Connected TV for mobile game launches
Connected TV can work as a launch week awareness layer for mobile games, but only with QR-led creative and a proper geographic lift study.
A launch week media plan that skips connected television is a defensible choice. A launch week media plan that buys connected television and expects it to behave like a performance channel is not.
That distinction matters more now than it did when CTV first showed up on a mobile UA buyer's radar. Streaming inventory has kept growing; ad-supported tiers on the major services have normalised interruptive video ads in living rooms; ad tech vendors have built dedicated CTV products aimed squarely at app marketers. AppLovin CTV Ad Demand: Building Behind a 73% Quarter, covered in this outlet last year, showed how seriously the ad tech side now treats the format. None of that changes what CTV does well. It delivers reach with frequency against a broad audience at the moment a studio wants attention, and it doesn't report cost per install with any precision, however much a dashboard may imply otherwise.
Why launch week is the right moment
The case for CTV in mobile UA has always been narrower than the case for CTV in general brand advertising. Nobody taps a living room screen. There's no install button, no deep link, no immediate action a viewer can take from the ad itself, so what CTV buys is awareness density in a short window, which is exactly what a launch week needs and what an always-on performance channel does not. A studio pushing a new title toward the top charts benefits from a viewer having seen the game somewhere before they reach the app store listing, even if that viewer converts later through search or through a friend's recommendation, or through an entirely different paid channel the CTV flight will never get credit for.
This is also why CTV rarely earns a permanent line in a UA budget the way Meta or Google does. It behaves like an event cost. A buyer spends it in bursts around a launch or a major content update, or around a platform moment worth amplifying, rather than optimising it week to week against a CPI target.
What the creative actually has to do
Standard mobile UA creative, built around a hook in the first two seconds followed by a clear call to action, doesn't transfer cleanly to a ten or fifteen second CTV spot. There is no tap, so the ad has to carry the entire job of prompting a search later. Two mechanisms do most of the work. The first is a prominent QR code that stays on screen long enough for a second screen habit to complete the scan; that converts a slice of viewers immediately. The second is a spoken, memorable app name written into the script, because voice-driven smart TV remotes and connected speakers mean a viewer reaching for a phone is now more likely to say a name aloud to a voice assistant than to type it out.
Beyond that, CTV creative should carry more brand identity than a typical performance cut would: a consistent character, say, or a colour palette or an audio sting a viewer can recognise later on a store listing or a friend's screen. The asset doesn't need to prove gameplay depth the way a rewarded placement does. It needs to be recognisable.
Measuring it honestly
Most of the operator discipline sits here. CTV inventory doesn't carry the click or view-through signal an MMP can stitch to an install with any confidence, and forcing attribution onto it produces numbers that look precise and mean very little. The credible approach is a lift study: a geographic or household-level holdout, spend concentrated in test regions, then a comparison of install volume and search volume against regions that received no CTV exposure over the same window. That's more setup than most buyers want for a launch-week tactic. Which is exactly why it belongs before the spend rather than after, when the campaign has already run and nobody can reconstruct a clean control group in hindsight.
Even without a full lift design, two proxies are worth watching in the days immediately after a CTV flight: search volume for the game's name, and app store listing traffic. A visible bump in either, especially one that doesn't appear in the control regions, is a reasonable signal the spend did something. Flat search plus flat listing traffic against a real CTV flight is the signal to stop.
The budget threshold
Below a certain spend level, CTV in mobile UA is closer to theatre than to a working channel: too little reach to register above noise in search or lift data, too small a frequency build to leave any recognition behind. The exact number depends on market and inventory cost. The operating rule doesn't. If the planned spend can't support a genuine geographic holdout with a control group large enough to read a result, the money is better spent extending an existing performance channel or funding one more week of creative testing, both of which will at least produce a number someone can act on. CTV earns its place in a launch plan when a team sizes it to answer a question, not when it fills a line item marked awareness because the plan looked incomplete without one.
That's the discipline connected television demands from a UA team that most other channels don't. It isn't a channel to run continuously and optimise. Deploy it occasionally, measure it properly, and retire it the moment a launch window closes.
