On Monday, Pinterest made a change that reads like plumbing and lands like an org chart. Its display and video inventory can now be bought from inside the same demand-side platforms big advertisers already use for streaming TV, open web display and retail media. For years, if you wanted Pinterest, somebody on your team logged into Pinterest. Now a trader in DV360 or Amazon’s DVA+ builder can add it as one more line item without ever opening Pinterest’s own ad tools.
That is good news if Pinterest has been stuck in the “test it someday” column of your plan. It is a mess waiting to happen if two people in your building, or at your agencies, now think Pinterest is theirs. And it is landing in the last week of October, right when holiday plans stop being plans and start being spend.
What Pinterest switched on
The Pinterest newsroom post on October 26 announced general availability of Programmatic Deals in North America, South America and Asia Pacific. Buyers get “selected Pinterest display and video inventory” through outside partners. Magnite is the primary supply-side platform. In Latin America and Asia Pacific, Google Ad Manager also handles supply. On the buying side, Pinterest names “Amazon Ads DVA+, Google Display and Video 360, Yahoo DSP and DeepIntent,” with the word “including” in front, so the list is not closed.
Adweek reported that buyers can transact through private marketplaces and preferred deals, that Pinterest is testing programmatic guaranteed, and that the setup has been in testing with agencies and brands since early this year. Marketing Dive confirms the same partners and regions.
Now the missing pieces. PPC Land’s read of the launch lists them plainly: no prices, no floors, no minimums, no description of which placements count as “selected,” no country list, no date for Europe, and no explanation of how conversions from DSP-bought impressions get counted. It is not self-serve either. Pinterest’s post ends by telling advertisers to contact their Pinterest sales representative.
So you know where the door is. You do not yet know what is behind it or what it costs to walk through.
Why Pinterest wants a second door
Pinterest is not shy about the motive. The same newsroom post says the product gives enterprise advertisers “another way to include Pinterest in broader media plans” and, in the same sentence, describes it as “broadening potential demand for its inventory.” Jason Fairchild, who runs programmatic and affiliate at Pinterest, told MediaPost the goal is to “broaden sources of demand” while “complementing our first-party ads business.”
Read that as a seller would. More buyers competing for the same supply is how a marketplace lifts price. The context backs it up. Pinterest guided third quarter revenue growth to 13% to 15%, down from 18% in the second quarter, according to both Adweek and eMarketer. PPC Land adds that in the second quarter, Pinterest’s ad pricing rose only 1% while impressions grew 16%. Pinterest reports third quarter results Thursday, October 29.
None of that makes the inventory worse. It tells you a platform that needs demand is a platform that will negotiate.
Two doors, one house, two keyholders
Here is the part that matters to anyone who owns a media plan.
Before Monday, Pinterest had one natural owner in most companies: whoever ran paid social, usually in house or at a social agency, working in Pinterest’s own Ads Manager with the Pinterest tag, Pinterest’s reporting and Pinterest’s frequency counting. The programmatic desk, in house or at a holding company, mostly had no clean way in.
After Monday, both of them can buy Pinterest. They will do it in different tools, with different audiences, under different reporting, on different invoices. Pinterest has not said whether its own frequency counting sees impressions bought through a DSP, or whether DSP frequency caps know about the Ads Manager campaign running next door. Until it says so, assume nobody is counting both, because nobody has told you they are.
That creates three very ordinary problems.
You can end up bidding against yourself. If your social team and your programmatic desk both chase the same holiday shoppers, and both routes draw on the same supply, as Pinterest’s “broadening demand” language suggests, you are two buyers for one pool. Best case, that is waste. Worst case, your own teams push up the price you pay for the same people.
Your reporting splits in two. Ads Manager will show you Pinterest’s version of the results for its campaigns. The DSP will show its version for the deals. They will not use the same attribution rules, and Pinterest has not said whether its tag or Conversions API data applies to DSP-bought impressions at all. Your weekly deck will have two Pinterest numbers that do not reconcile, and somebody will add them anyway.
Nobody owns the frequency. The pitch for the DSP route is that Pinterest can finally sit under one frequency cap with your streaming, display and retail media. That is a real benefit, but only for the share you buy through that DSP. Whatever runs in Ads Manager sits outside the cap.
eMarketer put the risk well: “A disjointed strategy, in which Pinterest is operating as an isolated channel, could lead to higher reach but little ability to turn it into business outcomes.” The new failure mode is subtler. Pinterest is no longer isolated. It is duplicated.
What the toll costs
A buy in Pinterest’s own tools has one company between you and the ad slot. A buy through a DSP and an SSP adds two technology layers, each taking a cut. Neither Pinterest nor Magnite has disclosed the take rate on these deals.
The DSP you pick changes the math. PPC Land, citing Needham analysis reported by The Wall Street Journal in 2025, put Amazon’s DSP fees at about 1% against 12% to 15% at The Trade Desk. The Trade Desk is not on Pinterest’s launch list, so for now you compare the doors Pinterest actually named.
The Amazon door has its own wrinkle. Pinterest names DVA+, not Amazon DSP. PPC Land reported on September 29 that DVA+ is a campaign builder inside Amazon Ads Agent, then in an extended closed beta in the US, with an open beta expected in late October and general availability across 35 markets in the first quarter of 2027. Pinterest does not say whether conventional Amazon DSP line items can target the deals. If your Amazon trader is not in that beta, the Amazon door may be painted on the wall for now.
Then there is deal type. Private marketplaces and preferred deals behave differently. A private marketplace lets invited buyers bid above a floor. A preferred deal gives you a first look at an agreed price without guaranteed volume. Adweek says programmatic guaranteed, the version with fixed volume and fixed price, is still being tested. For holiday weeks, that difference decides whether you get the impressions at all.
Read the case study like a buyer
Pinterest’s proof point is a June 2026 Moroccanoil campaign that used a deal enabled by Magnite and bought via Amazon Ads to add Pinterest to an existing video and display campaign. Pinterest reports 15% incremental reach with 1% audience overlap, 63% more new-to-brand buyers among users whose purchase paths included Pinterest, and a 20% higher click-through rate than the brand’s average video and display campaigns.
Those are encouraging numbers, and they are worth exactly what their footnote allows. PPC Land notes the source is Amazon internal data for June 1 to 30, 2026, that Pinterest describes the results as specific to the campaign and not indicative of future performance, that the 63% figure has no stated baseline, that “overlap” is not defined, and that the click-through comparison covers the whole campaign against the brand’s own average, not Pinterest placements against a control.
Take the shape of the finding, not the size. Pinterest audiences may not overlap much with the people you already reach on streaming and display. That is a hypothesis you can test cheaply, not a forecast for a finance model.
One sentence worth asking about
Buried in the announcement is a quote from Sean Buckley at Magnite: “In addition to connecting DSPs to Pinterest’s on-site inventory, Magnite can help extend the value of its intent signals across the open web.”
That describes something bigger than selling Pinterest ad slots. It points at Pinterest-derived intent informing buys on other publishers’ inventory. Neither company has explained what data would leave Pinterest, in what form, under what consent terms, or whether any of it is live. Pinterest’s inventory is also part of Magnite Orchestration, a layer where buyer and seller software agents can discover and package inventory. Pinterest gives no detail on whether any of its inventory has been bought that way.
If you have a privacy team, that sentence belongs in front of them before anyone books a deal that depends on it.
Who should care most
Not everyone needs to move this week.
If you already buy Pinterest well in Ads Manager, with your own shopping catalog, creative made for the feed and a measurement setup you trust, the DSP route is a second option, not an upgrade. Pinterest itself calls it complementary. The main risk for you is someone else in your organization opening the second door without telling you.
If your planning runs almost entirely through a DSP and Pinterest has never made the cut because it meant another login, another contract and another report, this is the first real reason to test it. The cost of activation just dropped.
If you are an agency running both a social team and a trading desk for the same client, you have the biggest homework. You are now the most likely party to put two bids on the same person.
What to do this week
Name one owner for Pinterest. One person, across in house and every agency, decides which door each Pinterest dollar goes through. Write it down before November 1. This is the single step that prevents most of the problems above.
Split by job, not by team. A clean default: Ads Manager keeps shopping, catalog and conversion campaigns where the Pinterest tag and catalog matter. The DSP route gets reach and video that needs to sit under a cross channel frequency cap. If both teams want the same holiday audience, the owner picks one.
Ask the rep five questions in writing. Which placements are in “selected” inventory. What deal types and floors are on offer for November and December. Whether Pinterest frequency and conversion measurement see DSP-bought impressions. What the combined supply-side and DSP take looks like on your chosen route. What data, if any, leaves Pinterest under Magnite’s open web language.
Price the route before the test. Get the fee stack in writing and compare effective cost with what you pay in Ads Manager. Vague answers are your answer for Q4.
Run one small test with a clean read. If you have never bought Pinterest, add one Pinterest deal to an existing DSP campaign and measure it against the same incrementality bar you use for other discovery channels. Do not compare it with the Moroccanoil numbers. Compare it with your own holdout.
Check what your Amazon trader can actually buy. If DVA+ is your intended door, confirm access before anyone promises the deal in a holiday plan.
Watch Thursday’s earnings. Pinterest reports on October 29. Listen for how much of its growth story leans on programmatic. A seller that needs this route to work is a seller with room to negotiate.
Pinterest has spent years arguing it deserves a line in the main media plan. On Monday it made that easy. Making it easy for two of your teams at once is the part you have to fix yourself.
