The test was a basic puzzle. It was the sort of question an intro computer-science student could solve, and the applicant, Robert Jacobson, was a math professor with a Ph.D. in complex variables. It should have been straightforward, and Jacobson should have aced the entry-level coding screen. But with the clock ticking away, the puzzle became an impossible task, and he froze. At the end of the hour, he bombed it completely.
The failed interview sent the former Roger Williams University professor into a tailspin. He'd declined the months of interview prep time offered by Google and had done little more than glance at the hundreds of pages of prep documents sent by the recruiter. He'd been offended when they offered this help, upset that the company would expect anyone to offer that amount of time and study just for the privilege of an interview, let alone ask it of someone who taught the offered materials to his own students.
So, suspended in a state of defensive fear, feeling like he had to explain himself somehow, Jacobson sent an email to the recruiter detailing his cognitive differences: among them, an anxiety disorder and ADHD.
"There's this natural sort of need to justify yourself and to explain like what happened. And so I said, it was really difficult for someone like me, who has ADHD, to perform in this kind of environment. And then, the next day, after I had calmed down, I realized what I had done. I had violated my very strongly felt dictum that I would not and should not have to disclose any of my cognitive differences to my employer or any potential employer, but here I am feeling like I need to justify myself, like I need to explain why I did so terribly," he said.
Jacobson has no trouble admitting he totally blew a basic test at the beginning of an interview for a fairly elite job at Google. He has no expectation that had he passed the test, he would have earned a place at the company down the line. He has a problem with the fact that a company with the power, prestige and wealth of Google has developed a recruiting process that is so large and systematic that it can both ask for large amounts of time and energy from prospective candidates and then easily or accidentally hurt or dismiss those same qualified candidates because of a difference, like Jacobson's cognitive disorder. (Google does provide accessibility accommodations for people who ask for them during the recruiting process).
Protocol spoke with a range of applicants to technical jobs at Google, all of whom had similar perspectives on the experience, despite widely varying outcomes: The company asks that prospective applicants give enormous amounts of time and energy to the process, while often purportedly failing to return the same commitment because of the sheer number of people who are applying for jobs. While this tension is not unique to Google, everyone interviewed for this story said that the issues are more widely discussed and pronounced at the company than any large competitor (like Facebook), and that they believe that Google has a unique level of prestige that should create a corresponding sense of responsibility.
"To help set up candidates for success during the process, we provide them with interview guides, and video tutorials sharing best practices and tips from our careers site. Our interviewers receive specialized training and over the years we've worked to improve the pace at which we review candidates and hire, for example by reducing the number of interviews per candidate," a Google spokesperson wrote in a statement to Protocol.
The competition for the top tech talent is intense among the largest tech companies, and the demand to fill ever-increasing numbers of jobs (Google interviews millions of people every year) make the technical recruiting process for a company like Google objectively difficult. "It isn't really a human process. It doesn't really rely on understanding the individuals. It's almost a production line that you're trying to automate. I think that's basically what they've got, a recruitment process that's very much scientific management," said Stuart Watt, currently the chief technology officer at AI-startup Turalt and a former prospective Google applicant. "Google is applying [scientific management] to people, that's the fatal mistake that they've made. They are applying metrics and measurements to people without understanding that it has consequences."
Watt personally liked his recruiter, took the recommended time to study for the interviews, and, during two separate interview processes, aced the initial screen and was then flown once to Dublin and once to Google in Silicon Valley for a second day of interviews. It was the design of the recruiting system itself that became his problem: At some point, Google decided Watt (who was working as an associate professor at the time) was interviewing to be a site reliability engineer, despite never indicating an interest in that role himself (he was interested in a more management-focused position).
"It just never seemed to get through. They were so focused on whatever categorization they had chosen and it was fixed," he said. "I think something in their processes meant they weren't really looking for a fit between a person and a job. It felt to me that they probably had a recruiter who was looking for a certain role. Once they put you in the pipeline, that's the role you're in."
No one interviewed for this piece — and almost no one on the many Twitter and Reddit threads discussing issues with Google interviews at great length — blamed specific recruiters or interviewers for any of their particular issues with the process. "None of the people I felt had an agenda, or were kind of aware of how the whole thing behaved. They kind of become cogs as part of the machine, and nobody knows how the whole thing works," Watt said.
The Google process can take many months for some people (and demand months of prep work beforehand), making it sometimes inaccessible for people in dire financial need, recently unemployed or working another demanding job at the time. This inaccessibility could eliminate the types of people from the process who would be otherwise qualified for the job, and these candidates may also provide the diversity in experience and socioeconomic background that Google professes to seek.
"Since most engineers are male, and since most interviewers and hiring committees are more senior, and senior engineers are more likely to be male, you end up with a pretty skewed panel," said one developer advocate who has interviewed twice with the company and asked not to be named for fear of harm to her future job prospects. She often mentors other women or young people who identify as members of underrepresented groups when they prepare for interviews at big tech companies, and she usually advocates against interviewing at Google. "Since Google is a very large [organization] with lots of entrenched leadership, it's difficult to effectively advocate for any changes in the interview process," she said.
A Google spokesperson told Protocol that the company is committed to continuing to invest in onboarding and recruiting programs, and to building a diverse and inclusive culture.
The sometimes brutal nature of the tech recruiting machine itself is not unique to Google and has been well documented. Complex interview questions about algorithms, and both physical and virtual "whiteboarding" tests (like Jacobson's) create the kinds of high-pressure environments that don't necessarily have any relationship to the actual job. While they are intended to examine whether someone actually has the skills they need to do the work (an important question tech companies always need to ask during the hiring process), they usually don't reflect the way most people actually code, which usually includes looking up and using existing resources. Basecamp founder David Heinemeier Hansson once famously said, "I would fail to write bubble sort on a whiteboard. I look code up on the internet all the time. I don't do riddles."
The whole constellation of issues means that some people who can afford to take the time to do the prep, like CloudKite.io founder Victor Trac, approach the Google interview like an educational opportunity instead of a job hunt. He wasn't in need of a job, so he took two months to review the hundreds of pages of prep materials that Google advises candidates study.
"I learned a lot. And it was exactly what I was expecting," he said. "It's a very thorough process, so there's lots of false negatives. I'm sure lots of very highly qualified people don't get offers. And it's a grueling day too, and you're there in the office all day. And yeah, you get tired and your brain's not working, and now you have to whiteboard something or do some behavioral situational interview."
While Trac sees the Google process as standard, that doesn't mean he likes it. "I think it really is for most people a prestige thing. That's something Google can take advantage of for so long," he said. Trac's own startup offers asynchronous interviews with real-world problems for candidates to solve. This way, the process tests applicants' abilities to do the job at hand without asking people to take time to study, answer questions that might not have any relevance or create a scenario that exacerbates a cognitive difference.
"A lot of it depends on the interviewer you get; it just takes one person for it to go wrong," he said. "Google has this reputation, this 'try again in six months.' I've talked to a bunch of people, it took two or three times before they got that offer. Then you are just extending it to a year or two-year-long process."
"If anybody had gone and called people up and interviewed them a year later, I actually think [Google] would know what is going on and maybe be able to improve their processes. And I genuinely think that would improve their ability to get the people they want. It's almost damaging their ability to recruit good people," Watt said.
