The Trade Desk Stock After Q3 2026: Q4 Still Guides Down 11%

I think the Q3 beat slowed The Trade Desk's decline without ending it. The stock jumped because revenue came in 13% above the company's own floor for the quarter. But about 3 points of growth came from a partner that won't recur, and the Q4 guide of at least $755 million is 11% below Q4 2025, in what is normally its biggest quarter and with midterm political ads inside it. Until a quarter prints back near last year's level, I treat the rally as a bet on a turn the numbers haven't shown yet.

The Trade Desk is a demand-side platform: software that agencies and brands use to buy ads across the open internet, from streaming TV to websites and podcasts. It keeps a cut of what clients spend. So its revenue is a read on how much ad money clients choose to route through it rather than through Amazon's or Google's own buying tools.

Q4 is when brands spend for the holidays. In 2025 The Trade Desk's revenue rose from $739 million in Q3 to $847 million in Q4, about a 15% step. The new floor implies about 3%. CFO Nate Olmstead said on the call that the company isn't expecting the normal seasonal uplift this year, and pointed to large global advertisers that remain under pressure.

Political ads make the floor weaker than it looks. They were a mid-single-digit share of Q3 revenue, and management expects them to be slightly higher in Q4, so the rest of the business is guided lower still. The Q3 beat also had help. Management said about 3 points of revenue came from a strategic partner it didn't name and doesn't expect to recur. Without that, Q3 was down about 3.5% on a year earlier.

The moat case for The Trade Desk is a network: more buyer money on the platform draws more publishers and data partners, which makes it better for the next buyer. Revenue can fall for three reasons. Clients leave, the cut shrinks, or clients send less through it. Customer retention stayed above 95% in Q3, as it has for over a decade, so clients aren't leaving.

The cut is harder to rule out. CEO Jeff Green said the take rate has stayed within about 1 to 1.5 points of its average over roughly ten years as a public company. A point of take rate is a real slice of revenue, so some of the decline could be price. But a band that has held for a decade doesn't explain a swing from 12% growth in Q1 to an 11% decline at the Q4 floor. My reading is that most of it is clients sending less spend through the platform.

That's the pattern I'd expect if the platforms that own their own shopper and search data are winning the marginal dollar. Amazon and Google can target ads with data they collect themselves; The Trade Desk has to rent it or rely on open identity tools. I laid out that split, with sourced Q2 figures, in the Netflix piece. The framework's live scores for the four names are below.

The strongest objection is that I'm making the same mistake the market made in August. The Q3 floor of $650 million was 12% below Q3 2025, and the company printed roughly flat. The Q4 floor is 11% below Q4 2025, which looks like the same playbook from a management team that missed badly in Q2 and doesn't want to be caught out again. If Q4 clears its floor by the same 13%, revenue lands around $855 million, slightly above last year, and I'm wrong.

There is operating evidence behind that view. Green said the number of joint business plans, multi-year spending agreements with big advertisers, is up more than 30% on a year earlier, with more than 100 in the pipeline, and that clients on them grow meaningfully faster. An upgraded bidding model in Kokai, its AI buying tool, cut cost per acquisition by a median 21% against the prior model in A/B tests across 62 campaigns. And a workforce cut of about 15%, announced in September, should take more than $100 million of gross cost out in 2027.

I take that seriously, but I don't think it reaches Q4. Strip the partner revenue out of Q3 and the beat over the floor was nearer 10%. The same beat on $755 million gets to about $830 million, still about 2% below last year, with more political spend than Q3 had. That already assumes a repeat of the best beat of the year; Q1's was under 2% and Q2 missed. Joint business plans and better tools are reasons the decline could end, and I'll change my mind when a print shows it.

1. 1February 2027, the Q4 2026 print: revenue against Q4 2025's $847 million. At or above it, the slowdown was a blip and I'm wrong.
2. 2The same release: the Q1 2027 floor, and whether management still describes large global advertisers as under pressure.
3. 3Retention and take rate: if either slips, the problem stops being where spend goes and becomes clients leaving or price.
4. 4The open measurement product Green said launches later this quarter, The Trade Desk's answer to walled gardens grading their own results.