Moloco launches Performance CTV: how an app marketer can test connected TV without a blind budget
Connected TV has long had a strange reputation. Everyone agrees that there is a lot of attention, the screen is large, and the brand coverage is beautiful. But as soon as an app marketer asks about installs, CAC and ROMI, the conversation usually shifts back to reach and frequency. That's why the launch Moloco Ads for Performance CTV is interesting in itself: the company is trying to sell CTV not as a beautiful showcase, but as a performance channel for applications.
On April 22, 2026, Moloco announced that it was bringing AI-powered Performance CTV to app marketers. In the press release, the company promises what the market has been lacking for a long time: optimization at the level of each impression, measurement through the usual MMP and transparent publisher-level reporting. On paper this sounds strong. But for a marketer, something else is more important: which of this can really be verified, and which remains a vendor claim.
What exactly did Moloco launch?
B official announcement Moloco says that CTV campaigns can be run through the same AI infrastructure that already runs in their mobile performance stack. This means that the goal is not just to buy impressions on TV, but to find households and devices that are more likely to proceed to install, register, or other app-event.
On Performance CTV product page Moloco specifically emphasizes three things:
- attribution goes through MMP, which the advertiser is already using;
- you can check the channel contribution through incrementality and ghost-bidding;
- CTV is viewed not as a separate world, but as a continuation of mobile UA.
This is the main point of the launch. Moloco is trying to remove the old problem of CTV, when buying and measurement live within the same platform, and the advertiser sees beautiful after-the-fact reports, but does not understand what the channel actually did.
What numbers are in the announcement?
Moloco provides strong numbers, but they need to be read carefully. The company writes that in early tests, advertisers received up to 1.5x higher ROI on CTV than on mobile when running campaigns on both channels in parallel. Another point from the announcement: approximately two-thirds of users who installed the application after viewing Moloco Performance CTV ad did so within six hours.
It's useful to remember that this is data from the platform itself. They're not useless, but they're not a neutral industry benchmark. Therefore, you need to treat them as a hypothesis for your own pilot, and not as a promised norm.
When should an app marketer test CTV?
Not everyone and not right away. CTV makes sense when mobile performance is already limited by saturation, and the cost of a quality user in familiar channels is growing. The second good scenario is when you have a clear post-install funnel, and not just an install volume. Otherwise, you'll just end up with an expensive top of the funnel with no confidence that it's getting to the money.
The test is especially appropriate if:
- the application already has a stable mobile UA;
- the team can count not only last click, but also view-through and incrementality;
- there are MMPs and resources for setting up cross-device measurement;
- the product can quickly see downstream events: registration, deposit, trial start, subscription or ad revenue.
How to check a channel without self-deception
The most common mistake in a CTV pilot is trying to evaluate it according to the same rules as a regular paid social. It doesn't work that way. The screen is different, the user path is longer, there are more intersections with mobile.
A normal test looks like this:
- Set one business goal: install, first deposit, trial start or D7 ROAS.
- Link the measurement to your MMP, and not to the default platform numbers.
- See not only attributed installs, but also incremental contribution.
- You compare CTV not with the “ideal channel”, but with what is already in your mix.
- Fix stop/go criteria in advance: how much the user costs, how many days you wait for a signal, what kind of increase in the quality of the audience is needed to continue.
If this is not the case, CTV will almost inevitably seem “too expensive” or, conversely, “too pretty.” Both conclusions are usually false.
What's really interesting about the Moloco launch?
The most useful thing here is not the fact of the release of a new product, but the shift in the logic of the market. CTV is no longer sold only as an outreach channel. Suppliers understand that app marketers do not need the gloss of television, but the same verifiable economics as in mobile.
This can also be seen from Devsisters case, where Moloco shows a decrease in CPI and the achievement of ROAS goals, and by the product architecture itself, tailored for MMP and transparency. Even if this specific vendor does not suit you, the direction of the market is already clear.
Conclusion
The launch of Moloco Performance CTV does not prove that CTV has suddenly become cheap and perfectly measured. But it shows that the market is finally speaking to the app marketer in normal language: install, value, incrementality, MMP, ROAS. And this is already good news.
If you look at the channel soberly, CTV should be tested not as a fashionable add-on, but as a controlled experiment with strict attribution and a clear stop tap. Then it has a chance to become not an image toy, but a working performance layer.