Pinterest, Inc. PINS

Key data

The business

Pinterest is a visual search app. People save images of things they want to make, buy, or do, boots, recipes, nurseries, wedding tables, and sort them onto boards. Because they arrive planning a purchase rather than scrolling to socialize, the audience carries commercial intent, and intent is what Pinterest sells: shoppable ads placed against a user who is already leaning toward buying. That is the moat, a first-party record of what half a billion people plan to spend money on, built for free by the users themselves and hard to copy without them.

The catch is geography. Most users sit outside the United States, yet no country other than the US contributes even a tenth of revenue. America is the engine, the world is the option. In February 2026 Pinterest bought tvScientific, a connected-TV ad platform, extending its ads into streaming and pulling in third-party ad demand to help close that gap.

The numbers

Five quarters show the problem and the answer at once.

Strong top line, ugly bottom line. Revenue grew 18% year over year in Q2 2026, the fastest in over a year, and yet both 2026 quarters booked a loss. That loss is not the business breaking. It is stock-based compensation running at 28% of revenue, set against the absence of 2024's one-time tax gain. Q4 is the tell: the seasonally strongest quarter earned real money. The adjusted profit the market tracks came in above expectations in three of the last four quarters.

Revenue compounded about 13% a year since 2021 and reaccelerated to 16% in 2025. Ignore the $1.86 billion of 2024 net income, it is a tax artifact from releasing a valuation allowance, not cash that came in the door. The honest earnings power lives in the cash statement.

Pinterest turns about 36 cents of every revenue dollar into operating cash and spends almost nothing to do it, capital spending runs about 1.5% of revenue, a rounding error next to the data-center bills of larger platforms. That left roughly $1.6 billion of free cash over the past year against a $13 billion market value. Management is spending it retiring stock, $1.95 billion in the first half of 2026 alone through an accelerated repurchase, cutting the diluted share count from about 690 million to 563 million over the past year. The market prices the 60x label and the two loss quarters. The filings price a mid-teens grower that converts a third of revenue to cash and is shrinking its share count a fifth. The single print that settles it is whether operating cash margin holds above 30% while the buyback runs.

Management

No insider has bought a share in twelve months, and thirty-six sales took out $28.8 million, including co-founder Ben Silbermann trimming $1.1 million in August. Plan status not disclosed in available filings; routine sales for a founder-led platform, not a signal on their own. The pay is the whole story here: stock compensation at 28% of revenue is why GAAP shows red ink, and it is the dilution the buyback exists to mop up. On that score management is doing the right thing, buying back at about 8 times cash rather than issuing into strength, and the count is genuinely falling. The thing to watch is that the buyback keeps outrunning the grants.

How it fails or surprises you

US advertising demand carries the whole thing. The United States is the only country above a tenth of revenue and the source of nearly all the profit. A domestic ad pullback, or Meta and Amazon pricing Pinterest out of shopping budgets, and mid-teens growth halves fast. The tell is US revenue growth: two quarters under 10% and the cash thesis loses its engine.

The losses could be structural, not seasonal. The operating loss widened across the first half of 2026 even as revenue grew 18%, a combined $135 million of red ink. The read here says the second half turns positive as it did in 2025, when Q4 alone earned $277 million. If it does not, costs are growing faster than revenue and the cash machine is a story, not a fact.

Rest-of-world users start paying (right tail). Most users sit outside the US and earn Pinterest a fraction of what an American user does. The tvScientific deal and third-party ad demand aim straight at that gap. If revenue per international user climbs, growth reaccelerates past 20%, and the market pays nothing for that today. Watch international revenue outpacing US.

Closing thoughts

This is the kind of setup a single number resolves. If operating cash margin holds above 30% through the buyback and the second half swings back to profit the way it did in 2025, the 8-times-cash price is simply wrong and re-rates. If the first-half losses are the new normal, the 60-times label was right and the cash was flattered by timing. The record favors the first reading: three straight years of rising operating cash, $613 million to $1.28 billion, is not the pattern of a business quietly breaking. The fatter tail points up. What is genuinely at risk is a US ad slowdown dragging growth into single digits, which would halve the cash trajectory and strand a share count that is smaller now but was not free to shrink.

The bet is still that advertisers keep paying to reach people who arrive at Pinterest already planning to buy, and that the rest of the world slowly starts paying the way America does. It breaks if US revenue growth slips under 10% for two quarters while operating cash margin falls below 30% at the same time, the two numbers that would show the intent is worth less than the price assumes. Until then the buyback compounds the value of every remaining share, and that is the whole case.