Ready For Liftoff

研报英文原文证据摘录

Ready For Liftoff

29 June 2026

Liftoff

Meta, YouTube, Pinterest, Snap, and Amazon -- we estimate a market

opportunity of ~$77bn in FY25. Within this $75-$80bn market in FY25,

Liftoff generated ~$682mn of Core Advertising revenue; if we assume a

30-35% take rate, that translates into advertising spend of ~$1.9-$2.3bn

or 2-3% market share. Competition is abundant within this market, but

among its closest peers -- Applovin and Unity -- that are also leveraging

neural-net-based algos within their DSPs, the algorithms are trained over

different data sets, and for different objectives. Hence, they each

approach targeting mobile app users in a different manner consistent

with their unique training and objectives. By 2030, the company

estimates the market for independent platforms, such as Liftoff, could

increase to ~$135bn, or at a ~11% CAGR, supported by shift of ad dollars

from platforms such as traditional TV which monetize at much higher

rates, increase in the number of mobile app users (which could be

modest), and app engagement (which could be more substantial), as well

as growth in newer verticals (such as GenAI apps/chatbots or prediction

markets).

• Neural-net based Cortex delivers strong performance for

advertisers. The mobile app economy is very fragmented with over 4mn

apps and more than 5bn mobile app users who spend ~3hrs per day on

apps. The complexity of serving an ad to the most appropriate person at

the right time and the right place can be inferred from the fact that Cortex

currently powers >1bn predictions and handles 11mn ad requests per

second. After launching Cortex, Liftoff has been running 4x more tests

compared to pre-Cortex, processes ~21x more data, and has cut model-

training cycles from ~2 weeks to under a day. This productivity