After selling Instagram for $1B, he found his next big lesson in a startup that wasn’t worth the years ahead
For years, Mike Krieger had lived inside the startup arc most founders dream about – the explosive growth charts, the billion-user milestones, the billion-dollar acquisition that turns a product into infrastructure.
But after later being faced with what seemed to be a much quieter momentum and graph with his personalized news app Artifact, he chose to pack it all in and walk away.
Not because the product was broken, and not because the users were gone, but because the future he’d imagined never materialized in the data. And while the typical decade-long startup playbook often teaches you how to persist, Krieger had spent a decade learning how to listen.
Taking the long road to overnight success
Understanding how Krieger arrived to where he is today doesn’t start necessarily take you back to Menlo Park. He grew up in São Paulo, Brazil, and eventually moved to California to study symbolic systems at Stanford – a degree that sits somewhere between computer science and psychology.
There he met Kevin Systrom, a fellow tinkerer with whom he would spend years swapping ideas, prototypes, and half-serious side projects before finally deciding to build something together for real.
Their first attempt at what’s now known as Instagram was called Burbn, a crowded check-in app where photos were almost an afterthought. It wasn’t long, however, until its users told a different story – ignoring nearly everything except the pictures.
That small insight forced a big change, prompting the two founders to decide between clinging to their original idea or pivoting to bet on the thing people were saying they loved. So, they decided to kill features, rename the product, and ship a stripped-down photo app.
Instagram, in its new form, grew from a few thousand users to millions in months, eventually climbing to tens of millions.
In 2012, Facebook came calling with a $1 billion deal for the 13-person startup. Over the next six years, Instagram kept compounding, and by mid-2018, it passed 1 billion monthly active users.
Back to zero after a billion-dollar exit
After a few more years inside Facebook, Krieger and Systrom left in 2018. They had already experienced what runaway growth felt like, and this time they wanted something quieter – perhaps a small team, a blank page, and a question they had been circling for years.
What would a calmer, smarter news app look like if you built it today?
The result was Artifact, an AI-driven news app that learned your interests and surfaced stories accordingly. When it launched in early 2023, the coverage was enthusiastic, along with the inevitable nickname, “TikTok for text.”
Behind the coverage, the graphs were less flattering. Artifact had a devoted core, but the growth curves didn’t look like the foundation of another billion-dollar company that they could build for the next 10 or 20 years.
In their shutdown note just a year after launch, the founders said they’d built something a core group of users loved, “but we have concluded that the market opportunity isn’t big enough to warrant continued investment in this way”.
That sentence, however, came at the end of a much longer, behind-the-scenes process. Instead of drifting for years on hope, Krieger and Systrom wrote down the big experiments they’d “feel silly not having tried” before quitting.
Letting the data make the call
They tied each experiment to concrete metrics and timelines, then ran the tests. Some numbers moved, but not enough to change the shape of the business. And at that point, rather than talk themselves into just one more cycle, they followed through and chose to stop.
As Krieger said during an interview with EO Studio last year, walking away is “never a celebration – but it doesn’t have to be a tragedy either.”
Artifact may have shut down as an app, but it managed to survive as an idea. In April 2024, Yahoo acquired the company’s AI-powered recommendation technology to power personalization across Yahoo News, TechCrunch, Engadget and other properties.
Krieger moved on too. In May 2024, Anthropic announced he was joining as its first Chief Product Officer, overseeing product engineering, management and design as the company pushed its Claude models into the mainstream.
Today, after a reshuffle, he co-leads Anthropic’s internal Labs group, back in a smaller, scrappier unit focused on experimental AI products.
Seen in sequence, the through-line is hard to miss. A Brazilian student chases an idea across continents, helps build one of the most used apps in history, then voluntarily closes his next startup when the data refuses to match the dream – walking straight into another frontier.
Because sometimes, knowing when to stop building matters just as much as knowing when to start.
