence. Founders who can show real cohorts, segmented CAC, and net LTV walk into those rooms with leverage. Founders who cannot spend the meeting explaining away the holes.

Where this connects

If the math says your model works but you do not have the team to run acquisition, retention, and payments at this level, that is exactly the work High Intent Services takes on, run by operators who have done it at scale. If you are launching and want the payments, fraud, and liquidity handled from day one, that is what the platform is for. And if you are weighing a raise or a sale, these are the numbers a buyer or investor will test first, so it pays to have them honest before the conversation starts.

Key takeaways

• Track ARPPU separately from ARPU. They describe different things, and acquisition budgets belong on ARPPU.
• Monthly churn above 20 percent makes any long lifetime claim suspect. Check churn before you trust an LTV number.
• CAC payback under three months is the working operator benchmark, because cash, not ratios, is what runs out.
• The honest LTV to CAC floor in dating is around three to one net, after refunds, fees, and disputes, not gross.
• In dating you buy a marketplace ratio, not a user. Segment CAC by side and by geography or it will mislead you.

Related reading

Pair this with our glossary entries on ARPU, ARPPU, LTV, CAC, CAC payback, and churn, and use the calculator above whenever you revisit your model.