The decline of Food52, Goop, Hodinkee—and the internet’s dream of content-to-commerce

Remember how much fun it was to shop on the internet a decade ago?

If you visited the Goop website, Gwyneth Paltrow might introduce you to her favorite $75 candle or $95 vibrator. If you were looking for a lasagne recipe, you could find a good one on Food52—along with recommendations for a baking dish hand-selected by former New York Times food editor Amanda Hesser. Watch-lovers flocked to Hodinkee to see what founder Benjamin Clymer thought of the cool new Longines or Omega timepiece (with a handy link to buy it, in case you really liked it).

At their peak, around five years ago, all of these media companies landed millions of dollars in venture capital and had valuations well into the nine figures. Legacy media ranging from the New Yorker to Vogue took a page from their book, too, linking to products you could buy directly from the pieces published on their websites.

But over the last two years, this generation of content-to-commerce pioneers has fizzled out. Goop has gone through multiple rounds of layoffs and its website is a shell of what it used to be. In 2024, Hodinkee was sold at a fraction of its former valuation. And last month, Food52 declared bankruptcy and is headed towards a fire sale.

It’s worth asking what happened to these startups—and what comes next, as AI transforms the way we shop online.

The rise and fall of Food52

The rise and fall of Food52 offers insight into what went wrong with the content-to-commerce model. Founders Amanda Hesser and Merrill Stubbs had come from the traditional food media. They saw a gap between legacy magazines like Bon Appétit and Food & Wine, which prioritized the perspectives of elite chefs, and amateur food blogs, which were flooding the internet. With Food52, they invited home cooks to submit recipes, which their team would test. The best ones would be featured on the site, alongside beautiful photography. The concept resonated and site traffic grew quickly.

Initially, the company generated revenue from advertising and brand partnerships. But in 2013, the site launched a shop that sold kitchenware and artisanal ingredients that Food52 staffers recommended. This approach made sense says Dan Frommer, founder of The New Consumer. One of the biggest problems with shopping online is the overwhelming volume of products available. First generation content-to-commerce startups offered expertise and a point of view, which gave them the authority to recommend products. “They were offering curation, which was a valuable service at the time,” he says.

Goop and Hodinkee followed similar trajectories. They began as blogs centered around a particular perspective and aspirational lifestyle, driven by their well-known founders. Over time, they built up enough trust with their readers to sell them products. (Food52 declined to comment on the story. We reached out to Goop and Hodinkee, but neither got back to us by the time of publication.)

In 2019 and 2020, investors still believed this might be the future of retail. They pumped millions into their startups to grow their audiences, start new revenue streams like events, and start their own product lines. Food52, for instance, was valued at $300 million in 2021, after an $80 million investment from TCG (which also invested in Hodinkee).

But this funding may have inadvertently led to their decline. With the influx of cash, these startups had a mandate to scale, but they all struggled to grow sustainably. By the start of this year, Food52 had declared bankruptcy. America’s Test Kitchen has reportedly agreed to buy it for $6.5 million, of which $3.42 million is Chapter 11 financing.

Frommer argues that there were many idiosyncratic reasons why each of these companies failed. Food52, for instance, appeared to have bitten off more than it could chew. In 2019, it launched its own in-house kitchenware line; it also acquired two entirely new companies, the Danish cookware brand Dansk and the lighting brand Schoolhouse. “There was a lot wrong with the business,” Frommer says. “There were failures in strategy and execution.”

But taking a step back, it’s clear that there were also broader issues with the content-to-commerce model that affected all of these businesses.

What Didn’t Work—and What Did

These early content-to-commerce platforms accurately identified that consumers were overwhelmed with the avalanche of products available on the internet—and they also knew that taste could be monetized. Still, there were flaws with their model.

For one thing, consumers often didn’t come to these websites with the intent to shop. They were there to take in the content: the recipes, listicles of clean beauty products, or a conversation with Ed Sheeran about his favorite watches. Only a small proportion of consumers would feel compelled to buy a product. Often, when a publication’s famous founder recommended a product, it would sell better; but over time, as the sites grew to have teams of writers, the sites no longer conveyed the distinct sensibilities of Paltrow, Hesser, or Clymer.

Then there were the economics. It is hard to make money by marketing other brand’s products. These sites generated small amounts of revenue by selling products at a markup on their online stores or by making a commission by driving the customer to another brand’s website. All of these companies realized that a more profitable route was to make their own products, which they all did, from Goop’s beauty and fashion lines to Hodinkee’s watch straps and limited edition collaborations with brands like Longines. But this meant building out teams with expertise in designing and sourcing products, which was also a major investment.

Finally, there was all the competition. Other media sites quickly realized they, too, could create a new revenue stream by linking to products. And some began doing it much more effectively. In 2016, for instance, the New York Times acquired Wirecutter for $30 million. Unlike Food52, Goop, and Hodinkee, Wirecutter was designed to help consumers at the moment when they were ready to buy a product. New York Magazine built its own product recommendation site called The Strategist, which has a similar model.

“Content that really drives commerce is not just ambient recommendations around fun articles,” says Frommer. “It’s really purpose-driven content designed to help the consumer solve a problem. The majority of traffic to Wirecutter and The Strategist happens at the moment of need—they promote their humidifier recommendations when the winter air is dry.”

The content-to-commerce model hasn’t disappeared; it has shape shifted. There are now massive players like Wirecutter that dominate the landscape. And at the other end of the spectrum, there are armies of individual content creators who recommend products to their followers on Substack, Instagram, or TikTok. It’s just the middle of the market that has collapsed.

But as with everything on the internet, change is constant. And everything we know about how to shop online is about to get transformed by AI, which is already where many people begin their shopping journey. In many ways, AI agents are the ultimate blending of content and commerce: They offers product recommendations, personalized to the user, presented within a conversation. But what’s missing from AI is a unique point of view or sensibility—which is what the early content-to-commerce players excelled in.

In an AI-driven shopping future, the winners won’t be the smartest algorithms. It’ll be the ones that blend data with something that feels like taste.