The Rent-Versus-Buy Line on a Game Licence Sits at 18 Months — Until You Order Ten
Two HTML5 game licensors publish rent and buy licence prices side by side. The break-even lands at 18 months, and bulk discounts pull it under a single year.
Most licensing conversations get resolved on cash flow. Renting spreads the cost, buying front-loads it, finance prefers the former, and the decision gets made without anyone doing the arithmetic. The arithmetic is not hard, and where licensors publish both prices it is not even ambiguous.
It turns on one input that almost no operator has measured: how many months a licensed title actually stays live on their platform before it comes down.
💵 Two Licensors Who Actually Publish Their Prices
Public rate cards in HTML5 licensing are rare. Most licensors quote on request, which leaves the market with very little visible arithmetic in it. Two exceptions are worth reading closely, because between them they cover both commercial models.
RAVALMATIC, a Barcelona studio, prints the whole card. Two tiers. A Regular game licence is 700 USD as a one-time fee, or 470 USD per year. A Premium licence is 1,000 USD one-time, or 670 USD per year. Source code is an optional add-on on either model — plus 210 USD on a regular title, plus 300 USD on a premium one. The bundle discounts are listed too: 25% off at three games, 30% at six, 40% at ten or more.
htmlgames.com, operated by Zygomatic, runs one model only. Every one of its 1,000-plus titles licenses non-exclusively at 1,000 USD as a one-time fee, with bulk discounts above five games and a site-lock so the build runs on your domain and nowhere else. There is no rental option. One sentence further down that page matters more than the price, and I will come back to it.
Famobi sells both shapes — a one-time payment it describes as lifetime, or a fixed monthly subscription scaling from a single title up to its full 500-plus catalogue — but publishes no figures. Useful as a shape, useless as a benchmark.
🧮 Both Tiers Break Even at the Same Point: 17.9 Months
Divide the perpetual price by the annual price and you get the years of renting that cost the same as owning.
- Regular: 700 ÷ 470 = 1.49 years — 17.9 months
- Premium: 1,000 ÷ 670 = 1.49 years — 17.9 months
The match is not a coincidence. The annual licence is set at 67% of the perpetual one in both tiers. Whoever built that card picked a ratio and applied it, which means the licensor holds a view about how long a title stays useful — and that view sits somewhere around a year and a half.
It is a defensible place to put it. Rent for one year and you have paid two-thirds of what owning would have cost, with nothing to show at the end of it. Rent for two and you have overpaid by a third. Rent for three and you have paid double.
So the naive reading is: rent for campaigns, buy for portals. That reading survives about four minutes of contact with the discount schedule.
📦 Order Ten Titles and Renting Loses Inside the First Year
Read the discount tiers again and notice what they attach to. The RAVALMATIC page is explicit — opt for the one-time fee licensing option for a bundle and the average price per game drops. The 25/30/40 percent discounts apply to perpetual purchases. The annual price does not move.
Recompute the break-even at each tier, regular licences:
- 1–2 games: 700 against 470/year → 17.9 months
- 3+ games, 25% off: 525 against 470/year → 13.4 months
- 6+ games, 30% off: 490 against 470/year → 12.5 months
- 10+ games, 40% off: 420 against 470/year → 10.7 months
At the top tier the perpetual licence costs less than a single year of the annual one. 420 against 470. You cannot rent ten regular titles for twelve months as cheaply as you can own them outright, and the premium tier behaves the same way — 600 against 670, break-even at 10.7 months again.
This is the part that gets missed, because buyers evaluate rent-versus-buy on one title and then scale the conclusion to the whole order. The discount curve makes that invalid. Volume does not nudge the rental case; it deletes it. Past ten titles there is no holding period short enough to make renting cheaper, unless you are licensing something for a single season and then pulling it.
Which is a real use case. A seasonal campaign, a three-month sponsorship activation, a pilot with an operator who has not committed to anything yet. If that is genuinely the job, rent. If you are stocking a portal, the rate card has already decided for you.
⏳ The Number You Need Is Your Own Removal Rate
All of it turns on one figure: the months a licensed title stays live before you take it down. Go and count yours. Pull the games you licensed twenty-four months ago and check how many are still in the catalogue, still linked from a category page, still taking sessions. Three outcomes, three different answers.
You have removed almost nothing
Typical of operators who treat the catalogue as inventory: titles go in, nothing comes out, the tail just gets longer. Your effective holding period is indefinite and perpetual licences are obviously right. Your risk is a different one — you are accumulating maintenance obligations, which is the next section.
You rotate a fraction each quarter
Common on telecom portals and white-label deals, where a merchandised front page carries perhaps forty slots and everything else is depth. The holding period for the merchandised set is short. For the depth it is years. Those are two different purchasing decisions and they should not share one contract. Rent the front page, own the depth.
You clear the deck every campaign
Agency and event work. Nothing survives the brief it was bought for. Rent, and spend the negotiation on minimum term rather than on rate.
The middle case is where most operators actually sit, and it is the one they handle worst, because they buy the entire catalogue on a single commercial model when the two halves of it have holding periods an order of magnitude apart.
📈 The Supply Side Is About to Move
There is a serious argument for renting that has nothing to do with cash flow. It is about what the catalogue you compete against looks like in eighteen months.
Poki commissioned Atomik Research to survey 2,000 weekly web gamers and 400 game developers across the UK and US, fieldwork completed in May 2026 and results published on 25 June 2026. PocketGamer.biz reported the headline: 53% of the developers surveyed plan to port a mobile game to browser within twelve months, and 41% plan to port a PC or console title.
Treat those as intentions rather than shipments. Survey questions about the next twelve months always overstate. But even at half the stated rate, the browser catalogue available to license in 2027 contains a materially different class of title than the one available now, including ported products with production values that casual-first HTML5 studios do not match.
A perpetual licence bought today is a bet that the title still earns its slot against that supply. For an evergreen mechanic — solitaire, mahjong, bubble shooter, match-3 — the bet is safe. Those have survived every platform shift since Flash. For anything trend-adjacent, it is not.
The same survey found 49% of web gamers try two to three different games in a single session. Players are not loyal to titles, they are loyal to the surface. That cuts both ways: you need breadth, and individual titles are more replaceable than their licence prices imply.
🛠️ Perpetual Buys the Files. It Does Not Buy Maintenance.
Back to that sentence on the htmlgames.com page. After the price, it reads: Six months support is included in the price.
That is the whole trap in eight words. A perpetual licence is perpetual in the rights it grants and finite in the obligations it creates. You may publish the game forever. Somebody has to keep it working forever, and after month six that somebody is you.
This is not hypothetical maintenance. Browsers have spent the last three years partitioning storage. MDN's documentation on partitioned cookies sets out where that landed: in a cross-site context such as an iframe, cookies now have to be explicitly opted into partitioned storage or reached through the Storage Access API, because the shared cookie jar embedded content used to rely on is gone.
A licensed game running in an iframe on your portal is exactly that cross-site context. If it saves progress, high scores or settings the old way, it quietly stopped. The player-facing symptom is not an error message — it is a game that forgets you. Nobody files a bug for that. It shows up as a retention line drifting down, six months after the person who signed the licence moved on.
Who fixes it depends entirely on what you bought:
- Annual licence, still in support: the licensor's problem. This is the strongest argument for renting and it appears nowhere on the rate card.
- Perpetual, past six months, no source code: nobody's problem. You hold a build you cannot modify and a supplier with no obligation to touch it. The options are a paid fix or a delisting.
- Perpetual with source code: your problem, and tractable. On the RAVALMATIC card that is 210 USD on top of 700 — 30% of the licence fee to convert a dead end into an ordinary maintenance ticket.
Priced against the alternative, the source-code add-on is the cheapest line on the card and the one buyers skip most often. It is also the line that makes the perpetual case hold up at all, because a perpetual licence without source is only perpetual until the platform moves.
📋 A Decision Rule You Can Apply This Week
Four questions, in order. The first one that returns a clear answer ends the process.
- Is there a fixed end date? A campaign, a sponsorship window, an event, an operator pilot with a review gate. If yes, rent — and spend the negotiation on minimum term rather than on the annual rate.
- How many titles? Above the licensor's top discount tier, perpetual usually wins before the first renewal, so the rental option is only worth pricing if question one said yes. Ask for the discount schedule before you ask for the price.
- Evergreen or trend-adjacent? Evergreen genres justify perpetual on their own. Trend-adjacent titles are worth renting even when the arithmetic disagrees, because the arithmetic quietly assumes you will still want it.
- Who maintains it in year two? If the answer is the licensor, you need an annual licence or a separate maintenance agreement. If the answer is you, you need source code. If there is no answer, you have not finished negotiating.
One more piece of diligence that costs nothing. Check when the licensor's catalogue page was last updated — RAVALMATIC's carries a visible stamp reading 35 games, updated 19th September 2023, which is more transparency than most licensors offer and tells you something worth knowing. Renewal is not refresh. If the reason you are renting is to keep the catalogue current, get it in writing that new releases are included in the subscription, because on most rate cards they are not.
🌲 Where This Leaves You
Rent-or-buy was never a cash flow question. It is a forecast about how long a game stays on your platform, and the published cards price that forecast at roughly eighteen months for a single title, under a year once you are ordering at volume.
Forestry Games has been licensing HTML5 and Android titles since 2017, and the catalogue runs to 1,049 games across both — which puts most portal deals well past the point where the arithmetic above has already made the call. Our licence and pricing page sets out how the models compare on our side, membership covers the recurring option, and source code is the line to read if you intend to run these titles past their second year. If you are stocking a portal rather than running a campaign, start from the catalogue and work backwards to the commercial model, not the other way round.
Whoever you license from, do one thing before you sign anything. Open your own analytics, count the titles you licensed twenty-four months ago that are still live today, and divide. That number decides the contract, and it is sitting in a database you already own.
