Google+ Had 540 Million Users and Almost Nobody Inside. Here's the Difference.

When Google finally pulled the plug on Google+, it buried a single devastating number deep in the announcement: ninety percent of user sessions on the service lasted fewer than five seconds.² Five seconds. That is roughly the time it takes to land on a page, register that there is nothing there for you, and leave. A social network is a place you live. Google+ was a place people accidentally arrived at and immediately exited—again and again, hundreds of millions of times, for years. It was not losing the social war. It was barely in the room.

The official story is that Google+ was a credible Facebook rival that simply lost. The numbers were huge—540 million monthly users at its 2013 peak.⁷ The real story is that those numbers measured the wrong thing, on purpose, and that Google+ failed not because Facebook was unbeatable but because of what Google chose to build instead of a social network.

The 540 million who were never really there

Here is the sleight of hand at the center of the whole story. In October 2013, Google said roughly 540 million people were monthly active users of Google+. But that figure counted anyone who touched the Google+ social layer—the +1 button on a web page, a comment under a YouTube video, a flicker of identity inside Gmail. Buried in the same disclosure was the number that actually mattered: only about 300 million interacted with the Google+ stream itself, the feed that was supposed to be the rival to Facebook's home page.⁷ And even that 300 million dissolves the moment you ask how long they stayed. ComScore measured the average Google+ user spending 3.3 minutes per month on the site in January 2012. The Facebook user spent more than 7.5 hours.⁶ Same word—'user'—opposite reality. One was a citizen. The other was a tourist who never got out of the airport.

The headline counts were inflated by the very thing Google thought was its advantage: it owned the front door to the internet. Once Google began requiring a Google+ profile to comment on YouTube or to use other services, accounts accumulated by the tens of millions—not because anyone wanted a Google+ identity, but because they wanted to do something else and were made to carry one. Rapid account growth looked like demand. It was conscription.

Integration was the strategy, and integration was the disease

This is the causal mechanism, and it is the opposite of the usual lesson. Google's plan was to use its ecosystem—Search, Gmail, YouTube, Android—to seed Google+ instantly, leapfrogging the cold-start problem that kills most social networks. On paper it was brilliant: why build an audience when you already own one? But the move that solved distribution destroyed the product. A social network's only real asset is that people choose to be there with each other. When you arrive because a YouTube comment box forced you to, you bring none of the intent that makes a feed worth reading. The plumbing filled the rooms with bodies and none of them with attention. Google+ became a tool for Google's ambitions—unify identity, sharpen ad targeting, stitch the products together—rather than a place users wanted to inhabit. And users can feel that. The five-second session is what indifference looks like at scale.

The fourth time was not the charm

It is tempting to file this as a one-off stumble by a search company out of its depth. It was not the first time. Google+ was Google's fourth attempt at social networking, following Orkut in 2004, Google Friend Connect in 2008, and Google Buzz in 2010.¹ A company learns from a pattern; the pattern here is a company that kept reaching for the same prize with the same instincts—engineer a network the way you engineer a search index, bolt it onto what you already own—and kept getting the same result. The genius of Google was making information frictionless and self-serve. A social network is neither. It is messy, emotional, and chosen. Google could ship the rails. It could never manufacture the reason to stay.

The breach that picked the date

There is one more turn, and it is the ugliest. Google did not shut Google+ down on the day it noticed the engagement problem; that had been visible for years. It announced the shutdown on October 8, 2018—the exact day the Wall Street Journal reported a concealed flaw in the Google+ People API that had exposed the private profile data of up to 500,000 accounts to as many as 438 third-party apps. Google had discovered and patched the bug back in March 2018, and chose not to tell anyone.⁴ A memo reviewed by the Journal warned that disclosing it would likely trigger 'immediate regulatory interest' and invite comparison to Facebook's Cambridge Analytica scandal.⁸ So the timing is its own confession: the engagement problem was the disease, but the cover-up was the thing that made the disease impossible to keep hiding. A second, separate bug in November 2018 affecting roughly 52.5 million users then pulled the final date forward to April 2, 2019.⁵³ A social network nobody visited still managed to leak the data of the people who never came.

Google+ is remembered as the product that lost to Facebook, but it never really played the game Facebook was winning. Facebook built a place and spent years giving people reasons to stay. Google built a layer and spent years counting the people who passed through it. The five-second session was not a symptom of a weak competitor; it was the honest verdict on a network engineered to serve its owner rather than its members. You can conscript an audience. You cannot conscript a reason to stay—and in the end the only number that survived the wreckage was the one measuring exactly how fast everyone left.