Phia founders knew about cookie stuffing months before Bloomberg exposé

Phia, the shopping startup co-founded by Phoebe Gates and Sophia Kianni, is facing fresh scrutiny over its business practices. Bloomberg reporters have learned that the founders were aware of the illegal cookie stuffing scheme as far back as December, even though the company previously claimed it only found out after the outlet reached out.

Cookie stuffing is a tactic where a platform takes credit and commission for affiliate purchases it did not help generate, effectively siphoning referral revenue from other marketers. Such behavior is typically banned by contracts in affiliate networks because of its unfairness. In July, Bloomberg published an investigation into Phia, and a company spokesperson at the time dismissed the issue as a "bug."

But new reporting, based on leaked Slack messages and sources familiar with the matter, paints a different picture. Cookie stuffing was a deliberately built feature that could be toggled on and off. Gates and Kianni discussed the practice with other executives and engineers in chats. According to Bloomberg, the scheme accounted for a substantial share of Phia's sales, and daily revenue dropped significantly once it was stopped. Affected retailers included Nike and Nordstrom.

In response to the new allegations, a Phia spokesperson partially pushed back on the outlet's claims but said the company would "learn from this." Phia did not immediately respond to TechCrunch's request for comment.

Phia launched last April as a more efficient shopping alternative to Google Flights, helping users find the best prices across retailers. Days after the July investigation, Puck reported that Phia had lost nearly half its full-time staff since the start of the year, that several brands were unaware they were listed on the app, and that investors were turned off by the startup's aggressive affiliate marketing. The company has also faced past criticism for secretly collecting swaths of sensitive data, including tracking user behavior.