DoubleDown's web shop took over half the revenue. The marketing budget barely moved

The swing itself, and the casino marketing line that barely moved#

A fox does not abandon a good den because a better one is fashionable. It moves when the ground shifts underneath. DoubleDown Interactive moved, and the trail through its second-quarter numbers is the most useful casino marketing document an operator will read this month, because it separates two things this industry habitually blends together: buying growth, and keeping more of the growth already won.

Take the figures plainly. Direct-to-consumer revenue rose to $40.5 million in the second quarter of 2026 against $10.7 million in the second quarter of 2025, and direct-to-consumer revenue as a percentage of total social casino revenue increased to 52.4 per cent from 15.4 per cent, reported by PocketGamer.biz and set out in the company's own Form 6-K exhibit lodged with the Securities and Exchange Commission. In four quarters, the shopfront DoubleDown owns went from roughly a sixth of the business to comfortably more than half of it.

Now the line that stayed out of every headline. Sales and marketing cost $13.857 million in the quarter against $13.087 million a year earlier, a rise of 5.9 per cent, while total revenue rose 11.2 per cent to $94.3 million and social casino revenue rose 11.5 per cent to $77.3 million. Read those sentences together and the casino marketing conclusion writes itself. Nobody bought this shift. The budget barely stirred in its burrow while the revenue mix turned over completely.

That is the whole quarter in two bars. No new market opened, no new title carried the period, and no casino marketing campaign turned the mix over. The same players, on the same games, paid through a different door, and the company kept more of what came through it.

For anyone drawing up casino marketing budgets for 2027, that is the uncomfortable part. The largest margin improvement available to a social casino publisher this year did not live in a media plan at all. It lived in a payments flow and a lifecycle programme, the two lines least likely to be represented when the media budget is signed off.

Be precise about what direct-to-consumer means here, because the phrase has been stretched to cover almost anything. In DoubleDown's own definition it is revenue from purchases made through channels the company owns, including web storefront transactions and other direct payment flows. A different till, in a shop the company already had, selling the same goods to the same people.

A margin event, not a spending event#

The company's account of its profit improvement names the mechanism without ornament. The rise came primarily from higher revenue, a favourable movement on foreign currency, and lower cost of revenue attributable to a higher proportion of direct-to-consumer revenue, partly offset by higher operating expenses. Strip the currency line out and what remains is a cost-of-sales story. Adjusted EBITDA rose 17.2 per cent to $39.3 million, and adjusted EBITDA margin reached 41.6 per cent against 39.5 per cent a year earlier.

Set that beside the spend line and the shape of the quarter is obvious. Direct-to-consumer revenue nearly quadrupled. The casino marketing and sales line moved by less than a million dollars. Whatever a web shop is, it is not a media buy, and treating it as one is how an operator ends up briefing a performance agency to solve a payments problem.

There is a broader hedgerow here, and the scent runs along it. PocketGamer.biz's quarterly survey of publisher disclosures puts DoubleDown at the sharp end of a general move rather than out on its own. Playtika reported direct-to-consumer revenue of $286.9 million at 39.3 per cent of its total, up from 25.3 per cent a year earlier. Playstudios reported $14.7 million, or 34.4 per cent of virtual currency revenue, against 13.9 per cent.

Playtika also supplied the number that explains the appetite. Through its own platforms it puts payment processing fees and related expenses for in-app purchases at roughly 3 to 4 per cent, where third-party providers typically charge between 5 and 15 per cent, per the same survey. Against a store commission, either figure reads as found money, which is why casino marketing plans across the sector are being rewritten this autumn.

Stillfront put the trade-off more honestly than most, describing the migration of players onto its own platform as negative for net revenue but accretive to gross profit and gross margin. That is the sentence a casino marketing plan should be built around. Selling direct frequently costs top-line revenue, because somebody has to be given a reason to move, and it pays you back below the line. A plan that promises both at once is selling a story rather than a model.

What the filing does not say about social casino games#

That is the company's own account of the why, and there is little reason to doubt it. It is not, however, a claim that the web shop improved the underlying game, and the filing draws a distinction most of the coverage skated straight past. Average revenue per daily active user for the social casino and free-to-play games rose to $1.42 from $1.33, and the payer conversion ratio rose to 9.4 per cent from 7.0 per cent. The company attributes those improvements primarily to the inclusion of WHOW Games.

In other words, the two metrics an operator would most like to read as proof that selling direct makes players better are, on the company's own telling, largely an acquisition arriving in the comparison. PocketGamer.biz flagged the same caveat, noting that DoubleDown completed the WHOW Games deal in July 2025 and that this may have affected the results. Anyone quoting the ARPDAU rise as evidence that a storefront lifts monetisation is quoting a merger, and a casino marketing plan built on that reading will not repeat the result.

There is a blunter number underneath. Average monthly revenue per payer for the social casino and free-to-play games fell to $218 in the second quarter of 2026 from $286 a year earlier. More payers, each worth less. That is frequently what a payer conversion ratio jumping from 7.0 per cent to 9.4 per cent means in practice, and it is exactly the detail that decides whether a retention programme is deepening a relationship or merely widening a funnel.

The audience figures agree. Average monthly active users came in at 1,252 thousand against 1,163 thousand, and average daily active users at 597 thousand against 578 thousand. Modest growth, nothing resembling the four-fold move in direct revenue. The business did not stumble into a thicket full of new players. The quarry was the same; only the till moved.

None of this makes the swing less impressive. It makes it a different achievement from the one being applauded. Margin earned by changing the payment route is real, durable and repeatable by anyone willing to do the plumbing. Player quality arriving with a merger is real too, and entirely separate, and any casino marketing agency presenting both as one story should be asked early which is which.

This matters beyond one publisher, because social casino games are the category where the direct-selling case is easiest to make. Payment friction is low, session frequency is high, and the player already has a relationship with a brand rather than a store. If the honest reading is margin plus a merger, a casino marketing pitch promising both in a weaker category deserves more scrutiny.

The mechanism is a court order, and it carries a date#

Here is the part that turns a company result into a sector decision. The reason a web shop became viable at this scale is not clever design. It is an injunction. In Epic Games v Apple, decided by the United States Court of Appeals for the Ninth Circuit, the panel found that Apple claimed to comply with the injunction while prohibiting App Store developers from using buttons, links and other calls to action without paying a prohibitive commission, and restricted the design of those links to make them difficult to use.

The panel concluded that charging a 27 per cent commission had a prohibitive effect, in violation of the injunction, and held that the district court did not abuse its discretion in finding Apple in contempt. Apple then updated its rules: an Apple developer news post in May 2025 recorded a guidelines update made for compliance with the decision, and the App Review Guidelines now state that entitlements are not required for developers to include buttons, external links or other calls to action in their United States storefront apps.

So on Apple's United States storefront, at this moment, a linked-out purchase carries no App Store commission. That sentence is the economic basis for the swing in DoubleDown's mix, and for the same swing at Playtika, Playstudios and Stillfront. It is also the sentence most likely to change, since PocketGamer.biz notes Apple has proposed charging up to 15 per cent on purchases made outside the store, and any final decision moves every model built on today's zero.

Google's side of the hedgerow never matched that and does not pretend to. Google Play's own policy record for United States developers confirms the Ninth Circuit upheld the injunction on 12 September 2025, that Google will not require the use of Google Play Billing in apps distributed on the Play Store, and that on 22 July 2026 it gave notice that developers enrolled in the external content links and alternative billing programmes must report transactions and successful downloads and pay the relevant service fees starting on 1 October 2026.

The rates are published, and they are not nothing. Google's alternative billing page for users in the United States sets out three tiers, and which one applies turns on when the player installed the app rather than on anything the operator does today.

Read the right-hand column twice. A player who installed before 30 June 2026 carries a 25 per cent standard fee on an alternative-billing purchase of digital goods, against 20 per cent for a newer install, and Google Play's service fee documentation sets the same cohort split across its wider schedule, alongside a 5 per cent billing fee and a 20 per cent alternative for external web links. The external content links programme adds a fixed fee per install of $3.65 for games and $2.85 for apps.

There is one more moving part. Google's policy page records a new settlement agreement entered with Epic on 4 March 2026, with the parties asking the district court for a revised modified injunction. Elsewhere the picture differs again: Google's user choice billing programmes cut the service fee by 4 per cent across more than thirty-five countries, and its external payments programme in Japan runs on separate rules. There is no single global casino marketing number to model.

The practical casino marketing reading is short. The window is asymmetric, dated and provisional. Build the direct route now, while Apple's United States commission on a linked-out purchase is zero, and write the revenue model in pencil, with the 1 October fee start sitting in the assumptions rather than the footnotes.

Player acquisition, casino CRM and what a casino marketing agency should model#

Ask the person doing the work and the answer is not about payments at all. Faith Price, director of growth marketing, user acquisition and user engagement at DoubleDown Interactive, told PocketGamer.biz that selling direct has not broken the company's user acquisition model, but has surfaced how little of the surrounding infrastructure was built with it in mind, naming attribution, mobile measurement partners and media partners as unprepared for the shift.

That is the first line item a casino marketing agency should price properly. When a meaningful share of purchases completes on a website rather than in the app, the attribution chain that justifies every acquisition bid runs straight through a gap. Price described DoubleDown pushing its own vendors on how the information gets attributed, an odd position for a publisher and a plain warning to any smaller operator assuming the tooling exists.

The second line item is retention, and this is where casino CRM stops being a support function and becomes the growth engine. Price is explicit that lifecycle marketing has always carried the job of explaining the website's value to the player, across the sixteen years DoubleDown Casino's site has existed. Players want something they understand, that they trust and that is easy, she says, and the lifecycle job is to give them exactly that and clear the roadblocks.

Which reframes the whole exercise. A web shop is not a channel to be bought, it is a habit to be taught, and the teaching happens in email, in-game messaging, retargeting and promotions working from one view of the player. Price runs user acquisition, lifecycle, retargeting, app store optimisation and answer engine optimisation together, and says the org chart matters less than whether those teams cooperate rather than compete.

The third item is the one nobody wants on the slide. Price says she cannot point to anything in acquisition measurement that has substantially improved since the tracking changes, only that the industry adapted, and that finding new players is much more expensive and much more difficult than five years ago. Player acquisition economics in social casino games have not recovered. For many publishers the margin unlocked by a shop they own is the only thing that has moved in their favour.

Her scepticism about channel tests follows from that. Another demand-side platform is unlikely to be the answer after a decade of trying them, she argues, because a player watching ads in games has probably already seen the ad. That points casino marketing budgets towards paid social and owned content built around a known audience, rather than another line in a paid search plan already optimised to its floor.

So the honest casino marketing sequence runs backwards from the till. Fix the payments route and the fee model first, because that is where the money is and where the court order put it. Rebuild attribution second, because every acquisition decision after it is otherwise guesswork dressed as a dashboard. Then spend, into a lifecycle programme that already knows how to explain the shop.

That order is why folkfox keeps returning to sequencing in this vertical. Ontario taught the same lesson from the other direction, where a sportsbook brand entered a market with no sportsbook because casino held four fifths of the revenue. The welcome offer taught it again, where the loudest number in the funnel turned out to be the most expensive one. And Allwyn's quarter taught the honesty part: read the footnote before you credit the headline.

The pattern repeats outside iGaming. PepsiCo's media agency consolidation is a margin decision in a growth costume, Grindr's nine per cent who pay is the same lesson about an audience you already hold, and creator selection scored on citation and Arbitrum's partner revenue turn on one question. Follow the scent to the money, then see who takes a cut.

Frequently asked questions#