Ad Tech's Q3 Divide: What AppLovin's Surge Means for SEM Budgets

Ad Tech's Q3 Divide and What It Means for SEM Budgets

As of November 2026, the Q3 filings of six publicly traded ad tech companies show a market that has stopped moving together. AppLovin's revenue rose 55% to $2.19 billion, while the other five combined generated about $1.94 billion (Digiday, 6 November 2026). Two competitors grew in double digits, and Teads fell 26%. For a buyer deciding where the next quarter's budget sits, and what a sem agency owes them in reporting, that gap is the clearest signal in the quarter: app and connected TV (CTV) inventory is growing, while much of the open web is contracting.

Key takeaways

AppLovin alone reported $2.19 billion in Q3 revenue, up 55%, more than the other five companies in the cohort put together.
Growth and profitability did not travel together: Unity grew revenue 32% while Criteo's net income fell 85%.
CTV was the one format that grew for almost everyone, even where total revenue sank.
Comparing revenue across these companies is a poor guide to performance; ex-TAC contribution and gross margin decide who is actually earning.
The gap between a growing platform and a struggling one is a service-level difference, which is why the reporting your agency hands you matters more than the headline growth rate.

The six companies in one table

The Digiday analysis of the Q3 results groups the six companies into two cohorts, mobile app platforms that grew and open-web businesses that shrank. The table below puts the reported changes side by side so a buyer can see the scale of the split.

Company	Change	Profitability note
AppLovin	+55%	Net income up 72%
Unity	+32%	Grow Solutions up 44%
Taboola	−6.9%	Gross profit rose 9.6%
Criteo	−11%	Net income fell 85%
Teads	−26%	Gross profit down 28%
Magnite	+14%	CTV ex-TAC contribution up 39%

Two of these rows tell opposite stories about the same format. Magnite, a supply-side platform selling inventory it does not own, grew by 14% because of CTV. Teads, a video advertising business with its own publisher relationships, lost more than a quarter of its revenue despite CTV strength.

A buyer at a desk compares a bright phone and television screen with a dim laptop showing web banners.

What the widening divide actually shows

App and CTV inventory versus the open web

The two strongest performers, AppLovin and Unity, both sell mobile app inventory. Unity's Grow Solutions division, which hosts its media offering, generated an increase of 44%, and its Vector ad platform grew 126% year over year and 40% sequentially. AppLovin told analysts its Q3 revenue exceeded the combined revenue of the other five companies.

The three decliners all depend on open-web placements: Criteo lost 11%, Taboola 6.9% and Teads 26% of revenue. Teads executives pointed to continued weakness in open-internet advertising and did not reinstate the company's financial guidance. Taboola's decline came with falling advertising impression volumes.

The shift is not a verdict on any single company's competence. It is a change in where attention and inventory sit. Apps and CTV are environments where a media buyer can address a specific viewer on a specific screen; open-web display is priced against a page rather than a person.

Why revenue alone misleads

Taboola lost 6.9% of revenue but grew gross profit 9.6%. Criteo lost 11% of revenue and still saw margin fall from 55% to 50% and net income drop 85%. Two companies with similar revenue direction ended up in very different places on the bottom line.

Traffic acquisition costs explain much of it. Criteo's contribution ex-TAC, the figure after paying publishers for traffic, fell 17%, a steeper drop than its revenue. When a ratio like that deteriorates faster than the top line, the business is buying growth it cannot keep.

For an advertiser, the lesson carries down to the account level. Revenue-side growth stories do not tell you whether your own spend is producing sales. The buyer's equivalent measure is return on ad spend (ROAS): revenue or transaction value generated divided by media spend. It is the one number that lets a growing platform and a shrinking one be compared on the same terms.

Where the CTV growth is coming from

Magnite's CTV contribution ex-TAC rose 39%, against 5% growth in its DV+ business of display and other digital video. CTV accounted for about 53% of Magnite's total contribution ex-TAC, up from a lower share a year earlier.

Outbrain's CTV revenue grew 37% year over year, but the channel was just 12% of quarterly revenue, so it could not offset a 26% overall decline. Teads reported nearly $100 million in year-to-date CTV revenue.

The pattern is consistent: CTV grows for almost every seller, but it only rescues a business whose revenue mix is already weighted toward it. A buyer should read a CTV growth claim the same way, as a share-of-mix question, not a headline.

A family watches a streaming service on a large television while a planner's laptop shows a channel mix.

Deciding whether to move budget

A marketing manager weighing a reallocation should work through three questions before shifting spend.

Establish the baseline. Pull the last two quarters of ROAS by channel, split by app, CTV and open web. If the client does not track revenue or transactions back to the source, the comparison is impressions against impressions, which tells nobody anything.
Check where the inventory sits. Ask the agency which exchanges and networks the budget actually clears through, and what share is app, CTV or open-web display. A campaign labelled "video" can sit in all three.
Set the test window. Move a defined slice, not the whole budget, and agree in advance what ROAS or cost per acquisition would justify keeping the new mix.

The Q3 numbers point toward app and CTV as the growth channels, but they do not say what a Singapore advertiser's own funnel will do. The test is the only thing that answers that, and before a larger shift a buyer can hold the existing inventory to the same standard through an sem agency singapore review of the current mix.

Where ad tech ends and martech begins

Readers often ask what separates ad tech from martech, and the Q3 results make the split worth spelling out because it decides which vendor is responsible for which number.

Ad tech is the machinery that buys and delivers advertising inventory, and also the machinery that sells it: demand-side platforms on the buying side, supply-side platforms like Magnite on the selling side, ad networks like Taboola and Teads filling the space between, and mobile ad platforms like AppLovin and Unity. Its money is media spend plus the fees taken on it. Its health shows in revenue and contribution ex-TAC, with gross profit telling you what the platform kept.

Martech is the machinery inside the marketer's own stack that stores customer data, runs email and CRM, and tracks lifecycle journeys. It is bought on subscription or usage rather than media spend, and its health shows in retention and seat count rather than ad revenue.

The boundary sits at the moment a person is handed to the ad platform. Martech knows who the customer is after they visit, sign up or buy; ad tech decides whether and how much they are shown a message before that. One vendor rarely owns both ends, which is why a Criteo-style blend of retail media and retargeting, now hit by two unnamed clients cutting business by $20 million and retail media revenue falling 23%, illustrates how awkward the middle can be.

sem agency singapore teams sit on the buyer side of that map. This is a service layer: the agency manages the client's media budget inside the platforms, is paid by management fee or service packaging, and reports against ROAS rather than publishing its own quarterly revenue. When a client asks whether the Q3 numbers mean they should move budget, the agency is the party that has to translate a platform's growth rate into the client's account performance, and Topkee's own material on how AI and privacy change Google SEM is written for that translation.

What Topkee Media does about it

Topkee Media works as a Google Ads partner agency, running search campaigns across Google, Yahoo/Bing and Baidu from a single service desk, and the Q3 split makes its measurement discipline the relevant part of the offer. The agency positions its work around return on ad spend rather than impressions, publishes its pricing as a monthly fee without a long lock-in, and keeps the media budget in the client's own ad account so the buyer retains visibility into where the money clears.

Its account audit service is designed to answer the question this article raises above: which channels are actually producing revenue, and which are producing impressions. That work runs through the agency's revenue tracking and lead conversion setup, which connects a transaction back to the campaign that paid for it. Its funnel audit is built to turn what it finds into a written list of changes for the account, and its conversion tracking is set up to attribute transactions back to the campaign that produced them, the same comparison the Q3 figures encourage.

For a buyer reading platform results, this matters because Topkee Media's fee structure does not reward volume for its own sake. A search campaign that produces fewer clicks but a higher ROAS is the outcome the service terms are built to chase.

An analyst and a marketer review a campaign dashboard at a desk beside printed audit notes and a laptop.

Reading a platform's numbers before you believe them

A vendor's own quarterly release is a marketing document as much as a financial one. Three habits keep a buyer honest when a platform claims growth.

First, look for the ex-TAC or gross-profit line, not just revenue. Taboola's profit growth alongside a revenue decline shows what that line can reveal. Second, separate a division from the parent. Unity's 32% growth hides a media business that grew 44% and a rest of the company that did not keep pace. Third, ask what share of revenue a growth channel represents, as with Outbrain's 12% CTV share, because a small mix cannot carry a large decline.

How SEM fits in a market that is splitting

Search remains the channel most directly tied to declared intent, which is why the buying decisions a well-run sem agency makes sit upstream of most other channels. A buyer searching for a service is closer to a transaction than a viewer scrolling a feed, and that is why the Q3 divide does not argue for abandoning search for app or CTV inventory. It argues for measuring all three against the same yardstick.

A Singapore advertiser can run this comparison without a large budget. Keep the search account as the baseline because it is the easiest to attribute, set a defined test slice for app or CTV inventory, and compare the two on ROAS after the same period. When an sem agency singapore team reports, ask for that comparison rather than a channel-by-channel impression count.

FAQ

Why did AppLovin grow so much faster than the rest?

AppLovin sells mobile app inventory, and it reported $2.19 billion in revenue, up 55%, with net income up 72%. Its Q4 forecast of $2.365 billion to $2.395 billion suggests its executives expect the trend to hold. The companies that declined, Criteo, Taboola and Teads, all carry significant open-web exposure.

Does a shrinking open web mean display advertising is finished?

No. Magnite grew 14% and its CTV business grew 39%, and Taboola grew gross profit despite lower revenue. What the filings show is that open-web display is no longer the default growth engine, not that it has stopped producing returns for well-managed campaigns.

Should I move my budget out of search?

Not on the strength of another company's numbers. Search sits closest to declared intent, and the Q3 results say nothing about a specific advertiser's funnel. Test a slice of app or CTV inventory against the search account's ROAS, then decide on your own data.

What does ex-TAC mean?

Ex-TAC means after traffic acquisition costs, the payment a platform makes to publishers for the inventory it sells. Contribution ex-TAC is the revenue the platform keeps, which is why Criteo's 17% fall in that figure is a more telling signal than its 11% revenue decline.

How can I tell whether my agency is doing a good job?

Ask for ROAS or cost per acquisition by channel, not impressions or clicks, and ask which exchanges or networks the budget clears through. A useful test is whether the agency can show the actions it took after an audit and what those actions changed.