What is more important to a successful online search business: the computing algorithm that decides what results to display or the data that tracks the results of user clicks?
Even within Alphabet Inc’s Google, the world’s largest search engine, that question has been hotly debated for years. And now it’s a key feature in a landmark antitrust trial, where the US Justice Department claims Google spends billions of dollars to stifle competition and preserve its monopoly over online search.
Google’s Chief Economist Hal Varian, who testified during the first days of the trial this week and has been a key spokesman since 2002, began arguing publicly in 2009 that the company’s algorithm was more important than the amount of data collected, referred to as “scale”, according to documents presented as government evidence in the case Wednesday.
“The scale arguments are pretty bogus in our view because it’s not the quantity or quality of the ingredients that make a difference, it’s the recipes,” Varian told tech news publication CNET in 2009. He made a similar point in a 2011 email to a New York Times journalist: “It’s certainly true that data is important for search engines but it is not the entire story, knowing what to do with that data is more important,” according to the trial evidence.
But Varian’s view didn’t sit well with some colleagues, including Google’s chief scientist Peter Norvig, who was the former head of search quality, according to documents presented by the government.
“We don’t have better algorithms than anyone else,” Norvig said at a March 2010 public event, according to the trial evidence. “We just have more data.”
Arguments about the scale of data collected by Google are key to the government’s case. Antitrust enforcers allege Google illegally dominates online search by paying as much as US$10bil (RM46.81bil) a year to be the default option on web browsers and smartphones.
Those agreements prevented rivals like Microsoft Corp and DuckDuckGo from gaining enough data to effectively compete, the US alleges. Google gets 16 times as much data as its next closest competitor, Microsoft’s Bing, the Justice Department said in its opening statement Tuesday.
The trial, expected to last through mid-November, focuses only on whether Google violated the law. The Justice Department hasn’t yet indicated what remedy it might seek at a second proceeding if US District Judge Amit Mehta rules in its favour. The government could push for the biggest forced breakup of a US company since AT&T was dismantled in 1984. Another possibility would be forcing Google to share some of its data to help rival search engines improve their quality.
The company denies it thwarts competition. Its attorneys told the judge Tuesday that there are other options on the Internet, and users choose Google because it has the best search tool.
But senior executives argued for the importance of data in internal emails, government evidence showed.
“It’s absolutely not true that scale is not important,” Google’s head of search, Udi Manber, who joined the company in 2006, wrote in a 2009 email to Varian.
‘That’s a fact’
“The bottom line is this,” Manber wrote. “If Microsoft had the same traffic we have their quality will improve significantly and if we had that same traffic they have, ours will drop significantly. That’s a fact.”
In another exchange in 2020 with Google engineer Daniel Russell, Varian said that “journalists and regulators” weren’t sophisticated enough to grasp his point that the quantity of data is less important than the algorithm, according to the trial evidence.
“They believe if we just handed Bing a billion long-tail queries they would magically become a lot better,” Varian wrote.
Russell replied: “The data quality effect IS real.”
In his testimony this week, Varian said he still believes the algorithm is more important, but has never said data isn’t also a factor.
“Data is valuable,” Varian said. “It’s not the be-all and end-all. Human skills are valuable. Scale is important but it’s not the only thing.”
– Bloomberg
-TheStar