Why one Google designer encouraged his team to get drunk before testing products
That means that Googlers will often test new gadgets or software pre-launch to work out bugs, validate their worth, and give feedback and suggestions on their capabilities.
Ideally, the dogfooding process helps designers and engineers spot issues that they can weed out before unleashing the product on the world.
But several recent Google employees tell Business Insider that the company's super-nerdy, relatively homogeneous make-up can can sometimes lead to questionable product decisions.
"To get a product to launch, you have to garner the will of a bunch of Googlers," one former Google exec says. "So, it had to be something that Googlers would want to use."
One sure-fire way to get Google employees excited about something in the dogfooding stage, he continues, was to add data-centric features and charts.
"I really believe products over-index to what a Googler would want, so that dogfooding goes well," he says. It's easy to add features that appeal to engineers, so those features end up shipping, even if they may not be valuable to a wider audience. "Products that should be so simple and easy end up 'nerdy.'"
Despite some progress in increasing its diversity, the company is still mostly run by white males, with a culture that celebrates nerdiness (and once had a reputation of making non-engineers feel like second-class citizens). Embracing a passionate geek-out is all well and good, but it does mean that the company needs to continually check itself for unconscious biases.
Google actually enrolls employs in formal classes to learn to combat subtle prejudices, which can span boundaries like race, gender, physical abilities, and intelligence levels.
The problem showed up in failures like Plus and Wave
Another former employee says he noticed the problem when the company built "Circles" in Google Plus. The feature allows users to create groups to sort other users into, without letting those people know what group they're part of. Users found the methodology unnecessarily confusing.
This employee says that he thought part of the problem was that Googlers often forgot that most people didn't have their internet-savviness and tolerance for complication. (Google Plus has since all but fizzled out and been broken up into different pieces.)
"You were building tools for yourself," Lars Rasmussen, a former Googler who now works for an interactive music startup told Business Insider.
Rasmussen worked on Google Wave, a now-defunct product that was supposed to be the ultimate collaboration tool. The company eventually killed Wave because there wasn't enough user interest: It was just too complicated. As Business Insider's Nich Carlson put it, Google Wave made him feel stupid and angry.
The problem, Rasmussen says, is that the team forgot that not everyone thinks and works like a Googler - and doesn't want to!
"I still run into people who loved Wave - who thought it was the best ever and can't believe that Google canceled it. And whenever that happens, it's like I'm looking at a mirror-image of myself," Rasmussen says. "Someone who is similar to myself in skill, experience, and profession. And that's just not a mass market."
Google has tons of products where it nailed this. Android is the world's most popular phone software, Gmail had hundreds of millions of users at last check in 2014, and you can't get more basic and easy to understand than search.
But outliers like Google Plus, Wave, and Glass, do stick out as flops because of their complications or geekiness.
"People pay thousands of dollars to let lasers onto their eyes to avoid putting things on their faces," another former employee griped. "And then look what they did with Google Glass! C'mon."