Pinterest is now in your DSP. Buy it for reach, and price it per new person reached

KEY TAKEAWAYS

1. The DSP route is a reach and frequency purchase. Search, catalog and conversion bidding stay in Pinterest Ads Manager.
2. The Moroccanoil case shows Pinterest added people to a plan. It does not show those people bought because of it.
3. Price a deal per 1,000 incremental people reached, after fees, not per 1,000 impressions.
4. Keep a deal only if a holdout shows outcomes in your own data at a cost you would pay elsewhere.

On 26 October 2026 Pinterest made Programmatic Deals generally available. Advertisers can now buy selected Pinterest display and video inventory inside the buying platforms they already use, next to the rest of a programmatic plan.

Most coverage asks what Pinterest left out: prices, placements, countries, the signals that travel with each bid. Those gaps are real. The more useful question is what job this route is for. This piece argues it is a reach purchase, sets out what stays in Pinterest's own Ads Manager, and gives a way to price a deal per new person reached and then test it.

What did Pinterest open on 26 October?

The reported facts, pieced together from Pinterest's announcement as covered by PPC Land, Marketing Dive and Social Media Today:

• Inventory. Selected Pinterest display and video inventory. Pinterest has not listed the placements.
• Supply. Magnite is the primary supply-side partner. Google Ad Manager also handles supply in Latin America and Asia Pacific.
• Buying platforms. Amazon Ads DVA+, Google Display and Video 360, Yahoo DSP and DeepIntent. The Trade Desk is not named.
• Deal types. Private marketplace and preferred deals are live. Programmatic guaranteed is in testing, per Adweek, as cited by Relevant Audience.
• Regions. North America, South America and Asia Pacific. Europe is excluded, with no reason or date.
• Access. Through a Pinterest sales representative. There is no self-serve path.

One background point matters for the Amazon route. DVA+ is Amazon's new four-step display, video and audio builder, announced on 29 September 2026 and in extended closed beta in the US. Deals, frequency management and custom bidding sit in its advanced settings.

What does a preferred deal or private auction actually buy?

Neither type guarantees delivery. Google's Ad Manager help, which covers the supply side Pinterest uses in Latin America and Asia Pacific, sets out how each works.

A preferred deal gives the buyer a first chance to bid at a negotiated fixed price. If the buyer passes, the impression can fall through to a private auction or the open auction. A private auction sets a floor. Invited buyers must bid at least that much to compete. On the buy side, Display and Video 360 recommends one deal per line item.

The practical gain is not access alone. It is that a Pinterest line can sit inside the same insertion order as your other lines. Display and Video 360 applies frequency caps across all inventory types under a line item, insertion order or campaign. Two limits apply. The help page describes caps across apps on one device and does not say they follow a person across devices. And on connected TV, impressions with no known device identifier are left out of capped lines. Within those limits, Pinterest impressions count against the same cap as the rest of the plan, and reach can be read across lines together. Ads Manager cannot do that for impressions bought elsewhere.

What stays in Pinterest Ads Manager?

Pinterest's own conversion tools are documented as Ads Manager features, not as part of the deals:

• Performance+. Automated bidding toward conversion events you choose, for Consideration, Leads and Sales campaigns. It offers keyword and interest suggestions, customer lists and catalog creative. Pinterest recommends its tag plus the Conversions API, a budget around five times your target cost per acquisition, and 50 or more conversion events a week for Leads and Sales.
• Visual Search Ads. Announced on 17 September 2026, in beta. They run in search results and Pin closeups and work with Performance+. The announcement does not mention programmatic buying.

Pinterest's leadership points the same way. On the second-quarter 2026 earnings call, Bill Ready said about 30% of lower-funnel revenue ran through Performance+ as of the first quarter. Adopters grew lower-funnel spend more than twice as fast as non-adopters. Pinterest is also piloting links between advertisers' own measurement systems and its bidding. And it plans to fold tvScientific, its connected TV buying platform, into Performance+ in 2027.

So the performance machinery is being built inside Ads Manager. The deals sell inventory and undefined "audience signals" to outside bidders. That is the main tension in the launch, and the evidence settles it by job, not by winner. Use the deal for reach you plan and cap across channels. Use Ads Manager for search intent, catalogs and bidding on conversions.

What does the Moroccanoil case actually show?

Pinterest's only published result is a Moroccanoil campaign run from 1 to 30 June 2026. It used a Magnite deal bought through Amazon Ads and added Pinterest to a campaign already running. ContentGrip reports it combined Amazon first-party audiences with Pinterest contextual beauty targeting. The figures come from Amazon's internal data:

• 15% incremental reach, with 1% audience overlap.
• 63% more new-to-brand buyers among users whose purchase paths included Pinterest.
• Click-through rate 20% above the brand's average video and display campaigns.

No spend, baseline or method was published. PPC Land, ContentGrip and MarTech Edition all flag this.

The reach figure is the strongest of the three, and it can be unpacked. Suppose the 1% means 1% of people reached on Pinterest were also reached by the rest of the campaign. Pinterest has not defined it, so this is an assumption. Then the Pinterest line reached about 15.2% as many people as the base campaign (0.15 divided by 0.99). That is our arithmetic, not Pinterest's.

The 63% figure is weaker. It compares buyers whose paths included Pinterest with those whose paths did not. That is an exposed versus unexposed comparison. In Gordon and colleagues' 2016 analysis of 12 randomized Facebook lift studies, such comparisons overstated lift badly. In one study a true lift of 77% came out at 416%. "New to brand" also has a narrow meaning on Amazon. It counts ad-attributed purchases from shoppers who had not bought the brand on Amazon in the previous 12 months, per Adbrew's 27 October guide. It is a count of attributed orders, not a measure of lift.

The click-through rate is compared with a brand-wide average, not a matched line. It says little about a new channel.

Does 1% overlap mean Pinterest users are a separate audience?

No, and this is where the case and survey data seem to disagree. GWI data reported by DataReportal says the typical social media user visits about 6.66 platforms a month. Pinterest's user base cannot be cut off from everyone else's.

Both can be true. Overlap between platforms' users is high. Overlap between the people one campaign reaches in one month is small, because each line reaches only a slice of each platform. So a low overlap figure is partly mechanical. It shows the deal added different people to this plan. It does not show Pinterest users are unusual, or that reaching them changed what they bought.

Which route fits which job?

How do you price a deal per new person? A hypothetical example

This example is invented to show the method. No real advertiser or price is involved.

A brand runs a $200,000 video and display plan in its DSP. It has $30,000 more to place and two options.

Option A. A Pinterest deal at an effective $12 CPM. That is the deal price plus DSP and data fees. The $30,000 buys 2,500,000 impressions. At an average frequency of 3, that is about 833,000 people. If 20% were already reached by the base plan, about 667,000 are new. Cost per 1,000 new people is about $45.

Option B. More spend on an existing open-web line at $8. The $30,000 buys 3,750,000 impressions, or 1,250,000 people at the same frequency. But this line already serves the same audience, so assume 70% overlap. That leaves 375,000 new people, about $80 per 1,000.

The cheaper CPM loses on the measure that matters for a reach buy. Fees decide more than they seem to. ANA's second-quarter 2026 Programmatic Transparency Benchmark, built from impression-level log data, put transaction costs at 27.2% of programmatic spend. Compare routes on the price you actually pay, not the deal floor.

New people are still not new customers. The example earns the deal a test, not a renewal.

How should you test a Pinterest deal? Seven steps

1. Get the terms in writing. Countries, deal type, floor or fixed price, placements, the signals attached, and whether the deal ID shows in your seat.
2. Work out the effective CPM. Add DSP, data and verification fees to the deal price.
3. Build it inside the existing plan. One deal per line item, in the same insertion order or campaign as the lines it should extend, so caps and reach reports cover both. For how browser identity loss can blur those reach counts, see our Safari blocklist line-item audit.
4. Read reach after two weeks. Pull unique reach and overlap from the DSP. Compute cost per 1,000 incremental people and compare it with your next-best line.
5. Hold out. Split the audience at random, or run regions with and without the Pinterest line. Our guide to choosing a geo test covers sizing and cooldown.
6. Judge on your own data. Count orders, new customers or qualified leads in your store or CRM. Do not compare platform figures with different windows. Pinterest's default, when none is chosen, is 30-day click and 1-day view. Amazon's long-horizon metrics bring their own problems, covered in our piece on Amazon's Long-Term ROAS.
7. Decide. Keep the deal if the cost per incremental outcome beats your marginal channel. If it only wins on reach, keep it for reach goals and budget it there.

What is still unknown?

Pinterest has not published prices, floors, minimums, placements, a country list or how its audience signals are passed. Programmatic guaranteed has no date. Europe has no date. Pinterest's third-quarter results call, scheduled for 29 October 2026, may add detail. Until the terms are public, treat each deal as a test with a written exit rule.

We track how brands appear across paid and AI surfaces as part of our early access programme. Questions go to esh@ethosengine.io.

Questions

What are Pinterest Programmatic Deals?

A way to buy selected Pinterest display and video inventory through outside buying platforms. Pinterest made them generally available on 26 October 2026 in North America, South America and Asia Pacific. Magnite is the main supply partner, with Google Ad Manager also used in Latin America and Asia Pacific. Access goes through a Pinterest sales representative.

Which DSPs can buy Pinterest inventory?

Pinterest named Amazon Ads DVA+, Google Display and Video 360, Yahoo DSP and DeepIntent, and introduced the list with the word including. The Trade Desk was not named. Private marketplace and preferred deals are live, and programmatic guaranteed deals are still in testing, according to Adweek.

Can I buy Pinterest Programmatic Deals in Europe?

Not yet. Europe is not among the launch regions and Pinterest gave no date or reason. Advertisers in Europe still buy Pinterest through Ads Manager. Country lists inside the three launch regions were not published either, so confirm your market with your representative.

Does a DSP deal include Pinterest search ads and Performance+?

Pinterest has not said so. The launch describes selected display and video inventory. Performance+ and Visual Search Ads are documented as Ads Manager features, and Pinterest says it will fold tvScientific into Performance+ in 2027. Ask which placements and signals your deal carries before you plan around them.

Sources

1. PPC Land, Pinterest sells ads through Amazon and Google DSPs in 3 regions, not Europe ppc.land
2. Marketing Dive, Pinterest strengthens bid to media buyers with programmatic solution marketingdive.com
3. Relevant Audience, Pinterest Programmatic Deals now in DV360 and Amazon DSP relevantaudience.com
4. Social Media Today, Pinterest introduces new programmatic buying solution socialmediatoday.com
5. ContentGrip, Pinterest programmatic deals expand through Magnite contentgrip.com
6. MarTech Edition, Pinterest makes Programmatic Deals available through Magnite and four DSPs martechedition.com
7. Pinterest, Q2 2026 earnings call transcript s204.q4cdn.com
8. Pinterest Newsroom, Pinterest unveils Visual Search Ads newsroom.pinterest.com
9. Pinterest Business Help, Pinterest Performance+ help.pinterest.com
10. Pinterest Business Help, Review Conversion Insights help.pinterest.com
11. Google Ad Manager Help, Preferred Deals overview support.google.com
12. Google Ad Manager Help, Private Auctions overview support.google.com
13. Display and Video 360 Help, Purchase private inventory support.google.com
14. Display and Video 360 Help, Manage frequency with frequency caps support.google.com
15. Amazon Ads, DVA+ display, video and audio campaigns launch announcement advertising.amazon.com
16. Adbrew, How to use Amazon new-to-brand metrics in your ad strategy adbrew.io
17. Fiducia, ANA Programmatic Transparency Benchmark Q2 2026 fiducia.eco
18. DataReportal, Global social media statistics datareportal.com
19. Gordon, Zettelmeyer, Bhargava and Chapsky, A comparison of approaches to advertising measurement (white paper) kellogg.northwestern.edu

NEXT STEP

See where your brand stands.

Your Digital Authority Score, who AI recommends for your target questions, and the placements that close each gap. Your first report is free.

Get your free report →