In September 2025, a shopping app founded by two Stanford graduates raised eight million dollars from Kleiner Perkins. Four months later it raised thirty-five million more, led by Notable Capital, at a valuation around a hundred and eighty-five million. On the cap table sat Kris Jenner, Hailey Bieber, Sheryl Sandberg and Michael Rubin.
The company is Phia. One of the founders is Phoebe Gates, whose father built Microsoft.
That last fact is the one every headline leads with, and it is also the least interesting thing about what happened next.
What the company actually does
Phia is a browser extension and app that compares prices across retailers, surfaces promotional codes, and looks for secondhand equivalents of whatever you are about to buy new. Gates has described the ambition as something like Google Flights for clothing: one interface that reads the whole market and tells you where the same item is cheapest.
The secondhand piece is the strategically interesting part. Resale stopped being a values choice some years ago and became a price mechanism, which is why every serious shopping tool now indexes it. A comparison engine that ignores resale is comparing an increasingly incomplete market.
Her co-founder is Sophia Kianni, a climate activist who was the youngest person to advise the United Nations on climate. The pairing matters for the product's positioning, because a resale-first shopping tool reads differently coming from someone with that record than it would from a pure commerce founder.
The privilege question, asked properly
Gates has been direct about the surname. She has said she wants Phia to succeed with no ties to her privilege or her last name, and that she has taken no money from her parents for the company.
Both things can be true at once, and the useful distinction is between capital and access. She did not take family money. She did not need to, because a Stanford network, a recognisable name and an introduction to Kleiner Perkins are worth more at the seed stage than the seed itself. The scarce input at that moment is not cash. It is the meeting.
The fair version of the criticism is not that she cheated. It is that the counterfactual is unknowable, and that a founder with the same product and no surname would have needed eighteen months to reach the room she started in.
The fair version of the defence is that access gets you the first meeting and nothing after it. Notable Capital led a thirty-five million dollar round in January 2026, seven months after launch. Growth-stage investors are not writing that cheque on a name.
Then the trouble
In mid-2026, Phia was accused of cookie stuffing, an affiliate practice in which a tool drops tracking cookies on a user's browser without a genuine click, so that the tool collects commission on a purchase it did not actually influence.
The company said the behaviour was unintentional and that it had been resolved.
Take that at face value and the incident is still instructive, because it exposes the structural problem inside every shopping assistant. These products are free to the user and monetise through affiliate commission. The tool that finds you the cheapest price gets paid by the retailer you buy from. The incentive to widen attribution is built into the business model, and the line between aggressive tracking and fraud is a matter of degree rather than of kind.
Honest operators in the category deal with this by capping attribution windows and requiring genuine click-through. It costs revenue. That is the point.
The case against the whole category
A piece that only makes the argument for its subject is advertising, so here is the serious challenge.
Price comparison has been tried repeatedly and mostly does not stick. The graveyard is full of extensions that solved the same problem, got acquired, and quietly degraded. Retailers hate them, because a tool whose entire function is to move a sale to whoever is cheapest is a margin tax on the whole industry. Retailers have every reason to make the data harder to read.
There is also a distribution problem. Browser extensions are a shrinking surface. Shopping increasingly happens in apps and, more recently, inside AI assistants that make the recommendation themselves. If an agent picks the retailer, an extension that compares retailers has been disintermediated by the layer above it.
The counter is that Phia is building the comparison layer that those agents will need to call, rather than competing with them for the user's attention. Whether that is the plan or a retrofit is not yet observable from outside.
What it signals
Gates appeared on the 2026 Forbes 30 Under 30 list for retail and ecommerce. The list still moves careers and still produces a reliable share of collapses, so treat it as a marker of attention rather than of durability.
The more durable signal is the cap table. A consumer shopping tool raising forty-three and a half million dollars across two rounds, in a period when consumer investing had been unfashionable for three years, says something about where capital thinks the next interface sits. It is a bet that the shopping decision itself, rather than the store, is the thing worth owning.
That bet may be wrong. It is not a small one, and it was not made because of anybody's father.
Sources
Fortune, on Gates and the privilege question. Yahoo Finance, on the affiliate allegations. Funding figures reflect reported rounds from September 2025 and January 2026.
