Startups

Marissa Mayer’s eternal Sunshine

Behind the rocky launch of a photo-sharing app was a co-founder about to quit and months of employee warnings that its new project wasn’t ready

Zoë Schiffer, Casey Newton

Apr 4, 2024 — 13 min read

Marissa Mayer attends TechCrunch Disrupt in San Francisco in 2019. (Steve Jennings / Getty Images)

For a brief moment in the tech world last week, it felt like 2012 again: a new photo sharing app had just launched, and everyone was talking about Marissa Mayer.

Mayer, a prominent early Google employee who created its famed associate product manager program before leaving to become CEO of Yahoo, had returned with a new app from the startup studio she co-founded six years ago. Shine, an app designed to let people automatically share photos taken at events they attend together, was announced March 27 — and was immediately mocked by commentators for its bare-bones design and anachronistic feature set.

“Whenever you feel like you're not good enough, always remember, the creator of the Associate Product Manager program at Google just dropped this,” wrote venture capitalist Deedy Das on X, alongside a screenshot of the app.

In the office, employees placed bets on how many downloads the app would get on its first day. Mayer set hers high, at 12,000 (12 is her lucky number). But the actual number was closer to 1,000, Platformer has learned.

This week, the company suffered another blow: Mayer’s co-founder, former Google and Yahoo executive Enrique Muñoz Torres, resigned. “Startup life is not easy, but it has been hands down the most rewarding experience in my professional career, which made my decision to leave a very tough one,” he wrote on LinkedIn. Until Wednesday, he was leading the development of an organizing app for families, which has yet to launch.

Building a hit app is hugely challenging, and chaos is a common feature of many or even most of the startups that set out to build them. Sunshine, which Mayer founded with Muñoz Torres a year after leaving Yahoo, has been no exception, according to interviews with employees, who all asked for anonymity to speak publicly about the company.

Mayer’s original idea had been to create simple utility apps powered by artificial intelligence. (She almost named the company Mundane AI, she told The Information last year.) In 2020, two years after the company launched under the name Lumi Labs, the pair announced the company’s first product: an app that promised to be “the world’s most advanced, intuitive contact manager,” according to The Verge.

The company raised a $20 million seed round in 2020 from a small group of investors, including Mayer. At the end of that year, it officially rebranded as Sunshine.

Despite high hopes, the contacts app has yet to take off. It was buggy at launch, duplicating contacts or introducing other errors into data fields. And it rattled some users by filling in contact information using public databases, raising privacy concerns.

Sunshine eventually released an improved version of the app, but it hasn't managed to reach a critical mass of users. Last summer, Mayer laid off 20 percent of the company, leaving it with about 15 employees.

She also went into brainstorming mode. Last summer, she had the idea for the app that would become Shine. Employees have been working on it ever since.

But its path from concept to execution has been a rocky one.

II.

In October, Mayer threw a Halloween party at her house in Palo Alto.

Mayer’s public Halloween bashes are famously lavish: complete with professionally carved pumpkins, jumbo bags of candy, and plush stuffed animals.

The 2023 party was a quieter affair. Guests — a mix of friends and employees — gathered to carve pumpkins and test out a prototype of Shine, according to a former employee.

Photo sharing became one of the most popular categories of apps in the early 2010s, as rapidly improving smartphone cameras and cellular networks led people to take more photos than ever before. Startups raced to build social networks around the feature, and the tech graveyard is now littered with the ones that didn’t make it: Path, Picplz, DailyBooth, and many more.

One of the most famous flameouts in photo sharing was Color, which showed users every photo taken by fellow users in their immediate vicinity. The company raised $41 million — an eye-popping sum at the time — but got next to no traction, and eventually its engineering team was acquired by Apple in 2012 in a fire sale.

To a degree that surprised employees, Shine sought to borrow ideas from Color. Like that app, it would ask users to share their location and photos with the app and with fellow users. When Shine users were nearby one another and snapping photos, Shine would automatically sort those photos into shared albums.

“We’ll do Color, but we’ll have nudity filters so we’re not Color,” Mayer said, according to notes taken by the same former employee. (Color users used to troll each other by posting nudes, knowing that other people would see them when they opened the app.)

Some staff members were skeptical. First, they worried that few people would be willing to grant the app permanent access to their location and to their photo libraries.

Second, the team struggled to track users’ locations accurately enough to ensure that photos were shared as intended. Shine uses a Google API to track location, and it wasn’t always precise. Once, an employee found that photos of their daughter had inadvertently appeared in an album shared with colleagues, due to changes Sunshine made to its location parameters.

Third, employees say there was little clarity about how Shine would compete. Apple and Google now have system-level features for sharing photos privately with friends and family. Mayer leads product development for the app — but if there is a product roadmap, she has largely kept it to herself.

Employees say they learn what they are working on each week during Monday morning standup meetings, and that their mandate shifts frequently as Mayer changes priorities. They complain that Mayer often does not heed their advice, and say most product decisions are made based on Mayer’s own experiences using the app.

While disarray in product development is common to many startups, in other ways Sunshine has been built to resemble a much older company. Employees are required to work in the office five days a week, and have an all-hands meeting that is scheduled for 5PM on Fridays. (During ski season, we’re told that the meeting has started as early as noon). But Mayer is often late to the meeting, employees say, extending their work into the weekend.

When employees ask Mayer questions on Slack, employees say, she often prefers to answer them in person, delaying action. Sunshine manages its projects using Monday.com, but the app is not widely used internally, and in practice many projects are tracked on a white board, which the team calls “analog Monday.”

“We honestly wanted to work hard, but the company is just so inefficient,” said the former employee.

In any case, employee concerns about Shine proved to be mostly correct. After the party, during which few people gave the necessary permissions, Mayer agreed to try a different approach.

III.

After the Halloween party, Sunshine quietly made Shine available in Apple’s App Store. The requirement for users to permanently share their location with the app had been removed. Instead, it asked users to create or join individual events. If one Shine user created an event for a birthday party within the app, and another Shine user joined that event, the photos they took at the event would be shared with one another automatically.

The app evolved as Mayer tested it out with friends and family on vacation or at parties, employees said. Product changes were “explicitly about her social life,” one employee said.

In January, Mayer and her family went on a ski vacation in Wyoming. Mayer’s friend took some selfies on the trip — and was embarrassed when the photos automatically uploaded to a shared album on Shine, which Mayer had presumably created. When Mayer returned to the office, she told employees that photos taken with the front camera that only contained a single face should be excluded from shared albums.

Inside the company, frustration was mounting. Mayer wanted to announce Shine in February, but the app was still light on features. Every week, it seemed, she came into the office with new ideas that employees had to rush to implement. And Mayer had laid off the app’s only designer in January, according to former colleagues, leaving it without strong visual direction. At times, Mayer’s design suggestions involved a simple line drawing on a piece of paper that her engineers had to struggle to interpret.

When Shine finally did launch last week, commentators derided it for brutalist design, which seemed more appropriate to an app for filing expenses or recording timesheets. The most notable feature of its design was arguably its bright purple scheme, which many found reminiscent of Yahoo. (Mayer called the similarity “purely coincidental.”)

The frosty reception to Shine, along with the departure of Muñoz Torres, has clouded Sunshine’s future. While there is undoubtedly some appetite for the utility apps that Mayer envisions, the company’s history of buggy launches and tumultuous product development has left some close to the company with serious doubts. Six years in, the startup has yet to find a large audience.

Setbacks aside, work on both Sunshine and Shine continues. And earlier this week, in response to the criticism, Sunshine posted a mea culpa on X:

“We heard you all loud and clear on Shine's design,” the company said. “We're excited to share that later this week, we'll be updating our UI to address the feedback raised by Design Twitter.”

It was April Fool's Day, and the post included some joke screenshots of an even more brutalist-looking app. Later, the company called its own bluff.

"We did hear your feedback and are currently working hard on a design that will make Shine's interface more modern," it said.

This time, the company said, it was not joking.

More on the Grindr squeeze

Last week we told you about Grindr’s plans to ramp up its effort to get users to pay for more features. We can now tell you that employees at high levels of the company are currently engaged in discussions about putting some chat features behind the paywall, Platformer has learned.

Over the weekend, a bug restricted the number of free chats that users could send to people outside their geographic area through a feature called Explore. That bug has now been fixed, but while active it resulted in a surge of subscriptions — about 2,000 a day, compared to the typical 800. As a result, the company is now running a test to further restrict the number of free chats a user can send to someone outside their close vicinity, from three in a 24-hour period to as few as zero.

Employees are tracking how these tests impact both subscription sign ups and user sentiment about the app — and will likely try to find a sweet spot between increasing the first without tanking the second. Today, about 19 percent of Grindr’s daily active users use the Explore feature.

These tests are separate from Grindr’s discussions about paywalling chats more broadly, we’re told. So far, these discussions haven’t risen to the level of testing. But if the company continues to see sign ups surge after restricting free features, I wouldn’t be surprised to see aspects of the core chat function slip behind the paywall as well.

— Zoë Schiffer

On the podcast this week: Kevin and I examine new studies on how AI may already be transforming the economy. Then, filmmaker Paul Trillo joins us to discuss how he used OpenAI's Sora to explore making movies with AI. And finally, Kevin catches me up on the incredible story of the XZ Utils vulnerability.