The Trade Desk (TTD) Guided Q2 to "At Least $750 Million." It Delivered $715 Million, and Q3 Is Guided 12% Below Last Year.
It missed the Street. More importantly it missed the floor it set for itself 91 days earlier, only the second time in seven quarters. Q3 is guided 12.1% below last year with adjusted EBITDA down 49.5%. Inside: the derived U.S. revenue decline and four scenario price zones.
Exactly one year ago today, on August 7, 2025, The Trade Desk touched $91.45 a share. As of 8:04 a.m. Eastern this morning, the Nasdaq real-time tape had it at $12.64, down 28.47% on the session. That is an 86% round trip in 366 days, and roughly $5 billion of market value removed overnight.
The company reported second quarter results after the close on Wednesday, August 6. The headline everyone ran with was the analyst miss: revenue of $715.1 million against a consensus near $751.5 million, and adjusted earnings of $0.34 a share against $0.40 expected. That framing is accurate, and it is also the least interesting thing that happened.
The Trade Desk does not guide to a midpoint. For years it has published a floor. Every quarter, the press release says the same thing: revenue at least some number. On May 7, 2026, the company told the market that second quarter revenue would be at least $750 million, with adjusted EBITDA of approximately $260 million.
It delivered $715.057 million and $241.279 million.
That is $34.9 million below its own floor, not the Street's estimate. The company set the number, had ninety-one days to watch the quarter develop, and came in 4.7% under the line it drew itself.
Why the word "at least" matters here
A floor is a different promise than a forecast. When a company says "approximately," it is estimating. When it says "at least," it is committing to a downside bound, and the entire informational value of that phrasing rests on the bound holding.
For The Trade Desk, it almost always has. Pull the last seven quarters of guidance out of the 8-K filings and set them against what the company actually delivered:
| Quarter | Guided revenue | Actual | Variance |
|---|---|---|---|
| Q4 2024 | at least $756M | $741M | -2.0% |
| Q1 2025 | at least $575M | $616.0M | +7.1% |
| Q2 2025 | at least $682M | $694.0M | +1.8% |
| Q3 2025 | at least $717M | $739.4M | +3.1% |
| Q4 2025 | at least $840M | $846.8M | +0.8% |
| Q1 2026 | at least $678M | $688.9M | +1.6% |
| Q2 2026 | at least $750M | $715.1M | -4.7% |
Two things jump off that table.
The first is that this has happened before, once. The Q4 2024 miss was reported on February 12, 2025, and it took roughly a third off the share price in a day. It is also, not coincidentally, the closing date of the class period in the securities fraud actions now consolidated against the company in the Central District of California, which allege misleading statements between May 9, 2024 and February 12, 2025.
The second is the shape of the cushion in between. After the February 2025 reset, management rebuilt credibility by beating its floor by 7.1%, then 1.8%, then 3.1%, then 0.8%, then 1.6%. The beats were real, but they were also getting thinner. By late 2025 the company was clearing its own bar by less than a percent. The margin of safety in the guidance had been eroding for four straight quarters before it broke.
The number in the release that is worse than the miss
The quarter that just closed is history. The guide for the one now underway is not.
The Trade Desk told the market it expects third quarter revenue of at least $650 million and adjusted EBITDA of approximately $160 million.
The Trade Desk did $739.433 million of revenue in the third quarter of 2025, with $317 million of adjusted EBITDA.
So the guidance is for revenue to fall 12.1% year over year and adjusted EBITDA to fall 49.5%. The implied margin is 24.6%, against 42.9% in the same quarter last year, a compression of 18.3 percentage points in twelve months.
That is the actual event. Not a decelerating growth company. A company guiding to its first meaningful revenue decline, with operating leverage running hard in reverse. Revenue guided down 12%, profit guided down half.
Chief Executive Jeff Green attributed the shortfall on the call to two things: macro pressure on the world's largest brands, and the company's own execution. Consumer packaged goods and automotive, which management said represent roughly a quarter of the business, are absorbing tariff and input-cost pressure and cutting media budgets. Some advertisers, Green said, are choosing cheap inventory over quality inventory. New Chief Financial Officer Nate Olmstead, in the job since this spring, told analysts that "visibility is somewhat more limited than it has been in recent history."
Both explanations are plausible. Neither is testable from a conference call.
The filings are testable. The Trade Desk filed its 10-Q the same day, and the balance sheet carries a specific, unglamorous line item that tracks how much advertiser money is actually moving across the platform, independent of what the company chooses to book as revenue. That number has just done something it has not done before.
So the question that decides whether $12.64 is a dislocation or a repricing is not whether management sounded confident. It is this: does the third quarter floor of $650 million hold, and if it does not, what is the company actually worth?
The rest of this briefing is for paid members: the balance-sheet line that shows platform spend contracting for the first time, the derivation showing U.S. revenue went negative while management highlighted international growth, the tax mechanic quietly turning a falling share price into falling earnings, what the buyback tells you about management's own price expectations, and four scenario-by-scenario price zones with the arithmetic behind each.
AlphaBriefing Paid gets you every investment thesis, scenario framework, and catalyst brief we publish, the analysis private intel clients pay four figures for, at a fraction of that.