Sports game UA runs on the licence calendar

Sports game acquisition is not primarily a creative problem or a bidding problem. It is an inventory-timing problem, and two things the UA team doesn't control set the timing: the sporting calendar and the licence agreement. A buyer who understands both will out-perform a better buyer who does not, because the auction for sports players is one of the most predictable in mobile. Predictability is where margin lives.

The uncomfortable corollary is that the moment most sports teams spend hardest, the opening fortnight of the season, is the moment they are paying the highest price for the least incremental user.

That is worth disagreeing with, so the mechanism deserves a proper explanation.

Why the season start is expensive

Every licensed sports title on a platform is contractually pointed at the same date. Roster updates land, the annual edition ships or the live game's season refresh goes out, and marketing rights that were dormant become active. Everyone opens the taps in the same week, bidding for the same fans, whom broadcasters and betting operators are also chasing, as are the leagues themselves.

Organic intent is also at its peak. Fans search for the game because the season has started, not because they saw an ad. A paid install at that moment has a high chance of being someone who would have arrived anyway. Sensor Tower's State of Gaming report in February described the industry shifting from download volume toward retention, and sports is the genre where that shift matters most, because the download curve is so heavily shaped by external events that raw install counts tell you almost nothing about what your spend did.

So the launch burst is expensive on the cost side and weak on the incrementality side. It is still usually necessary, because absence from the chart in that window has its own cost. But treat it as a defensive spend with a capped budget, not as the scaling opportunity of the year.

The approval loop is the real constraint

The second mechanism is slower and more damaging. Licensed creative goes through licensor approval, and approval takes days or weeks rather than hours. Any use of a club crest, a player likeness, a league mark, even a recognisable kit design, typically needs sign-off.

Set that against what the rest of the market is doing. AppsFlyer's State of Gaming for Marketers 2026, published in January, put the top advertisers at roughly 2,400 to 2,600 creative variations per quarter, up by a quarter or more year on year. A sports team whose every variant needs external approval cannot run that playbook. It will be iterating at a fraction of the market's speed while paying the market's prices.

Most coverage treats this as an annoyance. It is a structural disadvantage that changes what the creative programme should be. The correct response is to split creative into two streams with different governance:

• Licensed hero creative: few variants, long lead time, planned against the calendar, approved in batches before the window opens.
• Unlicensed system creative: gameplay mechanics, progression, team building, match moments shot without protected marks, iterated at market speed with no external gate.

The second stream is where most of the testing budget should go, because it's the only stream you can test at a rate that produces learning within a season.

A three-window calendar

Here is a framework that has held up across the sports titles this desk has watched. Every sports game's year divides into three windows, and each has a different job.

The pre-season window, roughly six weeks before the first fixture, is for building. Costs are low because competitors have not started, and organic intent is rising from a low base, so paid installs are more likely to be incremental. Hero creative gets approved here. Reactivation campaigns to lapsed fans of last season run here. Cohorts acquired in this window tend to have the longest tenure, because they arrive before the noise.

The launch window, the first two to three weeks of the season, is for defence. The team caps the budget in advance, sets bids to hold rank rather than chase volume, and resists the temptation to read a spike in installs as a success of its own making.

The off-season window, which for most sports is the longest of the three, is for the unlicensed stream. There is no external calendar, so the team can iterate on system creative, test new geos where the sport has an audience but the game has little presence, and rebuild the audience lists that the launch window will need.

An illustrative allocation, invented for clarity: forty per cent of annual budget in pre-season, twenty-five per cent in the launch burst, thirty-five per cent spread through the off-season. Teams that arrive at the planning meeting with the reverse shape usually find that the two most expensive weeks of the year have set their blended cost.

The trade-off most teams miss

Planning against the calendar creates a second-order problem: the game's cohort quality becomes lumpy. Pre-season cohorts are loyal; launch cohorts are noisy; off-season cohorts in new geos are cheap but slow to monetise. A measurement team that blends them into one monthly ROAS figure will draw the wrong conclusions in every direction, cutting the off-season spend that is actually building the following year's base and defending the launch spend that is mostly buying organics.

Label cohorts by window and read them against window-specific targets. That's a reporting change, not a buying change, and it's the one most sports teams haven't made.

There is also a quieter licensing risk. Licence agreements come up for renegotiation on their own cycle, and a marketing programme that leans too heavily on hero creative has nowhere to go if terms change. The unlicensed stream is a hedge as much as a testing ground. The teams that treat it as their primary creative engine are the ones who find licence renewal a negotiation rather than a crisis.

Related archive reading

• The UA industry calendar: events and earnings
• The misleading-ad arbitrage and why it still pays