“It’s interesting that the founders left Google, spent 2.5 years building something, then get acquired by their former company for $50 million. What is going wrong at Google that is costing them this kind of ‘innovation premium’?”
– David Lifson, Google Acquires Aardvark For $50 million
Far be it from me to defend Google’s acquisition practices. It’s difficult to assess the relative failure rate of their acquisitions externally, not least because it’s difficult to judge the process of acquisitions generally. Some are about the team, some the technology, some the market, some the data, some all of the above. Only an acquirer knows for sure what the motivation is behind a given transaction. It is clear, however, that Google’s assimilation processes are – at best – less efficient than we see from more mature technology firms. From ugly exits like Dodgeball to the lengthy integration time for GrandCentral, in the area of mergers and acquisitions, Google – a firm that prides itself on its ruthless pursuit of efficiency – would receive what my kindergarten teachers called an “NI.”
That said, I think the conventional wisdom, which seems to argue that the purchase of Aardvark represents a gross market inefficiency, is worth questioning.
Implicit in that assertion are a number of arguments. First, that Google is, in fact, paying a premium on the actual development costs, customer acquisition and – perhaps most importantly – the data generated. Second, that Google should be creating an environment that allowed these startups to grow internally. Last, that the economic model of losing employees, then subsequently reacquiring them and their creation is inherently flawed.
The calculus of the first assertion is difficult, given that we lack data on the employee count, employee costs, infrastructure & marketing costs, and so on. What we think we know – via CrunchBase – is that the firm took on first $750K, then $5.25M in funding. We do know (PDF) that in the seven months following their March 2009 beta, they grew to ~90K users who asked ~225K questions and received ~386K answers. We also know that their engineers spent just shy of three years building their product.
The question then is this: is the product of two and a half years development time, several hundred thousand questions and answers, and the insight on the interaction model that generated that worth fifty million dollars? For you or me, the answer is probably not. For Google, the data alone could be forth tens of millions of dollars. Throw in the product, the team and the users, and the acquisition price might well be fair.
But should Google be fostering an environment that allows these startups to flower internally rather than externally? In hindsight, it’s not much of a trick to see the attraction of this idea: do you think Google would like to have Twitter back? But here’s the problem with that: this assumes that Google’s organizational processes are as or more efficient than the market. Consider: Google hires smart people, than gives them the scheduling latitude with their 20% time to tinker and experiment. Some of these bear fruit, as in Buchheit’s Gmail. But remember how many Googlers there are now: around 19,835 at this writing. Even allowing for the fact that a large fraction of that population aren’t developers and thus not likely to be working on pet projects, that’s a very large volume of potential innovation. Which Google itself has acknowledged struggling with. From the Wall St Journal:
So with the U.S. economy in a recession, Google is ratcheting back spending and cutting new projects. “We have to behave as though we don’t know” what’s going to happen, says Google Chief Executive Eric Schmidt. The company will curtail the “dark matter,” he says, projects that “haven’t really caught on” and “aren’t really that exciting.” He says the company is “not going to give” an engineer 20 people to work with on certain experimental projects anymore.”
Translation: instead of giving an engineer 20 people to work with, they’ll let him walk and let the market cull the winning projects from the losers. Google, in other words, has fully transcended its startup roots, and embraced the tried and true big company innovation strategy: acquisition. The benefits to this approach, as Paul Graham nicely articulates, are many. For both parties.
Big companies also lose because they usually only build one of each thing. When you only have one Web browser, you can’t do anything really risky with it. If ten different startups design ten different Web browsers and you take the best, you’ll probably get something better.
The more general version of this problem is that there are too many new ideas for companies to explore them all. There might be 500 startups right now who think they’re making something Microsoft might buy. Even Microsoft probably couldn’t manage 500 development projects in-house.
The thinking behind an Aardvark acquisition, then, isn’t that different from what we just saw with their Super Bowl Ad: rather than guess, they let the market tell them which spot to run. Which is nothing more than what Schmidt told us to expect two years ago:
[Schmidt] believes that because Google is “run by three computer scientists we’re going to make all the mistakes computer scientists running a company would make. But one of the mistakes we’re not going to make is the mistake that non-scientists make. We’re going to make mistakes based on facts and data and analysis.”
Understanding that it cannot match the market for efficiency in picking winning projects from losing projects, Google’s effectively outsourcing the triage and the risk to would be users and customers.
In their perfect world, Google undoubtedly identifies and picks the winners on the open marketplace before we all realize – as with Twitter – that they’re the winners, but that’s a difficult game to win consistently.
Is this strategy, then, the one that sees the Aardvark’s of the world seeing $50M exits inherently flawed? Only Google can say for sure.
Big picture, what this all means is this: Google, intentionally or otherwise, loses lots of developers. A portion of that population start companies. Most of those fail. Of the ones left, some are interesting for the technology, some for the data, some for the people, and some for all of the above. But, just as with its Super Bowl ad, Google isn’t guessing any more at the value: it knows, or at least has a pretty good idea. My guess is that this acquisition, like everything Google does, is being done by the numbers. What appears to be a premium is actually a reasonable, in context, acquisition cost.
The market is far more ruthlessly efficient than even Google could ever be.
Could Google have retained all of these developers and then picked the best after the market told them which was the most popular? Certainly. But is that an economically good decision? By outsourcing the process of innovation, Google bears none of the costs and risks of development, benefits and organizational distraction for that process. The costs? A possible premium in asset price and the possibility that the startup will either grow at accelerated enough rate – like a Facebook or Twitter – that the acquisition costs tilt or that a competitor acquires the technology from them.
It remains arguably true that of the things that Google does well, M&A isn’t one of those them. But I don’t think the conventional wisdom that Aardvark is yet another example of this stands up to close scrutiny.