July 27, 2009

Buying not making innovation

Larry Magid has an apt summary of Google’s innovation strategy in his San Jose Mercury column this morning:
Google, after all, has done an amazing job with its search engine and, thanks to the profits from all the ads it sells, has an enormous war chest to invest in research and development. The company is so keen on innovation that it allows its engineers to spend 20 percent of their working time on projects that aren't necessarily part of their job description. It's that "20 percent time" that helped spawn such projects as Google Suggest, AdSense for Content and Orkut.

And what Google can't invent, it can buy. Its Google Voice application, which it acquired when it bought GrandCentral Communications in 2007, is a stellar product, as is YouTube, which Google acquired in 2006.
I realize this is just one commentator’s interpretation of a $20 billion/year company. Still, I find Magid’s final point particularly interesting. Most of the internal innovations are related to search (Orkut is a me-too social networking site that has pockets of success), while the new areas have come from acquisitions.

In other words, to successfully diversify outside its area of expertise, Google has to buy not make those businesses. Google is the most successful high-margin, high R&D, high-growth tech company of our era, just as Microsoft was in the 1990s, Apple and DEC in the 1980s, and IBM in the late 1960s. One way to look at this is if Google doesn’t have the resources to pull off internal diversification, who does?

Another way to look at it is that Google is copying Cisco — diversification through acquisition — because it’s painfully aware of its predecessors’ failures. Yes, IBM created some great businesses through internal R&D, as have Apple and Microsoft (in many cases by hiring key talent from outside). However, the “not invented here” model of internal innovation also brought us such notable flops as the DECmate, DEC Rainbow, IBM PCjr and Apple Newton.

So perhaps we should give Google credit for taking its cash and savvy for buying the best innovations that are available — assuming it spends its money more prudently than the drunken sailors in Washington.

This does come back to a minor academic controversy: is it “open innovation” to buy up innovative companies? It’s open innovation to buy products from such companies, and closed innovation to develop things inhouse. Although others might disagree, I think the integration (or diversification) by acquisition is in the end a form of closed innovation, because it reflects an ongoing desire to control key technologies through administrative hierarchies rather than source them using markets.


Unknown said...

Whether or not acquisition of potentially popular technologies is an example of open or closed innovation will migrate towards a semantic argument. I agree with you however, the next valid thrust would be that it is the synergies available to Google by having an umbrella control that will provide even more innovation in the future. I think its a weak argument nonetheless. This reminds me of the conglomerate form of enterprise of the not do distant past.
Thanks for the read.

Munish Gupta said...

As an enterprise, I should not be worried, whether the innovation is inhouse or is coming from outside( via acquisition). I believe, the important part, how is innovation fitting into the overall enterprise vision. If google's vision is "organize the world's information and make it universally accessible and useful", then the innovation should tie up with the vision.

Jayesh Badani said...

Good post Joel,

It made me think for a while on your question if buying a innovative companies can be considered as open innovation. I guess yes, but there is one fundamental difference, when you buy a company, you have not taken a risk on you. You see the success of that company and now you want to integrate that with yours. While in open innovation, you are asking for a innovation on a particular subject, you will decide if you like the idea and eventually you will take a risk to implement it.


Joel West said...


I see your points. Certainly once you have done the acquisition, you are now vertically integrated and you run all the risks of any vertically integrated company, including that the technology (or factory or people or …) that you have invested in turns out to be a dud.

But in OI you don't necessarily run that risk. Yes you could precontract for the technology, but you can also buy it (non-exclusively) on the open market after it's done and proven. See

Chapter 6 of the 2006 Open Innovation book talks about buying Intel chips on the open market rather than making your own CPU is open innovation


For a company like IBM, this would manifest a major shift away from vertical integration, even if for Dell it would be the normal way of doing things.