Pinterest Acquires tvScientific, Enters CTV Advertising Market

Pinterest Buys tvScientific, Steps Into CTV

Pinterest acquired tvScientific, a mid-market connected-TV ad platform (one that buys and serves ads on streaming television), back in February and is now letting advertisers run TV-screen video campaigns directly through Pinterest. It also produced a shoppable show on Roku with QR codes that open Pinterest boards. Wayfair and Michaels showed up for the experiment, which is the right category fit: home, decor, craft. Nobody is scanning a QR code to buy insurance.

The structural question is whether Pinterest's purchase-intent data turns into a real premium on CTV inventory, or whether this is just another way to buy the same open-market supply everyone else accesses. tvScientific never owned premium inventory. It sourced from the open auction like any other buyer. Pinterest brings targeting and a promise, no exclusive TV shows, and that promise has not been tested with real advertiser money at scale yet.

On the buy side, the immediate friction lands on media planners. A Pinterest CTV line sitting next to a Pinterest social line creates a de-duplication problem the platform has never had to answer for. If the same person sees the social ad and the CTV ad, does Pinterest count that as one reach or two? Brand safety on CTV inventory bought through a company known for mood boards is a second question. These are solvable, but early adopters will find the seams before Pinterest patches them.

For Magnite and PubMatic (the independent SSPs that run the open CTV auction), every dollar Pinterest keeps inside its own stack is a dollar that never hits the open market. This is the same pattern Meta and Amazon ran before Pinterest: platforms with a login and a purchase history corralling TV ad money into their own storefronts. The SSPs are underpricing how fast this erodes their demand pool.

Roku is both partner and rival here. A branded show on Roku today becomes a competing demand channel tomorrow. That relationship has a shelf life.

The whole bet rests on two things. First, whether advertisers will pay a premium for Pinterest-targeted CTV over a generic open-market buy. Second, whether Pinterest can tie a TV impression to a downstream purchase in a way buyers trust. Against Amazon and Walmart, which have actual transaction data, and Roku and Netflix, which own the screens, Pinterest is squeezed on both sides.

Our call: Pinterest will not report CTV as a distinct revenue driver by its Q3 2026 earnings call (early November 2026). The conversion proof takes longer than eight months, de-duplication and measurement plumbing are still being stress-tested, and agency test cycles do not move real budget onto an unproven loop that fast. Medium confidence. A data-fusion win could surprise, but the structural mismatch between Pinterest's browse behavior and lean-back TV viewing is real.

Pinterest bought a mid-market connected-TV ad platform called tvScientific back in February, and now lets advertisers run TV-screen video campaigns straight through Pinterest. It also made a shoppable show on Roku with QR codes. The real question for the rest of us: does a platform with purchase-intent data but no video habit actually change how CTV gets bought, or is this another walled garden trying to skim the open market?

This is easy to undo for buyers. A media planner can test a Pinterest CTV line for a quarter and walk away. It is hard to undo for Pinterest, which now owns a DSP and has to run it. What's being decided is not "does Pinterest get into CTV." It's whether first-party intent data becomes the price of entry to CTV demand, squeezing the infrastructure players who sell pipes. No hard deadline, but the second season of the Roku show lands in 2027, and Q3 and Q4 numbers will tell buyers whether the loop closes.

The Market Analyst. Add Pinterest to the list of platforms buying their way into CTV with a data moat instead of building infrastructure: Meta, Amazon, and now Pinterest. In plain terms, the companies with a login and a purchase history are corralling TV ad money into their own storefronts. For Magnite and PubMatic, whose business is running the open TV auction where anyone can bid, every dollar Pinterest keeps inside its own stack is a dollar that never hits the auction. That's the slow bleed. tvScientific's old rivals in mid-market CTV buying lose an independent player and gain a competitor with 500 million users behind it. Roku is both partner and rival here, and that gets awkward fast.

The Skeptic. Pinterest is a place people browse for inspiration. Nobody leans back on the couch to watch Pinterest. That's a real mismatch with how TV gets consumed, and no acquisition fixes it. tvScientific was a decent mid-market buyer, but it never owned premium inventory. It sourced supply from the open market like everyone else. So Pinterest is not bringing exclusive TV shows to the table. It's bringing targeting and a promise. The whole bet rests on advertisers believing Pinterest's interest data improves CTV targeting more than The Trade Desk or Amazon already do. That claim has not been proven with anyone's money yet. The Roku show is a marketing stunt. Clever, but a stunt.

The Operator. Tuesday morning, a media planner gets a new line item: Pinterest CTV, sitting next to Pinterest social. First thing that breaks is de-duplication. If the same person sees the social ad and the CTV ad, does Pinterest count that as one reach or two? Trafficking teams have to validate that across both legs of a single campaign, and Pinterest has never had to answer for TV-screen reach before. Then there's brand safety on CTV inventory bought through a company known for mood boards. At 90 days, the pressure lands on Pinterest's ad ops and partner support, because early adopters will stress-test the standalone campaign screen and find the seams. The measurement question is the whole game: can Pinterest tie a TV impression to a downstream purchase? If it can't, this is just another place to buy the same open-market inventory.

The Customer / End User. Two customers here. The advertiser and the person on the couch. For the advertiser, this is optional, low-risk, and mildly interesting: a full-funnel buy where the upper-funnel TV ad and the lower-funnel shoppable pin live under one roof. Wayfair and Michaels showed up for the Roku experiment for a reason. For the person watching, a QR code on the screen that opens a shoppable board is a genuinely new thing, and it might work for exactly the categories Pinterest is strong in: home, decor, craft. Nobody is scanning a code to buy insurance. The narrow use case is the point, and also the limit.

The CFO. The real cost here is not the tvScientific price tag. It's the margin structure Pinterest just signed up for. Running a DSP that sources supply from the open market means Pinterest competes on price and targeting, with no exclusive inventory to protect a premium. That is a thin-margin business unless the intent data commands a real premium. So the payback depends entirely on Pinterest charging more per TV impression than a generic buyer, and sustaining it. Against Amazon and Walmart, who have actual transaction data, and Roku and Netflix, who own the screens, Pinterest is squeezed on both sides. The conversion-proof side and the inventory side.

Where the council splits

The real disagreement is whether Pinterest's intent data is a moat or a talking point. The Strategist view in the briefing treats the shoppable-intent loop as a rare asset nobody else can copy. The Skeptic and the CFO say it's unproven and structurally thin, because Pinterest still buys the same inventory as everyone else and now has to charge a premium it hasn't earned.

Second split: the Market Analyst sees this as one more brick in the wall of first-party-data platforms starving the open CTV auction. The Operator says none of that matters until Pinterest can prove a TV impression drove a sale, and that measurement question is unsettled.

Third, quieter tension: Roku as partner versus Roku as rival. A branded show on Roku today, a competing demand channel tomorrow. That relationship has a shelf life.

What it actually hinges on

Two things. First, whether advertisers will pay a premium for Pinterest-targeted CTV over a generic open-market buy. Second, whether Pinterest can close the loop from TV impression to purchase in a way buyers trust. If both are yes, Pinterest owns a real slice of shoppable CTV and the open-market SSPs feel it. If either is no, this is a modest, optional line item that looks better in the press release than in the revenue mix.

The council leans skeptical on the near term and open on the structural signal. The bigger story is not Pinterest specifically. It's that CTV demand keeps migrating into platforms with their own login and their own data, and the pure infrastructure players who run the auction are the ones who pay for it over time.

What to verify before betting on this: watch whether Pinterest reports any CTV-attributed conversion lift, and whether agencies keep the line item past the first test quarter. If planners renew, the loop is closing. If they don't, it isn't.

Prediction: Pinterest will not report CTV as a distinct revenue driver by its Q3 2026 earnings call (early November 2026), and the CTV push will remain a small add-on to social/search ad revenue through that print.

Confidence: Medium. The mismatch between Pinterest's browse behavior and TV viewing is real, but a data-fusion win could surprise.

Why: tvScientific never owned premium inventory, so Pinterest is buying the same open-market CTV supply as everyone else and competing on targeting alone, which means it has to prove its interest data commands a premium before the dollars scale. That proof takes more than the eight or nine months between the February acquisition and the Q3 print, especially with de-duplication and measurement plumbing still being stress-tested by early adopters. The opposite outcome, Pinterest calling out CTV as a meaningful growth line this soon, would require advertisers to have already moved real budget on an unproven conversion loop, which is not how agency test cycles work.

Revisit by 2026-11-15: We're right if Pinterest's Q3 2026 earnings materials and call treat CTV as an emerging capability without breaking out or crediting it as a revenue driver. We're wrong if Pinterest names CTV as a material contributor to Q3 revenue growth or reports specific advertiser conversion results from standalone CTV campaigns.

One more thing worth saying plainly. The interesting call here is not about Pinterest. It's that the open CTV auction keeps losing demand to platforms with a login, and the SSPs running that auction are underpricing how fast it erodes.