Why did Google get rid of the company behind Pokémon Go?

Augmented reality — tech that lets digital worlds interact with our physical one — is Google’s Moby Dick.

Did the company let a whale slip through its fingers?

By now you’ve heard of Pokémon Go, the mobile game that exploded in popularity over the weekend. Its allure comes from the ability to chase Nintendo cartoons in real environments with a smartphone — a tangible application of augmented reality. If the game’s growth keeps up, it would be the first AR app for the masses.

And those masses could have been addicted users for Google.

But they’re not. Niantic Labs, the company that built the game, was formed inside the search giant but spun out on its own last fall with the formation of Alphabet. The reasons why are very particular to the gaming company — but they also reveal some hurdles Google faces in its new experimental corporate anatomy.

Niantic began, like many old Google projects, as a pipe dream of a proven exec. John Hanke, a key figure behind Google Maps and Google Earth, was given leeway to form Niantic in 2010. He and a small band of engineers had this task: Build stuff on top of Google’s mapping tech.

They released a couple of Android apps: Field Trip, a location-based notification service, and Ingress, a multiplayer mobile game around physical spots. Pokémon Go is basically Ingress repackaged with Pokémon characters.

Niantic was set up as an “autonomous business unit” — an early Google experiment, pre-dating Alphabet, in running projects at arm’s length from the mothership. (Google Fiber, its broadband unit, was one.) Niantic’s plan was either to roll its tech back into Google eventually or be spun out, said a person familiar with its origins.

Although Niantic spun out when Alphabet was announced, this person said the plans were in place before the corporate restructuring began. Hanke told Business Insider the talks “surfaced” as Alphabet did. Niantic had, on the back of Ingress’ traction, become more of a gaming company than a Google one.

Here’s what Google said at the time of the spinoff:

They’re now ready to accelerate their growth by becoming an independent company, which will help them align more closely with investors and partners in the entertainment space.

One key reason for Niantic’s split may be the latter group mentioned in that statement: Partners.

Niantic’s independence was likely the clincher in getting its partner, Nintendo, signed on for Pokémon Go. (We asked Niantic if that was the case but didn’t get a response.) Hanke told Business Insider his operation was “always kind of bumping up against Google’s desire to stay neutral.”

That is, Google loves platforms — like its app store and maps — and does not want to be seen as favoring one developer over others. And potential partners, like Nintendo, would usually rather partner with an independent company rather than a giant like Google.

Perhaps Google decided it already had a bet in the space. A year before Niantic spun out, Google put a ton of its own money into Magic Leap, a startup that, like Niantic, is working with gaming and AR. Unlike Niantic, Magic Leap is pitching itself as the definitive platform for the newfangled tech.

It may turn out that Google made the right move.

If Niantic ends up creating a hit because of its independence and flexibility, then Google can point to its bona fides as an incubator for interesting, successful companies. And Niantic could get far more people accustomed to AR — a win for Google. Tango, its 3-D mapping tool, is designed (in theory) to give any smartphone instant AR navigational powers.

Google could make some money too. It pitched in on Niantic’s $30 million round after the spinoff. It’s not clear if that investment includes a stake in the earnings from Pokémon, which, if early days are indicators, look lucrative. (I asked Google; no response.)

But if Niantic only found such lucrative returns because it was outside of Google, that might concern the tech giant. It is trying — both with Alphabet and its own incubation program — to foster and keep startups under its roof.

The Alphabet model, placing distinct companies and cultures under one corporate umbrella, has had some hiccups so far. Google may find that these startup founders want to make hits out on their own.

This article originally appeared on Recode.net.