Your Licence Ends on a Date. Removing the Game Takes a Quarter.

When a game licence expires, the contract ends on one date and the removal work runs for months. Here is what expiry actually costs a portal operator — and the clauses that decide how bad it gets.

Licence purchases get scrutinised hard at the front end. Buyers argue over price per title, exclusivity, domain counts, whether source code is included. Almost nobody argues about the end of the term, because the end of the term is three years away and the person negotiating will probably have changed jobs by then.

Then the date arrives, and it turns out that "the licence expires" is not one event. It is a removal project with a technical workstream, a data workstream, a billing workstream, and a set of third parties who never signed anything with you and have no reason to move quickly.

📅 Two Delistings This Year Show What "Term" Actually Means

Star Trek: Resurgence launched in May 2023 and was pulled from every digital storefront in April 2026, under three years later, because the publisher's licence to distribute it lapsed. Stevivor reported that the Xbox listing disappeared first, with Steam, Epic, PlayStation and Switch following over the following weeks — a staggered removal, not a clean switch-off. Players who already owned it kept their copies.

Three weeks ago, Noisy Pixel reported that Freezing What If…, a licensed strategy RPG from developer BuilderSoft and publisher TrinityGames, would come off Steam on 31 August 2026 on expiry of its IP licensing agreement — base game and all DLC. Same pattern: no new purchases, existing owners unaffected.

Neither is a scandal. Both are what a term-limited licence does when it reaches its term. The instructive part is the shape of the aftermath, because that shape is identical whether the product is a narrative console game or a 400-title casual catalogue sitting on a telecom portal.

It also isn't only character IP. When 2K removed Spec Ops: The Line from sale in January 2024, it confirmed to PC Gamer only that "partnership licenses" had expired — widely read as the licensed soundtrack. A single expiring music cue can take the whole product off sale. If you licensed games rather than built them, you probably do not know what is in the audio folder.

🔌 Withdrawal Is Not Deletion, and the Stores Disagree About Which

Most licence agreements say something like "Licensee shall cease distribution of the Licensed Games." Operators read that as a delete key. It isn't one, and what it means depends entirely on the surface.

On Android, unpublishing is the weakest action in the toolkit. Google's own help documentation is explicit: if a user has the app installed, "you can continue to use the app", but "you will not be able to update your app", and if they uninstall it they cannot redownload it unless the developer republishes. The line that catches people is the next one — "Google Play's billing system will not work while the app is not available in Google Play, even if the app is already installed."

Read that against a subscription-monetised APK. Unpublish on the expiry date and your existing subscribers keep the app on their home screen, keep playing, and stop being billable. You have not ended the relationship, you have ended the revenue from it while keeping every support obligation attached to it.

Apple behaves differently, and the difference is less well documented. Apple's public support material covers redownload troubleshooting rather than stating a blanket rule, but reporting and user accounts have consistently indicated that titles a developer pulls generally cannot be redownloaded from purchase history at all — the transaction stays visible, the install does not come back. Treat the two stores as having opposite defaults until you have checked your own case, because the compliance question a licensor will ask you ("is it withdrawn?") has a different answer on each.

🕸️ On the Web, Your Copies Have Copies

HTML5 is worse, and it is worse in a way that only shows up at expiry.

A native app has one distribution point you control. A web game has as many as you created. If your licence permitted sublicensing or partner embeds — and if you built a syndication business, it did — every partner who took an iframe is a separate removal task, executed by someone whose incentive is to leave the game up until they get around to it. Every static bundle you pushed to a CDN has an edge cache with its own TTL. Every partner who mirrored your build to their own hosting has a copy you cannot reach.

The practical consequence: your obligation to stop distributing is joint, but your control is not. If your agreement makes you liable for continued availability anywhere in your distribution chain, you need a removal mechanism that does not depend on partner goodwill. In practice that means one of three things, decided at build time rather than at expiry:

• Server-side game manifests. Partners embed a loader that fetches the title list from you. Removing a game is a config change on your side, live everywhere in minutes.
• Domain-locked builds. The game checks its host against an allowlist you control. Useful against mirrored copies, and it doubles as licence-scope enforcement while the term is running.
• Contractual takedown windows with your own partners. A specified number of business days, written into the partner agreement, mirroring whatever your upstream licence demands of you.

The failure mode is a catalogue delivered as a zip file to forty partners over three years, with no record of who deployed what where. That is not a hypothetical filing problem — it is the reason expiry becomes an email archaeology exercise.

🗃️ The Player Data Outlives the Game

Pull the games and you still hold accounts, saves, leaderboards, purchase records and whatever analytics you attached to them. If you were operating a white-label portal for someone else, you were a processor, and Article 28(3)(g) of the GDPR is unambiguous about what happens next: the processor must, at the controller's choice, delete or return all personal data "after the end of the provision of services relating to processing, and deletes existing copies unless Union or Member State law requires storage".

"Deletes existing copies" is the phrase that costs money. Backups. Warehouse tables. The analytics vendor. The support tool with the ticket history. None of that gets cleared by unpublishing an app, and none of it is cheap to do properly if nobody designed for it.

Decide two things before the term ends, not after: who the controller is for player accounts, and what the export format is. A telecom client who assumed they owned the subscriber gaming data, dealing with a vendor who assumed the same, is a dispute that runs longer than the contract did.

💳 The Revenue Stops Before the Cost Does

Expiry is asymmetric in a way that is easy to under-model. Revenue from the expiring titles ends on the date. Several costs do not.

• Subscription cohorts. If players subscribed partly for titles you are removing, churn shows up in the following billing cycle, and refund pressure shows up immediately. Carrier-billed subscriptions make this worse — a complaint goes to the operator, not to you.
• Ad inventory commitments. If you sold guaranteed impressions against a section of the portal you are about to empty, that shortfall is yours.
• Support and QA. Installed APKs keep generating tickets after they stop generating revenue, and they can no longer be patched.
• Replacement content. Sourcing, integrating and QA-ing substitute titles takes a normal content cycle, and you need it finished before the expiry, not after.
• Store presence. A listing pulled and later replaced does not inherit the old ranking.

The plain version: a portal that removes a meaningful slice of its catalogue on a fixed date without a replacement plan takes the retention hit twice — once for the missing games, once for looking abandoned.

📝 The Clauses That Decide How Bad This Is

Nearly all of this is cheap to fix at signature and expensive to fix at expiry. The clauses worth arguing over:

Sell-off or run-off period

A defined window after term end during which existing installs may keep operating, existing subscribers may keep being billed, and you may keep serving already-distributed copies. Standard in physical goods licensing, routinely absent from digital game licences because nobody asks. Ask.

Renewal notice window

Two dates matter: when the licensor must tell you their renewal terms, and when you must give notice. If the licensor can name a price 30 days out, you have no leverage and no time to replace the titles. Ninety days is a reasonable floor; the number should scale with how long content replacement actually takes you.

What survives the term

Confidentiality, audit rights, indemnities and reporting obligations usually outlive the licence. Check how long you must retain play and revenue records after termination — you may be required to keep exactly the data your own retention policy wants deleted.

Definition of "distribution"

Does serving an already-cached build count? Does an installed APK on a device count? Get it defined, because your compliance position after the date depends entirely on the answer.

Perpetual versus term, per title

Many non-exclusive HTML5 licences are sold as perpetual for a named domain, while branded and character-IP titles are almost always term-limited. Mixed catalogues therefore have mixed expiry exposure, and the branded titles — usually your best-performing ones — are the ones with an end date. Know which of your top twenty are term-limited before you build a marketing plan around them. Our licence and pricing page is the right place to start that conversation rather than assuming a catalogue is uniform.

🔁 Renewal Leverage Runs Out Long Before the Term Does

Here is the uncomfortable structural point. Your negotiating position is strongest on day one, when you have not yet integrated anything, and weakest at renewal, when the titles are embedded in your portal, ranked in your search, wired into your recommendation logic and named in your subscriber marketing. The licensor knows this. So the renewal conversation should happen at least two content cycles before the date, from a position where you have genuinely costed the alternative.

The concrete version of "costed the alternative": know which specific titles you would swap in, from which supplier, at what price, with how much integration work. An operator who can name replacements negotiates. An operator who cannot, pays.

This is also the strongest argument for keeping some depth in perpetual, non-branded content. A catalogue where the entire top tier is term-limited branded IP is a catalogue with a recurring hostage situation built into it. Ordinary HTML5 and Android content bought on perpetual terms is boring by comparison, and boring is what you want holding up the floor.

🚫 Five Ways Operators Get Caught

1. No expiry register. Nobody maintains a single list of which titles expire when. Build it in a spreadsheet on day one if you have to — the absence of this list is the root cause of almost everything else here.
2. Unpublishing a subscription APK and calling it done. Billing stops, the app stays, the obligations stay, the revenue does not.
3. Assuming partners will remove embeds because you asked. They will, eventually, and "eventually" is your compliance problem.
4. Treating expiry as a legal task. It is an engineering, content, billing and data task with a legal deadline attached. Legal cannot purge a CDN.
5. Discovering the renewal price in the renewal month. By then the games are load-bearing and the price knows it.

🎮 Where Forestry Games Fits

We have licensed games since 2017, across a catalogue of 1,049 HTML5 and Android titles, and we work with branded IP as well as our own in-house HTML5 development. That mix is precisely why we think this matters: branded content and perpetual content behave differently at the end of a term, and a buyer should know which they are getting before it is embedded in their product. Whether you are running a white-label portal, a telecom deployment or an app catalogue, ask us what the term is per title — a licensor who cannot answer that quickly is telling you something.

🧭 What to Do This Quarter

Open your licence folder and build the expiry register: title, licensor, term end, renewal notice deadline, whether sublicensing is permitted, and which partners currently hold a copy. Most operators discover during this exercise that they cannot answer the last column, and that discovery is worth more than the register itself.

Then pick the single licence with the nearest end date and dry-run the removal. How long to pull it from every surface? What happens to subscribers? What data survives, and under whose instruction? Do it on paper, in an afternoon, while the stakes are theoretical.

A licence you cannot exit cleanly is a licence you will renew on someone else's terms. The exit plan is not paperwork you do at the end — it is the thing that determines what you pay in year four.