I was reading this depressing analysis of the finances of Research In Motion, makers of the Blackberry phone, and reading some commentary that the apparent demise of RIM could have been avoided if the company had somehow been different.

It got me thinking about corporate DNA.

Pick any company, and there's a good chance you can distil its raison d'être down to a pithy, not entirely inaccurate sentence. Microsoft makes ubiquitous software intended to be installed by default on 90% of the world's personal computers. Walmart builds cavernous department stores that sell as many things as possible as cheaply as they can. Armani leverages the cachet of its high-end fashion business to mass-market clothing under ‘lesser’ labels. Electronic Arts throws armies of developers at big brand-name game titles. Facebook develops more reasons for people to visit Facebook.

For the years it was successful, RIM made utilitarian business phones that were really good at email.

Corporate DNA isn't determined by a mission statement or dreamt up on an executive retreat. It comes from what the company actually does.

DNA is a flawed metaphor, but it at least captures the core, self-replicating concept. If a company is built around doing X, it is going to attract the kind of people who want to work for a company that does X. It is going to select staff who are good at executing X. It is going to structure its divisions and management around the parts necessary to do X. It is going to build processes, conventions and traditions all around how best to do X.

Everything about a company's strategy—what new ventures are given attention, what old ones are scrapped, which of a range of possible solutions to a problem it might pick, the degree to which it tolerates risk—all get entwined in the question “How does this relate to executing X?”

This idea of Corporate DNA is writ large in Mat Honan's stunning evisceration on Gizmodo, How Yahoo! Killed Flickr and Lost the Internet.

But Yahoo's social success in those years was almost accidental. It wasn't (and isn't) a company with vision. Its founders Jerry Yang and David Filo's great contribution to the Internet? They built a directory of links and then sold ads on those pages.

It was a gateway, nothing more. This was hardly an innovative idea, or technically complicated to pull off. You don't have to write algorithms to build a portal. Yahoo was little more than an electronic edition of Yellow Pages.

The founders' influence on a company's culture is enormous, and Yang and Filo cared about business, not products or innovation. They didn't foster a culture of computer scientists, like Google's founders did, or cultivate hackers like Facebook. They grew a business culture. For many years that worked quite well—until Google came along.

So when overnight, and entirely to my surprise the smart phone market stopped being about making utilitarian business phones that were really good at email, RIM was in the worst position to deal with it because they didn't just have to change their strategic direction, they had to change their entire corporate makeup.

The phone companies that survived the iPhone shake-up most ably were the ones whose DNA most closely resembled “We flood the market with phones built from commodity parts.” Because they were already poised to beg, borrow, steal and copy the next generation of products.

Companies can successfully mutate their DNA, but it’s inevitably traumatic. Microsoft built a successful server business by throwing money at it until it damn well worked, and had an even more unexpected success with the X-Box by pursuing the same approach in gaming. But beside those successes, there has been a lot of red ink splashed on equally ambitious but far less successful reinventions.

IBM, whose DNA going into the 1980s was “We sell fuck-off big hardware to people who can afford it” spawned the PC industry almost by accident, immediately lost control of it to Microsoft and Intel, almost collapsed in on itself, then reinvented itself as a hybrid research/server/services company.

Microsoft mutates because it can afford to fail repeatedly, IBM mutated because it was, as Cringely said in Accidental Empires, the size of a small country.

The two tech companies with the most interesting DNA are the ones I rather obviously haven't mentioned yet. They're interesting and successful because their “X” isn't about building a particular thing, it’s about a particular approach to building things.

One is “We use exceptional engineering to execute our ruthless product vision, trusting that our aesthetics are shared by the market.” The other is “We stockpile incredibly smart people and throw them (or let them throw themselves) at difficult problems, trusting that enough of those solutions will be profitable.”

And both companies have undergone recent and major structural changes that challenge that core direction.