The speculative nature of these investments causes moments of frustration for some analysts, who are, after all, paid to provide opinions on whether or not Google is worth more than $1,000 a share. (It closed Friday at $1,015.20, giving the company a market capitalization of $339-billion.) It’s something its CEO acknowledges, while making no apologies for the company’s unconventional approach to spending money.
“People rightly ask how we’ll make money from these big bets,” Mr. Page told shareholders in a 2012 letter. “We understand the need to balance our short– and longer-term needs because our revenue is the engine that funds all our innovation. But over time, our emerging high-usage products will likely generate significant new revenue streams for Google as well as for our partners, just as search does today.”
While waiting for the payoff, however, Google has to keep its cash cow healthy, especially as users switch to the mobile Internet, where ad rates are lower.
The company’s current struggle to re-ignite click rates is magnified by it prior success. For five successive quarters, the search engine has failed to post growth rates in its core business of more than 20 per cent. For most companies, such rates would be considered astronomical, but as BGC Financial analyst Colin Gillis noted recently, just two years ago Google’s core revenue growth rate was around 35 per cent. “We see the slowing core business as a one reason why Google is investing so heavily in new ventures as new products are needed to re-ignite revenue growth.”
That’s why, as part of Google’s attempt to remain relevant in its next 15 years, even its core Internet business is undergoing a massive overhaul.
The Knowledge Graph
On a whiteboard, Ben Gomes sketches an illustration of the inner workings of Google’s most prized possession – its search engine.
Since its inception in 1998, Google has collected information. But for years, that information has been stored in separate databases depending on its content – from books to video to images.
Today, Mr. Gomes, who has been with the company for 14 years and is now one of the chiefs of its search department, is overseeing an effort to combine all of Google’s knowledge into a single entity, called the Knowledge Graph.
As an engineering task, the Knowledge Graph is substantial. Already, the data include some 18 billion items collected by Google from internal and external sources such as Wikipedia. It is essentially a database of the relationships among the world’s people, places and things. For example, Barack and Michelle Obama are connected by the relationship “spouse,” whereas Barack Obama and the value “185 centimetres” are connected by the relationship “height.”
Millions upon millions of these linkages exist in Google’s Knowledge Graph. The graph contains the connections between stores and their opening hours, companies and their market capitalizationss, singers and their discographies. In effect, it is Google’s most overarching model of the real world. (The data also allow Google to answer some natural language queries. Typing “How tall is Barack Obama?” into the search engine, for example, produces a box that says “1.85 m”.)
At its core, the Knowledge Graph represents the fullest realization yet of Google’s original mission statement – to collect and organize the world’s information. Perhaps the most beneficial use of the Knowledge Graph for Google today is its ability to facilitate voice-based searching from mobile devices – where a user simply asks his or her phone a question in plain language, and the phone responds instantaneously.
“In the future you want to give the user the answer as if they were interacting with a real person,” Mr. Gomes says. “But to do that … the computer must begin understanding a model of the world.”
It is, in other words, a prime example of the future Google hopes for – one in which the company is the middleman in a digital transaction that feels no different from a real-world one.
Like many of Google’s endeavours, the Knowledge Graph is in many ways a project without an end. Engineers continue to refine and expand the database, constantly on the lookout for new places to use it. That strategy – innovate first and monetize later – has been the Google way since the company’s birth. Indeed, when Google executives sold the first ad on the company’s search results page, few could have foreseen the same ads one day running on a YouTube videos, Gmail messages or high-powered smartphones.
Now, the question is whether Google can stumble onto the next profitable technology before its current one stops making money.