What is Google?
In many of the conversations I have online with commenters on my posts or on Twitter, it becomes clear people have very different conceptions of what Google is as a company. I’ve heard at least the following descriptions:
- “Google is an advertising company”
- “Google is a search company”
- “Google is a data company”
And so on.
The reality is there are at least three different ways of looking at Google as a company. I wanted to explore each of those briefly with a view to seeing what each of these lenses can teach us about why Google does what it does, where it might be going next, and what its future prospects might be.
Lens 1: Google is a search company
This first lens is a product lens: it shows Google through the products it provides and the most well known and biggest driver of revenue is search. Google doesn’t break out its revenue by product, but it does distinguish between ad revenues from its own websites and ad revenue from its network. Revenue from its own sites is about two thirds of its total revenues. Much of that likely comes from search with YouTube, Gmail, and others making up the rest. Search is likely also the most profitable part of Google’s business – again, it doesn’t break out profitability by product, but there are other signals to suggest this business is the healthiest. To a great extent, search advertising is Google’s highly profitable core, an idea I talked about two weeks ago.
This lens works beyond just the financial picture too: search is a key component of Google’s portfolio in other ways. Search is the window onto many other Google services, whether YouTube, Maps, shopping or Gmail. Search now operates across these properties and it’s a key feature within those other properties as well. Google is still seen first and foremost as a search engine that has other products.
So where would a future rooted in search take Google next? Well, into searching more and more domains. It’s already taken Google beyond websites and into Maps, items for sale, emails, books, videos, images, and much more. It’s now moving into search within apps, as announced at I/O, and it’s easy to imagine it extending further into both real world and digital information that is currently inaccessible. But Google has also been changing the way it responds to search queries, from pointing people to results elsewhere to presenting answers itself.
And some of Google’s activities are very hard to fit into a search-centric view of the world. While search itself, Gmail, Maps and YouTube all arguably come into focus through this lens, others such as Calico, Google’s investments in robotics, Google Fiber and self-driving cars remain stubbornly blurry. Can another lens help to make sense of those?
Lens 2: Google is an advertising company
Many more people describe Google as an advertising company and I often hear this description applied pejoratively. But this too fits and, arguably in some ways, fits better than search. One of my favorite quotes from Steven Levy’s book on Google, In the Plex, is from Andy Rubin:
“We don’t monetize the thing we create… We monetize the people that use it. The more people that use our products, the more opportunity we have to advertise to them.”
This lens, and Andy Rubin’s description of it, is primarily about business models – i.e. how Google monetizes its products and services. A business model lens can be a phenomenally useful way of thinking about companies, as I wrote a few months ago. From a financial perspective, advertising dominates Google: until about a year ago, Google derived over 90% of its revenue from advertising and even now the number is just barely under 90% (I’m ignoring the brief ownership of Motorola). The vast majority of Google’s revenue comes from advertising. This is clearly its major business model and so this is an entirely acceptable way to describe Google.
Again, many of Google’s products also fit within this way of looking at the company. Almost all of what it does has an advertising component: Search, Gmail, YouTube, Maps, etc. However, there are products where Google doesn’t serve up ads or it monetizes in other ways: Calendar, Android on both phones and newer devices, Chromecast and so on. And when you start to apply this framework to some of its newer initiatives, it breaks down almost entirely: Calico, robotics, Google Fiber and self-driving cars all fall outside the scope of ad-supported services.
Lens 3: Google is a machine learning and AI company
There is one last lens which I almost never hear anyone talking about, but which I think is actually the most useful way to think about Google. Google is first and foremost a machine learning and artificial intelligence company. The problem with this description is it’s very hard to see directly either in Google’s product portfolio or in its finances. Google’s products revolve around search and its revenue comes principally from advertising but, behind the scenes, essentially all of its products either have a foundation of machine learning and AI, or help to drive Google’s effectiveness in machine learning and AI. If the search description is about products, and the ads description is about business models, this description is about core competencies.
Of course, this is where Google started: Larry Page and Sergey Brin didn’t set out to build a search engine or an online advertising powerhouse: they set out to organize the world’s information and make it accessible, which was fundamentally a data ingestion and machine learning problem. Essentially all the iterations of the Google search engine since the early PageRank systems have come through machine learning that better interprets user queries and finds the most appropriate results. But of course, machine learning is used throughout Google’s other products too, both for search and for other purposes. Entering a new Calendar entry is made easier by interpreting normal human language. Google Now responds to queries and even strings of queries almost like a human being.
But here’s the key thing: a machine learning and AI lens is the only one of the three that helps to make sense of Google’s more recent investments and acquisitions. DeepMind, Nest, US Robotics, Calico, Self-driving cars, etc., can all be explained through the lens of machine learning and AI as the core of Google, but are unexplainable by reference to either business models (advertising) or past product focus (search). The chart below illustrates this starkly by showing in red those areas where a product, service or investment matches a lens:
This doesn’t mean we should ignore either the role of search in the development of Google as a company and its product focus today or the role of advertising as the main monetization model for Google. But, in my view, Google is best thought of as a machine learning and artificial intelligence company which monetizes primarily through advertising and whose most successful and lucrative product *to date* is search.
The problem with this view
The major problem with this view of Google is it tells you little about Google’s future products beyond those already public, and even less about future business models. If Google isn’t a search company, and advertising isn’t its only business model, how will it monetize these future investments in machine learning and AI? How will Calico, robotics, self-driving cars and these other initiatives ever pay off? Advertising doesn’t seem like the answer for any of these initiatives, but the bigger issue is it’s not yet clear what the business model is for any of these things. That’s one of my biggest concerns about Google: though this lens helps make sense of Google’s recent activities from a core competencies perspective, it doesn’t help us figure out where Google’s money will come from in the future. Away from both search and advertising, which have a very clear track record, what evidence do we have Google’s future will be as lucrative? For all that, I’m comforted by this understanding of what Google is as a company, because it makes sense of what it’s doing. I’m still concerned about its ability to turn profits from most of those activities, though.