Xandr and Yahoo DSP buying, with an exit plan

Yahoo DSP and Yahoo Japan's display advertising are different companies, different stacks and different money. Yahoo DSP is the former Verizon Media platform, bought as a seat, with Yahoo's owned properties and native formats behind it. Yahoo Japan display, YDA, belongs to LY Corporation, formed when LINE and Yahoo Japan merged in October 2023, and it shares no seat, no identifier and no reporting with the other one. Something close to half the Japan briefs we receive naming Yahoo mean the second while carrying a plan written for the first. Sorting that out is the first hour of this engagement, and we do not charge for it. Why either seat is worth holding: The reason is supply, not technology. Xandr Invest reaches Microsoft's owned inventory, including MSN, Outlook.com and the Edge new tab, at rates no resold path matches, and that audience skews older, more desktop and more work-context than anything you will find on the social platforms. Yahoo DSP's case is its own mail, finance and sports properties plus native placements that behave nothing like standard display and need their own creative and their own benchmarks. If neither of those descriptions matches the people you are trying to reach, do not open the seat. Platform risk, priced in: Xandr moved from AT&T to Microsoft in 2022, and both platforms in this tier have had roadmaps trimmed since. We assume any of them can be restructured or retired inside eighteen months and we build accordingly: deal IDs, audience definitions, creative and reporting stay portable, the measurement layer lives in your warehouse rather than the platform's interface, and the migration plan is written before the first flight rather than during the announcement. When a vendor tells you to build measurement inside their UI, what is being sold is switching cost. How we buy: Owned-and-operated supply through direct deals rather than open exchange, because the whole reason for the seat is inventory you cannot get elsewhere and the open exchange here is the same inventory you can already reach. Frequency managed across seats, not within one, which means the cap lives in the measurement layer and the platforms are told what is left. Creative built for native separately, never a display banner resized into a native slot. Incrementality tested with a geo holdout, because attribution on this supply is weak enough that the platform's number is not worth arguing over. Not included: Microsoft Advertising search on Bing, a different discipline with a different team. Yahoo Japan YDA and LINE Ads, which are separate engagements quoted separately; we will not fold them into a DSP retainer to make the invoice look tidy. Ad server work. Creative production, though we specify the native units and review what comes back. Not for you if: You want either of these as a primary seat. You want an always-on retainer where two seasonal owned-and-operated buys would do the same job for a fraction of the money, which is the honest answer for most advertisers that come to us for this. You need a written guarantee about a platform's roadmap, which nobody can give you. Or the audience you want is mobile, young and social, in which case this supply is precisely wrong and we will say so before you sign anything.