I generally look at Google as a vast machine learning engine that’s been stuffed with data for a decade and a half. Everything that Google does is about reach for that underlying engine - reach to get data in and reach to surface it out. The legacy web search is just one expression of that, and so is the search advertising, and so are Gmail and Maps - they’re all built onto that underlying asset.
Hence, most of the experiments that Google has launched over the years are best seen as tests to see if they fit this model. Can you apply a vast expertise in understanding data, large numbers of computer scientists and data scientists, lots of infrastructure and a model of total automation and get something interesting and useful - can you get massive amounts of new data in, can you do something unique with it, and can you surface it back out? And, for all of these, are you solving hard, important problems with global scale?
That is, Google tests new opportunities to see if they fit in the same way that a shark bites a surfer to see if they’re a seal. If not, you don’t change Google to fit the opportunity - you spit out the surfer (or what’s left of him).
Naturally, sometimes it turns out that you need other capabilities (radio advertising). Sometimes it ought to be a good opportunity but the friction in actually unlocking the data is too great (Google Health, where there were too many different and reluctant parties involved). Sometimes Google's skills are just a condition of entry and other skills are more important (Google Plus in social), and sometimes the opportunity is just too small (Google Reader). But equally, there are projects for which Google's core skills and needs have fit very well. Maps had little obvious to do with web search and nothing to do with PageRank, but were a big problem where Google’s assets could be applied (and of course, a decade later, Maps turned out to have huge strategic leverage in mobile). The same may be true of self-driving cars - this is not a search question, but it is a data and machine intelligence problem where Google is uniquely placed to do things (or at any rate, that’s what Google believes).
Android embodies all of this. Originally, it was about reach, in the sense of people being able to use Google. It existed to head off domination of mobile by any third party that might shut Google out (first Microsoft, then Apple) and to enable the expansion of the internet from 1.5bn PCs owned by relatively rich people to, in the next few years from now, 4-5bn mobile phones owned by almost every adult on earth. In both of those aims it’s been an enormous success, much more than any other Google side-project. Apple will have a lot of the top 15-20% of the market, but Android will have almost all of the rest and serves to keep Apple honest even on iOS.
Over time, Android has also evolved to provide reach in collecting data as well - you’re always logged in to Google on your Android phone, and it knows where you are when you do that search or open that app, and where everyone else who ever did that search was, and what they did next (this is one reason why retaining control of the Android UI, and heading off forks, matters to Google). There’s an old computer science saying that a computer should never ask a question that it should be able to work out the answer to - the sensors and other capabilities in smartphones in general and Android in particular massively expand the range of things that Google can work out. So, Android transforms Google’s reach both in collecting and surfacing data.
The interesting part, though, is that there are now lots of different kinds of reach.
First, as everyone has talked about for years, the way that mobile moves us away from the plain old web as the dominant interaction model of the internet challenges Google’s central ability to understand the structure of online information and to link to it (and sell links to it). Apps cut off Google’s reach, both to get data into its systems, since apps are opaque, and to surface data out to internet users, since any search in Yelp’s specialist app is a search that wasn’t on Google, and such apps are stronger on mobile than on the desktop. Apps reduce Google’s reach in both senses. This of course is why (like Facebook) it has been pursuing deep links, and is probably also one reason why it is keeping Chrome OS warm as a standby mobile platform. But it also means that Google has conflicting incentives - as a provider of services, should it try always to make things as part of the web, or embrace the new experiences that apps and everything else happening on smartphones can provide? What would the web search team say if Hangouts became a development platform, for example?
Furthermore, ownership of an actual mobile platform creates more basic conflicts - should Google make a new app for iOS first, given that’s where many of the most engaged users are? Should it provide it for ‘forked' devices such as the Kindle Fire, if they have enough users, though that erodes a point of leverage for control of Android? For these kinds of questions it’s easy to see how individual product managers might have incompatible objectives - the Maps PM probably wants Maps to be great on iOS and might well like it on Kindle, but people thinking about maintaining Google’s control over Android clearly would not.
That is, how much does Google need to pull things to the web, or, alternatively, to Android, and how much should it let the logic of individual product teams prevail - and where there's conflict between product teams, who wins?
These are really classic ‘strategy tax’ questions. A product feature conflicts with the company’s overall strategy, so do you leave out the feature (and so pay the strategy tax) or compromise the strategy? To give examples from other ages of the tech industry, Microsoft's Office for Mac and Apple's iTunes for Windows both in theory undermined those companies' core products, but both were the right thing to do for the broader strategy. The question for Google is to work out which part is the tactic and which is the strategy. What kind of reach do you want, and which sacrifices to you want to make?
This problem reminds me of a book published a few years ago by Pierre Bayard, a French academic, called ‘How to talk about books you haven’t read’ (review). His observation is that the question ‘have you read this book?’ is actually much less binary than it appears: if you compare a book you read as a teenager 20 years ago and half-understood with a book that’s just come out and that you’ve read 3 reviews of, but haven’t actually read, you might know rather more about the latter than the former. There are books you've read and understood, books you’ve read and half-remember, books you can’t remember that you’ve read at all, and books that you’ve read half of, or know the key ideas of, or have heard about, or that you know are exactly the same as three others by the same author that you really have read. Reading a book and knowing about a book are not the same thing.
In the same sense, Google needs reach, but mobile means that there are lots of different kinds of reach. Consider someone who has an ‘official’ Android phone, perhaps even a Nexus, and is completely logged in - so Google has ‘perfect’ reach to them as an end-point. But, as I wrote here, suppose they live in a quiet suburb and drive only to work and to a few shops, never use Calendar, open Maps once a month and get a few personal emails in Gmail each week. Now contrast that with a 20-something in a big city who loves their iPhone and is not logged into any Google service - but is on this phone for hours every day, uses Google Maps (or maybe just apps that embed it) and is doing web search all the time. What kind of reach does Google have for these two?
Then, consider a farmer in rural Myanmar who’s just got their first phone: a $30 Android, with enough spending power to get perhaps 50 megs of cellular data a month, if that. What is that reach worth - what do they search for, what can the information they provide to Google be used for and, to raise the boring, pedantic question, how much are they worth to the advertising industry? Are they a higher priority than extending Google Now to the Apple Watch?
The key change in all of this, I think, is that Google has gone from a world of almost perfect clarity - a text search box, a web-link index, a middle-class family’s home - to one of perfect complexity - every possible kind of user, device, access and data type. It’s gone from a firehose to a rain storm. But on the other hand, no-one knows water like Google. No-one else has the same lead in building understanding of how to deal with this. Hence, I think, one should think of every app, service, drive and platform from Google not so much as channels that might conflict but as varying end-points to a unified underlying strategy, which one might characterize as ‘know a lot about how to know a lot’.
Another way to think about this, perhaps, is the comparison with Internet Explorer. Microsoft was entirely successful in ensuring that its own web browser dominated the internet for a decade or so. But really, whatever browser people ran, they were going to run it on a Windows PC anyway, because what other mass-market global platform was there? So too for Google: what matters is to win at 'search', whatever that means and wherever and however far from PageRank that leads you. Nail that and reach will come to you - get it wrong, or find yourself irrelevant in whatever the new new is (as happened eventually to Microsoft), and nothing else will matter.