How Does Phia App Work? The 85M AI Startup Explained

In January 2026, a snowstorm shut down New York City. Phoebe Gates and Sophia Kianni were stuck in their apartments, unable to reach the Phia office. Their 20-person team showed up anyway, sending Slack photos of snow piling on the porch. Ten months after launch, Phia had just closed a $35 million Series A led by Notable Capital, with Khosla Ventures and Kleiner Perkins writing checks alongside more than 30 celebrity investors. The company was valued at $185 million. Gates hasn’t taken a dollar from her parents.

Phia is a free AI-powered shopping assistant, co-founded by Phoebe Gates and Sophia Kianni, that compares prices across 40,000+ retail and resale sites in real time through a browser extension and iOS app. It’s hit 1.5 million users, been named a TIME Best Invention, and attracted backing from Sydney Sweeney, Paris Hilton, Kris Jenner, and Jessica Alba. But what matters more than the celebrity roster is how the product actually works, how it makes money, and what the founder playbook looks like for other builders watching from the outside.

Last updated: June 2026

Quick answers

How does Phia work? Phia is a free browser extension and iOS app that scans prices across 350 million items from 40,000+ retail and resale sites when you click “Should I Buy This?” on any product page. It shows cheaper alternatives, applies coupon codes, and tracks price drops automatically.

Is Phia free? Yes. Phia costs nothing to download or use. The company earns affiliate commissions from retailers when users purchase through its recommendations, so the user never pays a fee.

How does Phia make money? Phia operates on an affiliate model. When a user clicks through to a retailer and buys something, Phia earns a commission from the brand. The company compares this to how travel booking platforms like Booking.com monetize recommendations.

How does Phia work?

Phia runs as a Chrome browser extension and an iOS app. The core feature is a button called “Should I Buy This?” that appears while you browse any shopping site. Tap it, and Phia’s AI scans its database of 350 million products across more than 40,000 retail and resale partners to find the same item (or something visually similar) at a lower price.

Say you’re looking at a $200 Anthropologie dress. Phia checks Poshmark, The RealReal, ThredUp, and hundreds of other secondhand marketplaces to see if the same dress is available for $80. It also compares new prices across retailers, so if Nordstrom has the same item on sale, that shows up too.

The app does three things beyond basic price comparison. First, it auto-applies coupon codes at checkout, including exclusive codes negotiated directly with brands. Second, it lets you set price drop alerts on any item, so you get notified when something you want gets cheaper. Third, it runs a rewards program where you earn points on eligible purchases that convert to gift cards.

Under the hood, Phia uses AI to analyze pricing trends and resale value. The system can tell you whether an item is priced high, low, or typically for its category. For a founder building an AI product, that real-time pricing intelligence is the technical moat. It isn’t just matching SKUs. It’s reading product images, descriptions, and brand positioning to surface “visually similar” alternatives that a keyword search would miss.

The Chrome extension works on any shopping website. You don’t have to be on a Phia partner site. Browse Zara, SSENSE, or any independent boutique, hit the button, and Phia pulls results from its index. The iOS app gives you a standalone feed where Phia recommends items based on your browsing history, saved items, and style preferences. Gates told TechCrunch that the long-term vision is a “holistic shopping agent” where you go to Phia first, get outfit recommendations based on what’s already in your closet, and decide what to sell or donate.

How does Phia make money?

Phia uses an affiliate revenue model. When a user clicks a product recommendation and completes a purchase through one of Phia’s 10,000+ retail partners, the company earns a commission from the brand. The user pays nothing extra.

Kianni has compared it to how the travel industry works. Booking.com doesn’t charge travelers a fee. It recommends the best option, and the hotel pays a percentage of the booking. Phia applies that same structure to fashion and retail. The company told TechCrunch it achieved 11x revenue growth since its April 2025 launch, and it’s on pace for nine-figure sales growth in 2026.

This model creates an alignment that traditional advertising doesn’t. Phia only makes money when it actually saves you money or finds you something you want to buy. If the recommendations are bad, users leave and revenue drops. That’s a stronger feedback loop than display ads or sponsored placements, where platforms get paid regardless of whether the user benefits.

The metrics Phia shares with brand partners back this up: 13% higher conversion rates, 30% stronger new customer acquisition, 15% increased average order value, and return rates cut by more than 50%. If those numbers hold at scale, the affiliate model could be more profitable per user than traditional e-commerce advertising, where customer acquisition costs have ballooned across Meta and Google over the past three years.

Who founded Phia?

Phoebe Gates and Sophia Kianni co-founded Phia in 2025. They met as randomly assigned roommates at Stanford, bonded over activism and shopping, and started brainstorming startup ideas in their dorm room before dropping out to build the company full-time.

Gates is the youngest daughter of Bill Gates and Melinda French Gates. She’s been vocal about building Phia without her parents’ money or connections. “I have a chip on my shoulder,” she told Fortune in February 2026. “I want this to succeed with no ties to my privilege or my last name.” She hasn’t taken funding from her family, insisting on raising outside capital even when some investors seemed more interested in her personal life than her pitch deck.

Kianni brings a different origin story. She’s an Iranian-American climate activist who founded Climate Cardinals, the world’s largest youth-led climate nonprofit, at 17. She became the youngest-ever advisor to the United Nations on climate change. The sustainability angle in Phia, specifically the push toward secondhand shopping as an alternative to fast fashion, comes directly from her background.

Together, they’ve built a social media presence of 2.7 million followers across platforms, including a podcast called The Burnouts where they interview figures from Bryan Johnson to Paris Hilton. That audience has been the primary growth engine. For other founders picking a business to build, the lesson is clear: Gates and Kianni didn’t just build a product. They built a distribution channel through their personal brands first.

Phia’s funding and valuation

Phia has raised $43.5 million in total funding across two rounds, reaching a $185 million valuation in less than a year since launch.

The investor list reads like a casting call. Over 30 celebrities participated in the Series A, including Ice Spice, Karlie Kloss, Halsey, Mindy Kaling, Olivia Culpo, Priyanka Chopra Jonas, and Olympic gold medalist Eileen Gu. On the tech side, Vladimir Tenev (Robinhood), Ankur Jain (BILT), the co-founders of Venmo and Linktree, and Mati Staniszewski of ElevenLabs all wrote checks. Alexandre Arnault, heir to the LVMH empire, also invested.

That celebrity roster isn’t just a PR play. Each investor brings an audience. When Sydney Sweeney posts about an app she invested in, her 23 million Instagram followers see it. Multiply that across 30+ investors with their own audiences, and you’ve got an organic distribution network that paid advertising can’t replicate. It’s the same strategy that helped brands like Hailey Bieber’s Rhode reach critical mass without traditional marketing budgets.

Phia vs. Honey vs. Karma

Phia isn’t the first browser extension that promises to save you money. Honey (now owned by PayPal) and Karma have been in this space for years. But Phia does something fundamentally different.

Honey finds coupon codes. Karma tracks prices. Phia does something neither of them does: it searches the entire secondhand market in real time and uses AI to find visually similar items across resale platforms. If you’re looking at a $400 bag on a brand’s website, Phia might surface the same bag on Poshmark for $180, or a nearly identical style from a different brand for $90. That’s a different value proposition than saving 15% with a coupon code.

The trade-off is scope. Honey and Karma work across electronics, home goods, groceries, everything. Phia is focused on fashion, clothing, shoes, and accessories. If you’re buying a TV, Phia isn’t helpful. If you’re buying clothes and care about getting the best deal across new and resale markets, it’s the only tool doing that in one click.

Is Phia legit and safe to use?

Phia is a real company with real venture backing, 1.5 million users, and a TIME Best Inventions of 2025 award. It isn’t a scam. But “legit” and “safe” are different questions, and Phia has had one notable privacy controversy worth knowing about.

In November 2025, Fortune reported that security researcher Maahir Sharma, a former Meta software engineer, discovered Phia’s Chrome extension was transmitting HTML snapshots of every web page users visited back to the company’s servers. That included non-shopping sites like Gmail. The extension was sending compressed page data through GraphQL calls to Google Cloud Storage, even when the user wasn’t interacting with e-commerce.

Phia’s own privacy policy stated the company “generally excludes personally identifiable information” and only collects data from “retail sites.” The actual behavior contradicted that claim. After Sharma contacted Phia, the company removed the feature but didn’t publicly disclose the issue or explain what happened to the data already collected. A separate NowSecure analysis of the Safari extension confirmed similar overreach.

Kianni told TechCrunch in January 2026 that Phia is “extremely transparent with users” and that “all data is aggregated, anonymous, and only used for the purpose of being able to help users find the best products.” Whether you accept that depends on how much weight you give to the original privacy policy gap. The feature is gone, and Phia has grown from 500,000 to 1.5 million users since the incident without further reports. But for anyone who runs sensitive browsing alongside shopping, it’s worth knowing the history.

What founders can learn from Phia’s playbook

Phia went from dorm room idea to $185 million valuation in under 18 months. That’s fast, even by Silicon Valley standards. Here’s what’s replicable and what isn’t.

Distribution before product. Gates and Kianni built a combined social following of 2.7 million before Phia launched. Their podcast The Burnouts gave them a weekly content channel to talk about building the company in public. By the time the product hit the App Store, they already had an audience ready to download it. Most founders building AI tools do the opposite: build first, then scramble for distribution. The lesson isn’t “become famous first.” It’s that audience building and product building should run in parallel from day one.

Celebrity capital as distribution, not decoration. Phia’s 30+ celebrity investors aren’t paying for brand association. They’re buying organic distribution. Each investor has an audience. When they mention Phia, their followers see it. That’s not the same as paying a celebrity for an endorsement deal. The investors have skin in the game, so their promotion is authentic. This mirrors what the CHAMP Fund does with athletes: equity over endorsement fees.

The naming problem matters. Gates went out of her way to avoid leveraging her family name. She hasn’t taken parental investment. She’s talked publicly about investors who were more interested in her father than her product. The irony is that every article about Phia, including this one, mentions that she’s Bill Gates’s daughter. The name generates attention whether she wants it to or not. The strategic insight is how she’s handled it: acknowledge it, refuse to lean on it, and redirect every conversation to the product. That positioning has generated more press coverage than if she’d simply embraced the connection. Contrast that with the Sydney Sweeney approach at SYRN, where the celebrity identity is the brand. Both can work. Gates chose the harder path, and it’s paid off in credibility.

Pick a model that aligns incentives. Phia makes money only when it saves the user money. That sounds simple, but most consumer tech companies make money by selling attention (ads) or subscriptions (paywalls). The affiliate model means Phia’s business interests and user interests point the same direction. For founders building AI products, the question isn’t just “what can the technology do?” It’s “how do we make money in a way that makes the user’s life better, not worse?”

What isn’t replicable is the unfair advantage. Gates has a last name that guarantees press coverage. Kianni has a UN credential and a climate nonprofit. The celebrity investor network flows partly from those personal connections. Other founders won’t have those levers. But the underlying playbook, building audience first, using investor equity as distribution, and choosing a business model aligned with user value, works without the famous last name. It just takes longer.