HOT OFFERINGS
| Getting your Trinity Audio player ready... |
In the news
On July 17, Microsoft Advertising announced it will shut down its DSP, Microsoft Invest (formerly Xandr), by February 2026. Officially, the move is about aligning with “a more private and personalized advertising experience for a conversational and agentic world.”
Unofficially? It’s about shedding a clunky asset that never quite kept up.
Instead of keeping Invest on life support, Microsoft will now funnel its buy-side business into a streamlined, AI-assisted self-serve tool: the Microsoft Advertising Platform. Think less like DV360, more like a Performance Max clone with Copilot whispering in your ear.
Meanwhile, third-party DSPs will still be able to access Microsoft’s inventory. On the sell side, the company says it remains committed to Microsoft Monetize and Microsoft Curate—tools that serve publishers and control premium supply.
Invest’s lineage is notable: born as AppNexus, swallowed by AT&T, rebranded as Xandr, then offloaded to Microsoft in 2021. Now, it ends its run as another casualty of the AI consolidation wave.
Layoffs accompany the closure. The industry’s reaction? Mostly: “Saw it coming.”
The Xenoss take
Let’s not kid ourselves—this wasn’t just about “agentic futures” or a philosophical shift in UI metaphors. Microsoft’s decision to kill off Invest is, first and foremost, a cleanup act.
The DSP never truly recovered from the AT&T years. AppNexus was once a kingmaker. But after years of underinvestment and strategic whiplash, it fell behind. Microsoft gave it a new lease on life—but not a new direction. This was the slowest car on a crowded freeway, and eventually, someone had to pull over.
So they did.
But what’s interesting isn’t what’s dying—it’s what’s being built in its place.
AI eats the DSP
Microsoft is going all-in on first-party demand tools—faster, cheaper, and easier to control. AI is the new buying assistant. Not in the “help me choose a CPM” sense, but in the “I’ll take care of everything” sense.
If that sounds familiar, it should. Google’s Performance Max, Meta’s Advantage+, and Amazon’s growing suite of “just trust us” tools have been singing that tune for years. The selling point is simple: better outcomes, less effort. The cost? Visibility, control, and any illusion of a neutral marketplace.
Microsoft is just the latest to lean in. Invest’s shutoff clears the deck for AI-powered, walled-garden media buying where Microsoft controls the knobs—and the margins.
A shift in power, not philosophy
The blog post talks a big game about privacy and personalization, but let’s call it what it is: a business model pivot.
The money’s not in being a neutral DSP. The money’s in owning demand, supply, and the data pipes in between. Especially if you’ve got LinkedIn–level identity data to play with.
And therein lies the rub. Microsoft can talk privacy all day, but LinkedIn is a logged-in behavioral treasure trove. It’s not that they’re backing away from data. They’re just walling it in.
Invest didn’t have a place in that strategy. It didn’t generate enough revenue, didn’t have a strong differentiator, and worst of all, didn’t fit the narrative of a future shaped by Copilot-style automation.
What this means for the rest of the market
In the short term, The Trade Desk, Yahoo, DV360, and others will scoop up the displaced enterprise clients. Expect a flurry of “we’ll make your transition easy” emails by Q4.
Longer term, though, the DSP category faces existential questions. If every major platform is shifting toward AI-native, closed-loop systems, where does that leave the open web?
The answer might be: in the same place display has been for years—fragmented, under-monetized, and losing attention to platforms that offer simpler paths to performance.
For publishers, a small silver lining
Microsoft isn’t walking away from the supply side. Curate and Monetize will still be in play. But let’s be honest—those tools now function less like open-market facilitators and more like wrappers for Microsoft’s broader data and demand strategy.
In this light, Invest wasn’t sunset because the DSP model is broken. It was sunset because Microsoft’s DSP wasn’t big enough, sticky enough, or profitable enough to justify the overhead.
Final thought
Microsoft didn’t kill Invest because DSPs are obsolete. It killed it because generic DSPs without differentiated data or scale have no future in a market sprinting toward AI-powered, vertically integrated stacks.
This is less about the DSP model failing and more about Microsoft’s specific DSP failing to compete. Invest never became essential to buyers. It didn’t offer exclusive data, couldn’t compete with the firepower of The Trade Desk or DV360, and wasn’t sticky like Amazon Ads or Meta’s Advantage+ tools.
In an era where control of identity, first-party data, and proprietary media environments is the new holy trinity, running a standalone DSP with none of those advantages is a losing game. Microsoft knows where its strengths lie: LinkedIn, Bing Search, Edge, Outlook. Logged-in users. Authenticated data. AI-powered ad automation. That’s where it can control the rails, extract more margin, and create performance without having to explain the sausage-making.
So no—DSPs aren’t dead. But the “me-too” DSP is. If you can’t bring exclusive demand, differentiated data, or a smarter way to buy, you’re just middle tech—and middle tech gets squeezed.
