A search engine that can’t keep up with the internet’s ever-expanding data sets is causing engineers to work harder and harder to get results.
In a new study published in Science, the authors found that Google engineers were spending an average of more than 15 hours per week dealing with a large data set, compared to a mere 3 hours per day during the same period last year.
Google engineers also spent more than 40% of their time on network engineering tasks, compared with less than 20% in the previous study.
The study also found that engineers spend an average 20 hours per month on the internet.
“Google is the primary engine that we use to do our search,” said the study’s lead author, Michael Schott, a data scientist at Stanford University’s Information Science and Engineering Department.
“It’s the one that we see and feel the most, and that’s why we’re spending a lot of time there.”
The findings echo those of a separate Stanford study released last year, which found that nearly half of the engineering teams in Google’s US office had spent more time dealing with data than any other department.
Google’s engineers are among the most dedicated in the world, but they’re working harder than ever to keep up, said Michael P. Osterman, a computer scientist at the University of North Carolina.
“We have to get this stuff done, and we have to keep pushing the envelope,” Ostermann said.
“That’s how we can get to a world where you can have an enormous amount of information available to us.”
But Ostermaier added that Google has a lot to learn from other organizations.
“When Google started out in 1995, they were basically an offshoot of Microsoft,” he said.
The company launched a search engine, called Search, in 1997 and focused on the “core search” part of its business.
Google started to make a name for itself in 2006, and its core search business has expanded to include everything from news to sports to shopping.
But the company has faced criticism for how it has handled data.
“People say it’s like Microsoft, and I think they’re absolutely right,” Otermann said, “but it’s not like that.”
For the new study, the Stanford team looked at how much data Google engineers had spent on each task during the last three years.
“Our research shows that Google’s core search team is the most productive of any engineering group in the US,” Schott said.
But when the researchers took a look at how Google employees were performing during other tasks, it was clear that Google employees didn’t do nearly as well as the data analysts in other departments.
Ostersman said that this pattern is not unique to Google.
“A lot of other large companies, Google is a great example of that,” he added.
The researchers found that the average engineer spent an average 3.7 hours per workday on data analysis.
That’s about the same amount of time as a data analyst would spend on their own data analysis tasks.
But engineers spent more hours on network work, which included handling DNS requests and managing network traffic.
Google also spent about 4 hours per year on network management, which includes keeping track of what web pages are currently being visited and how long they take to load.
That is likely because the company is looking at how it can improve how users can navigate the web, Ostermans said.
Oesterson said that while Google’s data analytics teams are “in a very good position” to handle the vast amounts of data that Google is working with, there is still room for improvement.
“They’re probably not the best of the best in this regard, but there’s still room to improve,” he noted.
Google said it is constantly reviewing its systems and processes to make sure they can deliver the best experience to users, and is looking for ways to improve.
“In the future, we will make better use of the Google APIs, and continue to build better solutions for data,” a Google spokesperson said in a statement.
The Stanford study found that engineering tasks that involved networking data and managing DNS traffic were most often performed by people who are not technically adept, such as network engineers who specialize in network troubleshooting.
Other tasks that involve complex network engineering, such like data analytics, were performed by engineering teams who are “highly technical,” Schotts said.
Schott agreed, saying that engineers are doing more work on complex tasks that require high levels of knowledge.
“Engineers are working harder and doing more data-intensive tasks,” he told ABC News.
“The question is, how much more?”
He added that the findings of the Stanford study “call into question” the notion that Google should be “an enabler of the Internet’s growth.”
“It would be nice to have the ability to do network engineering at Google, but the reality is that network engineering is not a very important part of Google’s business,” Schotte said.