Open is better than closed. We all know that, right? If you are trying to get into a house, when the door is closed you walk right into it. That sounds bad.

But if you are in the house, things are less clear. If the door is open, wind and rain and who knows what may come in. That sounds bad.

This is the fundamental problem with the ‘open versus closed’ debate when it comes to strategy. It eschews the purpose of the door which is to allow it to be open and closed at the choice of someone. The real issue is who and when that someone is.

These thoughts came to me as I read Jonathan Rosenberg’s latest article at Think Quarterly, Google’s online magazine. Rosenberg is the former VP of product development at Google and now an advisor to them. I have always found his thoughts to be provocative and enlightening. See here is his attempt to define openness, here are his thoughts about the future and also here a lecture on Google’s product philosophy. His posts were Obama like in their deference to alternative possibilities while, in the end, building up to support a position. That made them the sort of readings I could give to my students.

But those had been done three years ago and we haven’t heard much from Rosenberg in that time, until today. In ‘The Future is Open’ Rosenberg reflects on his earlier writings that suggested openness was important.

In the weeks that followed, I received thoughtful emails from a remarkably broad audience – professors and writers appreciative of the look inside Google, business leaders telling me how open affects their business, grad school students surprised that this was the very opposite of the lock-in strategy they were being taught. Cut to three years later. What leaps out at me from that manifesto now is something entirely different: How wrong I was.

Wow, I thought. This is going to be interesting. But it turned out to be a feint.

It’s not that open doesn’t improve Google and the world. It’s that this has happened far faster than I’d ever imagined. This realization came to me recently in the middle of the most mundane of twenty-first-century routines: I was checking my phone, a Droid Razr Maxx. Staring at the thing, I saw it for its sheer diversity: Two dozen apps – from the New York Times to Flipboard,Dialer One to OpenTableRunKeeper to SlingPlayer – created by a slew of different developers, on a phone built by Motorola. It occurred to me I wasn’t looking merely at a mobile device, but the physical embodiment of how an open ecosystem can ripple its way through the world nearly overnight.

Oh, he was wrong because he just wasn’t right enough. And so what emerges now is not balanced reason but more on the lines of evangelism.

Chrome and Android, which have taken off since ‘The Meaning of Open’ first appeared, exemplify this principle. With both, we’ve maintained one simple goal from the beginning: Make the product as strong as it can be. As we learned time and again, no route would get us there faster or more reliably than open – more hands working on a product will only improve it. Open allows for preto-typing a concept, or testing it in the earliest stages. What’s more, open systems tolerate failure better – and attract a more devoted user base. They know the primary motivation of an open system is product excellence; if the company tried to impose some other agenda on it, the developer audience would detect it immediately and revolt. In committing a product to openness, the company surrenders the ability to do anything but make it better for the user.

Now that’s fine. It is a line I would expect Google to take. But is it really warranted? It is claimed that openness has generated a diversity of experiments and applications for Android and along with that a high market share. And for Android that is true. But it is really hard to claim that openness is The Way rather than A Way for these things. Last time I looked Apple had more apps and the growth rate for apps on Windows Mobile was faster than both. This suggests there is more to the story.

Now Rosenberg, in one moment, acknowledges this a little:

The alternative tactic – most notably employed by Apple and our own search teams – is to keep systems more closed, and to exercise complete control. This approach requires its own set of unique organizational skills, beyond just moving fast, since product excellence and innovation must be sourced entirely from within. Both approaches can obviously be successful, but in our experience, when it comes to building global platforms, going open is a more sure-fire path to success.

Hang on a second. Look over the last decade. The two most successful platforms are Apple’s iOS and Google’s search. And both apparently are not open in the sense Rosenberg now sees as the “more sure-fire path to success.” Indeed, a proper reading is that both of these platforms have been open in that they have invested heavily in making participation easier and freer. Apple have lowered the costs of development and distribution for many app makers. Google has lower the transaction costs for many advertisers. By reducing costs, they have opened up the platforms. To be sure, both Apple and Google still hold the door handle but the point is that they are standing up and holding it open. For Android, the door opening is left for others. Its a trade-off and that means that sometimes both approaches may be warranted.

Saying to businesses that considering openness as a profitable strategy is one thing. Virtually requiring it is another. To take one of his examples:

Fortunately, a growing number of organizations have seen the writing on the wall. In Wikinomics, authors Don Tapscott and Anthony D. Williams recount the tale of Goldcorp, a Toronto gold-mining firm that, in the late ‘90s, appeared to be on the ropes. Facing a contracting market, a host of internal troubles, and what appeared to be a picked-over mine, CEO Rob McEwen did precisely what any business textbook would say not to: He started giving away what little the company had left.

Specifically, he dumped 400 megabytes of information about Goldcorp’s 55,000-acre property on the company website. Rather than jealously guard its last shreds of proprietary information, he offered $575,000 in prize money to anyone who could use their data to, in essence, find their gold. It was a tremendous success. More than 80 percent of the targets identified by the public yielded significant quantities of gold. From that small initial investment, the company has pulled over $3 billion worth of gold from the ground.

To say this is openness of the type Android has is very strange. Goldcorp opened up its exploration operations. But it did not open up its mineral rights. They never ever forced themselves into a position where they had to compete to be the best mining operation to match that exploration. The the company never surrendered “the ability to do anything but make it better for the user.” Of course, that is not to say that the moves it made were not sensible. I am just saying that they would have been unlikely to have followed a ‘fully open’ path. This was good for their profits even if it would have been better for the Canadian economy to have had competition brought in further down the value chain in mining.

Sadly, this time around Rosenberg hasn’t helped the communication and understanding of the types of strategies businesses should employ. But that he posted his thoughts at least might encourage discussions that do.

2 Responses to Being careful about openness as a strategy

  1. tomslee says:

    I’m glad you are picking up Mr. Rosenberg on his rhetoric. As in his previous article, he seems to have little insight into Google’s actual business model and how its selective openness serves it. The idea that selectively open==open is quite strange.

    Also, I do love the suggestion that the Canadian government is now a paragon of openness.

  2. […] Rosenberg. These two were for a decade the ‘adult supervisors’ at Google. I have written about Rosenberg before  The book is a set of anecdotes peppered with good advice. The best advice is to ‘default to […]

Leave a Reply

%d bloggers like this: