Ad Tech Briefing: AppLovin’s AI-fueled surge and The Trade Desk’s stumble show where investors are placing their bets

AppLovin’s third-quarter numbers were the kind that usually end an argument. Revenue jumped 68% year over year to $1.4 billion, with net income and adjusted EBITDA growing even faster as the company leans into its AXON-driven advertising platform and shrinks its legacy gaming footprint. The market’s reaction was briskly positive, as the company’s stock popped (see table below).

However, what makes that reaction stand out is what it seemed to ignore. Just a month earlier, a Bloomberg report revealed that the U.S. SEC was probing AppLovin’s data-collection practices, following short-seller allegations that it skirted platform terms to squeeze more targeting out of Meta and mobile ecosystems. The stock fell about 14% on that headline in October, and further coverage detailed how the company shut down an “Array” product that had become a focal point of those accusations.

Yet when the Q3 numbers hit, Wall Street effectively compartmentalized the regulatory overhang, although, some analysts explicitly advised caution: growth forecasts point to a deceleration from the current breakneck pace, and the SEC probe is still unresolved. But for now, the combination of scale, margins and AI narrative appears to outweigh the governance red flags in investors’ models.

Big Tech vs. The open web

Meanwhile, The Trade Desk’s Q3 tells a more nuanced tale about sentiment toward the demand side of the ecosystem. The company delivered $739 million in revenue, up 18% year over year (22% excluding U.S. political spend), beating consensus and continuing its long streak of double-digit top-line growth.   CTV outpaced the rest of the business, and video now accounts for roughly half of revenue, underscoring The Trade Desk’s role as the de facto DSP for the “open internet.”

And yet, the stock fell 4–7% in the wake of the results, depending on which post-earnings snapshot you look at. Coverage framed it as a classic “great quarter, high bar”: the company beat on revenue and EPS, guided Q4 to at least $840 million, but is still working off a bruising summer reaction to its Q2 earnings.  

On the company’s subsequent earnings call, CEO Jeff Green again insisted Amazon isn’t a direct competitor because most of its ads target owned-and-operated inventory rather than the “open internet.”

However, investors appear split on that framing, with some seeing an independent DSP with durable CTV and retail-media momentum; others worry that Big Tech and retailer networks can squeeze the open web on both inventory and data. The Q3 sell-off doesn’t signal that the market has “given up” on The Trade Desk — but it does show that continued backing for the sector’s flagship DSP now depends on more than just beating the quarter. The story investors are buying has to reconcile competition from Amazon, Google and retailer media with a credible path to sustained 20%-ish growth and healthy cash generation.

Collaboration across the middle layers of ad tech is another trend investors are monitoring, and while not explicitly raised on the call, the fact that concerned parties are seeking a “grand bargain” to end months, if not years, of sniping over supply-path and auction mechanics is likely welcome. Against that backdrop, The Trade Desk and PubMatic — which reported its own Q3 earnings as Digiday was going to press (see chart) — announced a new API-based integration for deal-ID campaigns using The Trade Desk’s Price Discovery and Provisioning, or PDP, API. Observers have characterized the move as a pragmatic reset: both sides acknowledge that legacy deal-ID plumbing is too brittle for modern CTV and curated marketplaces, and are willing to align on common pipes to keep spend flowing.

For markets, deals like this suggest that even when buy- and sell-side rhetoric turns heated, the largest players will still cut operational truces where there’s real money on the table. In a sector where investors have been burned by opaque auctions and fee stacks, cleaner integrations that promise more predictable yield and fewer failed deals are a modest but meaningful de-risking signal.

‘Death of the open web’ narrative 

The other axis driving valuations is how convincingly companies can connect AI to revenue — and how quickly they can prove it on the open web, not just in pitch decks. DoubleVerify’s Q3 is a case in point. Revenue grew 11% to about $189 million and margins remained robust, but sales came in a touch light versus some expectations, and Q4 guidance implied high-single-digit to low-double-digit growth. The stock promptly dropped around 19%.

On the company’s earnings call, DoubleVerify leaned hard into its new DV AI Verification suite: Agent-ID to classify LLM agents and bots, and “AI SlopStopper” to detect and block synthetic or manipulated media across the open web. Management framed AI not just as a risk but as a margin lever, arguing that generative tools can 20x–2000x labeling productivity while opening new revenue lines in social, CTV and LLM-mediated advertising. Analysts pressed on how quickly those innovations could move the growth needle, and DV sketched a medium-term goal of taking social, streaming TV and AI verification from under 30% of revenue to roughly 50%. The share-price reaction suggests investors believe the strategy, but aren’t yet prepared to pay up for it without clearer acceleration.

Elsewhere, Teads, provided the cautionary mirror image. Its Q3 revenue came in around $319 million, missing expectations and down roughly 15% year over year on a pro-forma basis, with EPS swinging to a loss — and the stock slumped by about 30%. On its earnings call, management laid out a turnaround plan targeting $35 million in annualized EBITDA improvement and highlighted CTV and AI-driven initiatives as the growth engine.

Analysts focused heavily on how quickly Teads could stabilize core markets like the U.S., U.K., and France and whether its AI and CTV bets would stem share loss to rivals. The verdict from Wall Street — at least for now — is that execution risk is high, with the highly leveraged 2024 deal that led to the union of Outbrain and Teads, plus the “death of the open web” narrative, dragging down its stock price.

In total, the above developments from last week point to a more nuanced view of ad tech’s prospects, rather than the blanket enthusiasm (or cynicism) of past cycles. On one side are scaled platforms such as AppLovin and The Trade Desk, which can still post high-teens to high-double-digit growth, invest aggressively in AI and CTV, and throw off substantial cash. On the other hand, there are infrastructure and media players that must prove AI is more than a cost line — that it can deepen moats in measurement, verification, and curated inventory rather than simply compressing margins.

What we’ve heard

“Every player is optimizing for their own incentives instead of the health of the ecosystem. And somewhere in that noise, publishers — the ones creating the content and context that make advertising possible — are being left out.”

– Renowned industry consultant, Jana Meron, of Lioness Strategies, says the quiet part out loud.   

Numbers to know

• $70 billion: Anthropic’s forecasted 2028 revenue. 
• $4 million: The amount People Inc. executives claimed they spend each quarter on litigation against Google. 
• 58%: The percentage of late invoice payments in 1H 2025, per data from OAREX. 
• 50%: The amount of agency staffers who use Amazon DSP, according to a Digiday survey. 

What we’ve covered

News Corp explores multi-LLM licensing playbook

News Corp is scoping out the potential for a multi-licensing LLM portfolio strategy, according to sources familiar with the matter.  

Ranking is out, visibility is in as publishers chip away at AI search optimization

The AI search era is rewiring the goal from ranking to visibility, and publishers are slowly but steadily tuning their playbooks.

What we’re reading

Inside Bayer’s ad tech and data strategy to find new customers and prepare for agentic AI

Bayer is bringing its custom algorithm in-house so it can be more hands-on with the model that undergirds its programmatic ad placements, working with Chalice and Snowflake to achieve this aim. 

Google is helping WPP build Websites for AI search engines 

Google is giving WPP-owned agencies access to tools that help marketers better appear on AI search engines, including Google’s AI Overviews and rivals such as OpenAI’s ChatGPT and Perplexity.

ID5 acquires TrueData, combining strengths to create the most comprehensive ID solution 

Financial terms of the deal were not publicly disclosed, but the U.K.-based company states that by bringing TrueData’s graph on board, the company is upping its U.S. operations and strengthening its Adaptive Identity technology.

The DOJ and Google sharpen their remedy proposals as the two sides prepare for closing arguments

Both sides filed new post-remedy trial briefs on Monday, with the DOJ and Google each citing the phrase in support of their conflicting viewpoints on a potential divestiture of Google’s sell-side ad exchange.