A seasonal game is commissioned against a date. Unlike a catalogue title, which can ship when it is ready, a title built for a campaign has a fixed opening that was chosen for commercial reasons long before anyone wrote a line of code. That opening is not a preference. It is the deadline for every dependency in the chain — the build, the certification submission, the published explanation of the rules in each market, the offer wording, the load test and the rollback plan — and each of those dependencies has its own lead time.
This article is about that window and the offer attached to it. It is not the operator’s brief, which is covered in the branded commissioning brief, and it is not the platform’s licence position or distribution grant, which are covered in the platform commissioning guide. It is the set of decisions that a seasonal title forces earlier than a catalogue title does, and the test that a seasonal title invites people to run badly.
A window is an interval, not a date
The first thing a commission needs is agreement about what the words mean. A campaign window is an interval: an opening instant and a closing instant, in a named place, expressed in a format that a build, a marketing calendar and a reporting query can all read without translation. That is what the standard exists for. ISO 8601-1:2019 “specifies representations of dates of the Gregorian calendar and times based on the 24-hour clock, as well as composite elements of them, as character strings for use in information interchange”, and it is “also applicable for representing times and time shifts based on Coordinated Universal Time (UTC)”.
Writing the window as an interval rather than as prose settles three arguments that a seasonal launch otherwise has at the worst possible moment.
The first is which window is being discussed. A seasonal title usually has at least two: the promotional window, during which the campaign mechanic is available, and the availability window, during which the game itself is in the lobby. They rarely start on the same day and they almost never end on the same day. A commission that says “the game goes live for the season” without naming both has given the studio, the operator’s marketing calendar and whoever configures the lobby three different deadlines.
The second is which currency the closing instant is expressed in. “Ends Sunday night” is not a specification; a timestamp with an offset is. Every downstream job — the campaign switch-off, the bonus expiry sweep, the final reporting extract — reads that instant, and each of them will get it wrong if it has to infer the offset from a sentence.
The third is whether the interval is inclusive at both ends. Seasonal promotions are usually written as though the closing day is fully included, and implemented as though it is not.
Local midnight moves, so it cannot be a constant
The tempting simplification is to treat “midnight” as an arithmetic fact. It is not, and the reason is documented. The IANA Time Zone Database “contains code and data that represent the history of local time for many representative locations worldwide” and “is updated periodically to reflect changes made by political bodies to time zone boundaries, UTC offsets, and daylight-saving rules.”
A winter campaign that opens before a daylight-saving change and closes after it spans two different offsets in the same market. That is not an edge case in seasonal commissioning; it is the normal shape of a season. It means the closing instant has to be stored with its offset, the market’s offset has to come from maintained data rather than from a constant somebody typed into a configuration file two years ago, and the reporting query has to be able to answer the question in both the market’s local time and UTC. The time and clock discipline guide is where that discipline belongs in a regulated platform; the commissioning task is to make the calendar a data deliverable rather than a paragraph in a brief.
There is a responsible-design consequence too. A mechanic that says “you have until midnight” is a prompt, and the moment at which it stops being true should be the same moment in every surface a player can see — the lobby tile, the promotional panel, the game itself and the notification. A campaign that runs past its own countdown has published a claim it did not honour, whatever the internal configuration says.
The offer is a licence condition before it is a creative idea
Most seasonal commissions attach something to the title: free spins, a tournament entry, a prize draw, a bonus. In Great Britain that something is an incentive, and the LCCP condition 5.1.1, Rewards and bonuses — SR code, treats it as one. In the version effective from 29 July 2026 it applies where a licensee makes available to a customer or potential customer an incentive or reward scheme or other arrangement under which a customer may receive money, goods, services or any other advantage.
What the condition requires of that arrangement is where a creative idea becomes a specification. Licensees must “Set out terms and conditions, in relation to an incentive, which are clear, transparent, and fair and readily accessible to any customer or potential customer to whom it is offered.” They must not “Apply wagering requirements, which requires a customer to play through bonus funds, over a maximum of 10 times.” They must not “Include more than one type of gambling product (betting, casino, bingo, and lottery) within an incentive.” And they must not “Alter or increase the receipt or the value, or amount of the incentive if the qualifying activity or spend is reached within a shorter time than the whole period over which the benefit is offered.” Schemes designated high value, VIP or equivalent must be offered in a manner consistent with the licensing objectives. Alongside it, the ordinary-code condition 5.1.2 states that “Licensees should only offer incentive or reward schemes in which the benefit available is proportionate to the type and level of customers’ gambling.”
Read that as the person commissioning a seasonal title and the order of work changes. The wagering cap, the single-product rule and the no-escalation rule are not terms that marketing can settle after the game is built, because the bonus engine requirements decide whether the campaign’s mechanic can express them at all. A seasonal ladder that rewards a player more for qualifying faster is precisely the structure the condition prohibits. A campaign that pairs a casino draw with a sportsbook entry is two product types inside one incentive. Either of those discovered at configuration time is a rebuild, not a copy edit.
That is also why the offer wording belongs in the commission as an approved artefact. Somebody has to write the terms, and somebody has to be able to show that the version a player saw on the opening day is the version that was signed off — which is a release-management question, not a marketing one.
The campaign is marketing, and marketing has a code
The second regulatory surface is the campaign itself. LCCP condition 5.1.6 requires that “All marketing of gambling products and services must be undertaken in a socially responsible manner” and that licensees “comply with the advertising codes of practice issued by the Committee of Advertising Practice (CAP) and the Broadcast Committee of Advertising Practice (BCAP) as applicable”, with the same principles to be applied to media not explicitly covered. The Commission’s own advertising and marketing guidance, last updated 29 June 2026, restates it: “The advertising of gambling products and services must be undertaken in a socially responsible manner and you must comply with the UK Advertising Codes issued by the Committees of Advertising Practice (CAP) and administered by the Advertising Standards Authority (ASA).” Its separate page on marketing and advertising adds that in the Commission’s view marketing for gambling “should not appear in places where it is likely to be seen by large numbers of children” — which is a placement question that a seasonal campaign, with its own bespoke landing page and its own paid placement plan, tends to create rather than inherit.
The CAP Code’s section 16 is where the seasonal mechanic meets a written test. Rule 16.1 requires that “Marketing communications for gambling must be socially responsible, with particular regard to the need to protect children, young persons and other vulnerable persons from being harmed or exploited.” Rule 16.3 states that marketing communications must not “portray, condone or encourage gambling behaviour that is socially irresponsible or could lead to financial, social or emotional harm” (16.3.1); must not “exploit the susceptibilities, aspirations, credulity, inexperience or lack of knowledge of children, young persons or other vulnerable persons” (16.3.2); must not “suggest that gambling can provide an escape from personal, professional or educational problems such as loneliness or depression” (16.3.3); and must not “suggest that gambling can be a solution to financial concerns, an alternative to employment or a way to achieve financial security” (16.3.4).
Seasonal creative concentrates its risk in one place: scarcity. A countdown, a “last chance”, a limited edition, a prize framed as the thing that fixes a difficult month — those are the shapes that 16.3.4 and 16.3.1 are written against, and the responsible-product-design work for a market is where the market-specific version of that judgement sits (responsible product design requirements). None of this is a legal conclusion, and none of it is settled by a supplier; the point for the commission is that the copy and the mechanic are reviewed against a code before the build, not after the campaign opens.
The window has to absorb everything that precedes it
Once the calendar and the offer are fixed, the commission has a schedule it can check, and the checks are the ones that fail late otherwise.
- Certification. A submission to a test house is a queue with its own lead time, and it is tied to a build. If the seasonal title is a variant of an existing one, the commission should say what changed and therefore what has to be re-submitted; the certification submission guide is where that cost lives.
- The published explanation. Each market sets its own requirements for what a player must be able to read before committing to gamble, and a seasonal variant that introduces a new mechanic, a new prize structure or a limited-time paytable changes what has to be explained. The rules specification guide keeps the explanation and the mathematics describing the same game.
- Peak load. A seasonal window is the traffic case the rest of the year does not test. The load and soak testing guide exists because a platform’s normal-week evidence does not describe its busiest week, and a seasonal opening is not the moment to discover that.
- Freeze, rollback and the game after the window. The commission needs a content-freeze date, a named rollback decision, and an answer to what the title does when the campaign ends: stay in the catalogue, retire, or come back next year. That last answer changes the build, so it belongs in the brief rather than in a post-campaign meeting.
A useful form for all of this is a single table with one row per dependency: the date it must be done, the role accountable for it, and the date after which the launch date itself has to move. Seasonal commissions fail on the last column far more often than on the first.
Test the window; do not narrate it
The seasonal title arrives with an irresistible story attached: the campaign ran, the numbers moved, therefore the campaign worked. HM Treasury’s evaluation guidance sets out why that inference does not hold, and it does so in one sentence: impact evaluation assesses “the extent to which the changes can be attributed to the intervention, over and above what would have happened had the intervention not taken place. This is complicated because there will be other influences that must be understood to claim that the intervention has had an effect.” The Magenta Book then names what is missing from a before-and-after comparison: “Experimental and quasi-experimental approaches infer the impact of an intervention through statistical comparison to a group or time period unaffected by the intervention. This unaffected group acts as a proxy for what would have happened to the affected group in the absence of the intervention and is commonly called the counterfactual.” Its annex adds the two families: in an experimental design the groups are “effectively identical (typically, through randomisation)”, while in a quasi-experimental design they “differ in known ways that can be accounted for analytically”.
A season is the worst possible environment for the comparison that ignores this. The window is defined by a calendar event that moves player behaviour on its own, the title is new at the same time as the promotion, and the baseline a year earlier was a different product mix. The honest version of the seasonal question is narrower and decidable: of the players who could have been offered the title, did the ones who actually saw it behave differently on the measure we named in advance?
That question can be tested, and it is worth writing the test into the commission:
- Name the measure before the window opens. Registration-to-first-deposit is a reasonable primary measure for an acquisition campaign because it is defined in advance and does not depend on the campaign’s own reporting. Cost per acquisition is a different measure and needs the campaign’s spend attached to it; pick one as primary rather than assembling a dashboard afterwards.
- Name the comparison. A holdout that is not offered the campaign, or a market or brand that receives it on a later date in a stepped design, gives the counterfactual the guidance describes. What is not a comparison is this season measured against the same season last year.
- Say what the design cannot carry. A single-market launch with no holdout produces observations, not results. Reporting “first deposits rose during the campaign” is an observation about the period; the causal claim needs the design, and if the design was not built, the report should say so rather than borrow the credit.
- Keep the responsible-gaming read in the same plan. The players a seasonal campaign reaches are the same players the market’s protections cover, and measuring the campaign does not suspend them.
None of that requires a statistician in the room; it requires deciding, before the window opens, what would count as evidence and what would not. A comparison chosen after the numbers arrive is a narrative.
Write the window into the commission
A seasonal commission is cheap to make decidable. Ask for six things by name and the rest of the chain can be scheduled:
- the campaign window and the availability window as dated intervals with offsets, and the market data they are read from;
- the promotional mechanic and its approved terms, written against the conditions that govern incentives in each market it will run in;
- the marketing review record for the campaign copy and placement, against the advertising code that applies;
- the certification scope for the exact build, and who owns the market’s published explanation of the rules;
- the freeze date, the rollback decision and what the title does after the window closes;
- the primary measure, the comparison and the limit of what the launch can support.
The studio and platform partner conversation is where those six lines get a name against them, and the earlier they are asked the cheaper they are. A seasonal title that ships on time with an offer nobody had to reword is not luck; it is a calendar that was decided before the build, in writing, with each dependency’s deadline attached to a role.
Questions operators ask
When does the build have to start before a seasonal campaign date?
Work backwards from the opening instant through the dependencies that have fixed lead times: certification submission and its queue, the market’s published explanation of the rules, the load and soak test, and the freeze date before launch. The build itself is usually not the longest item. The date after which the launch has to move is set by the slowest dependency, not by the software, so write that date down when the commission is placed rather than discovering it in the final month.
Can a seasonal campaign offer free spins, a bonus or a prize draw?
In Great Britain such an offer is an incentive, and LCCP condition 5.1.1 sets out what the terms must do: be clear, transparent, fair and readily accessible; cap wagering requirements at a maximum of 10 times; keep a single type of gambling product inside the incentive; and not increase the benefit because qualifying activity is reached sooner than the whole offer period. Condition 5.1.2 adds that the benefit should be proportionate to the customer’s gambling. Other markets set their own requirements, so the specific offer belongs with your own advisers — but the offer’s shape is a design constraint on the game, not only on the advertisement.
Does a seasonal countdown or “last chance” mechanic breach the advertising rules?
It can, which is why it is reviewed rather than assumed. LCCP condition 5.1.6 requires all marketing of gambling products and services to be undertaken in a socially responsible manner and to comply with the CAP and BCAP advertising codes. CAP Code rule 16.1 requires gambling marketing to be socially responsible with particular regard to protecting children, young persons and other vulnerable persons, and rule 16.3.4 prohibits marketing that suggests gambling can be a solution to financial concerns or a way to achieve financial security. Scarcity framing is exactly where that risk concentrates, so the countdown and the surrounding copy should be reviewed against the code in the market where they will run.
How do we know whether the seasonal title actually worked?
By naming the measure and the comparison before the window opens, and by reporting the limits of the design alongside the result. HM Treasury’s Magenta Book explains that attributing a change to an intervention requires comparing it with what would have happened without it — the counterfactual — and that a comparison group or a randomised design is how that is obtained. A season measured against the same season a year earlier is not that comparison, because the calendar event, the product mix and the baseline all moved with it. Without a holdout or a staged rollout, the launch produces observations about the period, not a causal result.
Does a seasonal variant change what has to be published for the game?
It can. A variant that introduces a new mechanic, a new prize structure or a limited-time paytable changes what a player has to be able to read before committing to gamble, and each market sets its own requirements for that explanation and for the information supporting an informed decision about the chances of winning. The commissioning decision is to name, per market and per launch language, who authors and signs the published explanation and who owns it when the paytable changes at the end of the window. The rules specification work is what keeps that explanation describing the same game the mathematics describe.
