What Google Does Right

I’ve appreciated Google’s mission and its modus operandi for a long time now.  I’ve avidly read Planet Google, many Wired articles, and a number of blogs and other pieces about the company.  But what I want to address here is how Google provides a great user experience, what enables it as a company to follow the path it does, and what smaller companies can learn from it.

It’s easy to state but hard to do right, and often requires deep design to accomplish, but it’s one thing that Google does extremely well: it keeps its interfaces simple.  This ability is, for sure, enabled and exploited by the very nature of the company: it’s a web-based company through and through.  That means it can radically simplify so many things that mass consumer computing users find so hard: a big example being navigating and using a file-based storage system.  Instead, of course, everything is stored by Google (in its “cloud”, if you like) – and this simply obviates the need for a Save button, a Load Button, and all that junk.  Nor do you need any IT infrastructure to use most of Google’s products (email, for example).  By saving your documents or writing automatically – quietly and regularly, the way it should be done – the user never needs to even think about the where or how of storing data.  Except, that is, if you need to categorise – but here again Google makes sure that its core capacity – that of Search – is always front and center and powerful enough to find whatever you need.  A user experience should aim to empower the user, not baffle or frustrate them, and in this regard Google generally succeeds admirably.

By providing many of its products free to the mass consumer market, Google owes its audience nothing.  This gives it free rein to change and improve (in short, to innovate).  By having lots of small but focused products, it can bring on or cull away products quickly (generally at the lightning-quick speed of the web world, and impressively fast for such a big company).  Here again Google understands right in its DNA both freemium and the web’s “Everything, free” tendencies.  Google is also very good at knowing what to keep hidden – its apps are great at hiding functionality that is less relevant to day-to-day usage from the user (they’re often there, but you have to dig a little to find them).

Google as a company has shown an almost terrifying ability to grow, but to grow without collapsing under its own weight.  One way that they do this is by – in the main – using low cost easily available hardware (which has financial benefits as well as intangible benefits), even in huge data centers; a Commodity Computing approach (they even store their servers in shipping containers).  Development has an open feel to it, and is often open sourced or provides public platforms and APIs; Google Labs and techniques that expose Beta versions show Google developing software often in public view – compare to the secrecy that often surrounds Apple development.  Product support is often scaled by using open forums where members of the public helps each other.  Internally there is an almost astounding lack of management hierarchies.  In fact one could conjecture that Google is probably not really a big company as such, but a network of highly connected small companies that share common DNA and some common base technologies (often through open sourcing or open standards).  The shelter of the larger entity (not to mention its profitability) give it the ability to take risks – if one of the smaller companies/products fails, it can be easily absorbed.

Of course, Google’s flagship Web Search also scales (it has to, to have any chance of covering billions of web pages).  But interestingly, it seems to me that the success of Google’s PageRank algorithm – the core of it’s Web Search function – is largely because at heart the algorithm combines both the human and the machine in a very effective way – the human aspect is the importance of a page due to linking (a result of human activity) along with a series of quantifications (the rank).  But we are now also starting to see meta-data aware algorithms, that are getting nearer to natural speech, for example the Wolfram Alpha service (a so-called “computational engine”).  You can be sure that if Google truly cracks the problem of natural language search (which may or may not be equivalent to – perhaps a very dumb – AI), it will change the world (again!).  Indeed Google’s founders have stated that Google’s aim is to develop an Artificial Intelligence by way of Search, and there have been some startling successes: Google’s language translation service is apparently very good, and has resulted from a statistical approach enabled by massive data sets.

Google might be construed as an “information” company – after all, it wants to “provide access to all the world’s data”.  But there’s an important distinction to be made here – Google only cares about data insomuch as it is useful to someone (typically, consumers or businesses) – it does not care about information per se.  (That’s not to say Google will lose your data!).  The point is that Google is above all a technology company – it is enabling and automating the use of technology, predominantly software, but increasingly also hardware, to solve all sorts of engineering problems, and lots of data just happen to be the input.  Storing millions of search results, using millions of documents in different languages in order to automate translation, and many other examples support this view.  Google as a company is a master engineering problem solver, including solving some of its own internal problems.  Many of its products are happy accidents, or the results of its famous “20% time”, where its employees are given one day a week to pursue their own interests.  Google is like a giant R&D lab that also happens to be a corporation.  It also places huge importance on hiring the right people (smart ones), because it knows that great solutions come from clever minds – in fact CEO Larry Page personally signs off on every new hire.

The humble web start-up right through to the big unwieldy enterprise can learn much from Google’s approach, particularly if your products or services are targeting the mass consumer or massive business arenas: