You Pay for the Catalogue on Signature. The First Ad Payout Clears 60 Days Later.

A licensed games portal pays its catalogue fee on signature and waits 31 to 61 days for its first ad payout. Model that working-capital gap before you launch.

Portal business cases get built as monthly P&Ls. Licence cost here, hosting there, revenue underneath, a margin line at the bottom. The margin line is usually fine. The business still runs out of money in month three, because a P&L records the month a play happened and a bank account records the month somebody actually sent you the funds, and in this industry those are never the same month.

The gap is not a rounding error and it is not unknowable. Every major revenue partner publishes its payout mechanics. Almost nobody reads them before signing the catalogue invoice.

💸 The Two Invoices That Never Share a Timeline

On the cost side, a licensing deal is close to instantaneous. You sign, you pay, you get builds. Perpetual licences are typically due in full before or on delivery. Annual licences are usually billed a year at a time up front. Either way the cash leaves in week one, alongside hosting setup, a domain, whatever integration work the portal needs, and the first month of anyone you are paying to run it.

On the revenue side, nothing is instantaneous. An impression served on 1 June is not money on 1 June. It is a number in a dashboard that becomes a finalised figure in early July, becomes a payment instruction in late July, and becomes cleared funds in your bank in early August. The play happened in Q2. The cash is a Q3 event.

Run those two facts against each other over twelve months and the shape of the business changes. Not the profitability — the survivability.

📅 "Monthly Payouts" Is Not the Same Thing as Monthly Cash

Every ad network and store says it pays monthly. All of them are telling the truth and all of them mean something different by it. Here is what the published terms actually say, as of August 2026.

Google AdSense and AdMob

Google's published payment timeline runs like this: earnings for a calendar month are finalised on your Payments page by roughly the 3rd of the following month. Your balance is checked against the payment threshold — US$100 by default — on the 20th. If you clear it, payment is issued between the 21st and the 26th. Electronic funds transfer then takes up to seven business days to land.

Do the arithmetic on a single month. An impression served on 1 June is paid around the start of August: call it 61 days. An impression served on 30 June is paid at the same time: 32 days. Across a full month of traffic, the average dollar you earn waits about a month and a half to become spendable.

The threshold matters more than it looks. It is not a delay of a few days — it is a hard gate. Finish June on US$80 and you are not paid in July. You are paid whenever your cumulative balance crosses US$100 before a 20th, and the whole balance travels with it. A portal doing modest early numbers can easily find its genuine first payout arriving in month four or five of trading, not month two.

Web game platforms

Distribution platforms tend to write longer terms than they actually run. CrazyGames' developer documentation states a minimum of €100 in earnings before a payout is issued, and describes the platform as operating on NET 60 terms while currently aiming to pay sooner — typically by the 10th of the following month.

Read that as an operator, not as a developer. The contractual right is 60 days. The current practice is faster. Practice is not a covenant, and a model built on the practice has no headroom if the practice reverts to the contract. Build the plan on NET 60 and treat early payment as upside.

App stores, if you also ship an APK

Google Play's payout documentation puts payouts around the 15th of the month following the orders, with a minimum earned balance of US$1 for local-currency payouts and US$100 for USD wire transfers, and no deposits initiated on weekends or banking holidays. That is the fastest large payer in this whole stack: a sale on the 30th waits about fifteen days.

Apple is slower and stranger. It pays 33 days after the close of a fiscal month, and Apple's fiscal months do not line up with calendar months. RevenueCat's 2026 Apple fiscal calendar gives a concrete example: sales in the fiscal month running 1–28 February 2026 are paid on 2 April 2026. A sale on 1 February waits 60 days. The same catalogue on two stores produces two materially different cash curves, which is a reason to know which store your revenue actually skews to before you plan around it.

🧮 The Cash Curve, Written Out

Call one month of net revenue R. Call the licence and setup cost L. Call monthly running cost — hosting, tooling, anyone on payroll, any paid traffic — C. You do not need to know what R is to see the shape.

• Month 0: you are out L, plus C. Nothing has been earned.
• Month 1: traffic starts. You earn some fraction of R. You receive nothing. You are out another C.
• Month 2: you earn something closer to R. You receive the month 1 money, if it cleared the threshold. You are out another C.
• Month 3 onward: receipts stabilise roughly one full month behind earnings, permanently.

The trough is not L. The trough is L plus about two months of C, and deeper than that if your early months sit under the payout threshold. That is the number that needs to exist in the bank on day one, and it is routinely two to three times what the founder budgeted, because the founder budgeted the catalogue.

There is a second-order effect that catches people who are doing well. The lag is proportional to revenue, so growth widens the gap. A portal growing 20% month on month is permanently financing a bigger receivable than it was last month. Slowing down improves your cash position. That is a deeply unintuitive property of the business and it is the reason profitable portals stall — every extra unit of growth has to be funded out of pocket for a month or two before it pays for itself.

🔗 What You Are Owed Is Not What You Are Guaranteed

Programmatic advertising runs on sequential liability: as CafeMedia's publisher explainer puts it, if an advertiser does not pay the agency, the agency does not have to pay the demand-side platform, and the shortfall travels down the chain to the publisher who already ran the ads. In practice it is close to impossible to recover money lost that way, because you have no relationship with anyone above your immediate partner.

How bad the upstream gets is contested, and worth stating as a range rather than a number. Digiday reported in April 2023 that publishers were describing contracted Net-30 terms being paid at Net-90 or Net-120 in practice, with some naming 150 to 180 days for the largest advertisers — one publisher's summary was that Net-30 "never worked" because the agency in the middle was always going to make it Net-60. Those are publisher-side accounts of direct and programmatic display, not a benchmark for casual game arcade revenue, and I would not plan a portal against them.

The transmission mechanism is what matters to you. Your ad partner cannot pay you faster than it collects, sustainably. When you are choosing between two networks on eCPM alone, you are ignoring half the decision. Ask what their own upstream terms look like and whether they carry the risk of an upstream default or pass it to you.

🌍 Carrier and B2B Money Is Slower Again

Operator revenue arrives through the longest chain in this business: subscriber, operator billing, aggregator, you. Every hop adds reconciliation time, and revenue-share agreements of this type commonly include holdback and clawback provisions so that a slice of each settlement is retained against later refunds, chargebacks and reversals. That is a normal commercial feature, not a red flag — but it means the settlement you receive is smaller and later than the gross number in the report, and the difference has to be modelled.

If you invoice European businesses — agencies, brands, event companies — the legal floor is knowable. Directive 2011/7/EU sets 30 days as the default B2B payment period, extendable by agreement to 60 days, with a statutory right to interest at the European Central Bank reference rate plus eight percentage points and at least €40 in fixed compensation per late invoice. The Commission proposed replacing the Directive with a stricter Regulation capping terms at 30 days in September 2023; Parliament adopted an amended version in April 2024, member states blocked it in Council, and it had effectively stalled by mid-2025. The Commission's late payment guidance is the current reference. Plan for 60 days on B2B invoices, and know you have a statutory claim when it goes past.

🛠️ Five Things That Actually Move the Trough

1. Buy the catalogue in tranches. A portal does not need 500 titles on day one to test whether anyone shows up. Take a starter set, prove traffic, and time the next tranche against a payout date rather than a launch date. Catalogue size is a lever you control; payout dates are not.
2. Negotiate the payment schedule, not only the price. Quarterly instalments on an annual licence at the same headline number is a materially better deal than the same number due on signature, and licensors will often say yes because it costs them little. Ask before you argue about the price. Our breakdown of licence models and pricing is the place to start on which structure suits your case.
3. Consolidate accounts so you clear thresholds. Three properties each earning US$60 a month against a US$100 threshold is three balances rolling over indefinitely. One account earning US$180 gets paid every month. Fragmentation is a self-inflicted cash delay.
4. Put one fast-settling revenue line in the mix. Card acquiring settles on a rolling basis measured in days, not months. A direct subscription or membership line pays for running costs while the ad money is still in transit, which is a different argument for it than the usual one about ARPU.
5. Treat a large deal as a financing question too. When an operator or a brand deal lands, the payment terms are as consequential as the revenue share. A carrier portal deal that pays quarterly in arrears needs more working capital behind it than a smaller one that pays monthly. Ask when, not just how much.

⚠️ The Failure Shape: Profitable and Insolvent

The portals that die from this do not look sick. Traffic is fine. Retention is fine. The margin per thousand plays is exactly what the model said. There is simply no money in the account in month three, because the catalogue was paid for in month zero, the first payout is threshold-gated into month four, and the running costs never paused to wait.

The founder's instinct at that point is to buy more games, run more traffic, chase more revenue. Every one of those moves deepens the trough before it fills it. The correct move is the boring one: stop widening the gap, collect what is owed, and only then grow into the headroom.

🌲 Where This Touches Forestry Games

Forestry Games has licensed games since 2017 and today runs a catalogue of 1,049 HTML5 and Android titles, sold to portal operators, telecom operators, agencies and brands. That means both sides of this problem are visible from here — the licence structure on one side and, through monetization work with portal operators, what the revenue side actually settles like. If you are sizing a first tranche rather than a full catalogue, that is a normal conversation to have, not an awkward one.

✅ What To Do Before You Sign

Build a twelve-month cash line, not a twelve-month P&L. Same revenue assumptions, but shift every receipt to the date it actually arrives under the published terms of the partner paying it — 21st-to-26th plus clearing for AdSense, NET 60 for a platform that says NET 60, 33 days after fiscal close for Apple, whatever the operator contract says for carrier money. Add the payout threshold as a gate, not a rounding assumption.

Then read the lowest point on that line. That is the amount of capital the portal actually requires, and it is the number to have settled before the catalogue invoice arrives — not after. If the trough is deeper than what you have, the fix is a smaller first tranche and a staged catalogue, not a more optimistic eCPM. Start with the tranche you can fund to month five, and let the payouts buy the rest.