Q2 earnings season delivered a verdict that had remarkably little to do with the quarter it was supposed to be judging. Companies that grew got punished. Companies that shrank got punished harder. And the only firms that escaped were, tellingly, the two big independent sell-side platforms.
The scoreboard reads like a category judgment rather than a set of individual ones. AppLovin grew revenue 53% and its stock fell 19.7%. The Trade Desk grew 3% and fell 22%. Criteo declined 11% and fell 24%. Taboola grew 2.4% quarterly and fell 27.5%. Against that backdrop, PubMatic (+20.8%) and Magnite (+8.6%) were the exceptions — the sell side, for once, the safe harbour. One analyst quoted in Digiday’s round-up put the mechanism plainly: investors are “pricing these companies on what it expects to happen next, not the actual facts of what just happened.”
What investors expect to happen next is not hard to decode, because the same coverage spells it out. Madison & Wall’s Luke Stillman observed that the walled gardens “were all still very solid” — and Meta, Amazon and Google are expected to take roughly 58% of US advertising this year, up from 56%. Two points of share in a market that size is an enormous transfer, and it is flowing away from the open-web companies whose stocks just got repriced, regardless of whether their own revenue lines were up or down.
A necessary caveat before the analysis: these are stock moves, not operating results. A share price is a bet about the future, and bets can be wrong — Wall Street has mispriced ad tech in both directions before. But publishers do not get to ignore the bet, because the companies being bet against respond to it.
The numbers in this piece
01The market is grading the pool, not the swimmers
The striking thing about this quarter is that performance stopped protecting anyone. AppLovin posted the kind of growth number that in any normal cycle buys a rally, and it fell nearly 20% anyway. The Trade Desk grew and fell. Criteo shrank and fell. When outcomes diverge that far from results, the market is not evaluating companies — it is evaluating the pool they all fish in, and concluding the pool is shrinking in relative terms as the walled-garden share climbs toward 58%.
The PubMatic and Magnite exceptions are worth reading carefully rather than triumphantly. The market currently rewards the independent sell side — which is a useful signal for publishers who route revenue through those platforms, and a mild vote of confidence in the supply layer’s durability. But two positive stock moves in a sea of markdowns is a thin foundation for strategy. What the exception more reliably supports is redundancy: the sell side is where the market sees resilience, and maintaining more than one healthy SSP relationship is how a publisher banks that resilience rather than merely observing it.
02Why a vendor’s stock price is your operations problem
A repriced vendor is not an abstraction. A company whose equity has been marked down 20–27% despite growing is a company under new pressure — from its board, from activist interest, from the simple arithmetic of employee stock compensation — to do something about it. The available levers are familiar: cut costs, raise take rates, chase a strategic pivot, or sell. Every one of those lands on publisher partners as something concrete — thinner support teams, changed fee structures, deprecated products, or a migration triggered by an acquisition nobody scheduled.
That is why partner equity health belongs on an ad ops dashboard and not just in a finance newsletter. You do not need to trade the stocks. You need to know which of your demand and technology dependencies just became more likely to change underneath you.
The scoreboard was never about last quarter, and neither should your planning be.
03Why this matters for publishers
| The macro trend is the real headline | A walled-garden share moving from 56% to roughly 58% of US spend means the open-web pool your programmatic revenue draws from is shrinking relative to the market every year. That is the structural argument for direct-sold, first-party-data and non-advertising revenue — not a talking point, a trend line. |
|---|---|
| Pressured vendors change behaviour | Repriced companies cut costs, adjust take rates, or get acquired. Any of those outcomes reaches your stack as degraded service, new pricing, or an unplanned migration. |
| The sell-side exception supports diversification, not complacency | The market rewarding PubMatic and Magnite is a point in favour of the independent supply layer — and a reason to maintain multiple healthy paths through it rather than consolidating to one. |
| Expectations become budgets | Buy-side planners read the same forecasts investors do. If the street believes concentration is the future, agency dollars will drift the same direction unless your inventory gives them a documented reason not to. |
04What publishers should do
05The bottom line
The scoreboard was never about last quarter, and neither should your planning be. Wall Street spent the week betting that open-web ad tech loses relative share to three walled gardens, and while the bet may prove too pessimistic, the companies on the wrong side of it will start behaving as if it is true — cutting, repricing, consolidating. Publishers cannot control that. What they can control is knowing exactly which of those companies their revenue runs through, having a rehearsed answer for each one, and building the revenue lines that no concentration forecast can touch.