I have a soft spot in my heart for Yahoo! After all, it grew out of the Stanford School of Engineering (where I work) and founders Jerry Yang and David Filo have been quite generous to the school. I have also always been treated nicely by people in the company, especially those in HR and Strategic Data Services. But as I got to know the company, I kept seeing deeper and deeper evidence of two flaws that, I believe, help explain how they got into their current mess:
1. A lack of a clear strategy. They would seem to say "yes" to any idea that seemed good, throw it into the mix on their increasingly confusing landing page, and often make remarkably little effort to fit the different "properties" together. Indeed, I blogged a few months back about Steve Jobs' comment that great companies don't just kill bad ideas, they have to kill a lot of good ideas to because it is only possible to do a small number of good things well and otherwise you end-up with a kind of feature creep (this isn't just in products) where there is so much stuff that –even if it is all good — the human experience of dealing with it becomes confusing and unpleasant. The one thing I didn't say in my post was that Jobs gave that talk to Yahoo! senior executives. They have made some progress if you look at the their landing page, but I still see lots of evidence of a lack of strategic focus and inability to stop doing good things that don't fit with their strategy — perhaps because they have had a weak strategy.
2. Dysfunctional internal competition. For years before Yahoo! went into its current steep decline, I heard senior Yahoo ! executives complain about how difficult it was to create cooperation across different Yahoo! "properties" such as mail, search, jobs, auto, and so on. As a result, there was often poor integration between the properties and lack of information that could have helped everyone. There were two reasons for this, from what I can tell. First, the properties had lots of power, and central management did not. Second, the reward system and culture pitted properties against each other, so there was actually an incentive for property managers to treat each other as competitors. I had a long cross country flight where I sat next to one property manager who spent much of the time complaining about how he was getting no help at all from several other property managers, even though working together would help them in the long run and the company. He explained that it had to do with how rewards for were handed it and with the fact that Yahoo! — although it was getting better — had few strong "cross-property" groups for integrating the technologies and actions of people. He also admitted that, if he just looked at his incentives (except for his stock options, which were underwater anyway), it wasn't really rational for him to help people in other properties. This was exacerbated in some cases by acquisitions that Yahoo! made that were turned into properties and never quite integrated into the whole.
Although decentralization is sometimes treated by Americans as like motherhood and apple pie, and it certainly is effective for spurring innovation and allowing effective responses to the quirks of local markets, any good organizational theorist will tell you that centralization is one of the known cures to the above two problems. This is especially true when there is time pressure, so there is no time to do a massive cultural shift so that peer pressure and a common world view can replace authority to cure the above problems.
This all brings us to the press reports today that Yahoo!'s new CEO Carol Bartz — who is known for leading with a strong hand — is revamping the structure to make it far more centralized. Given Yahoo!'s historic problems, the lack of strategic focus, the lack of a strong internal core to weave together ideas and software from different properties (although they have made progress on the software), and especially, the dysfunctional internal competition, strong centralized management is probably the best answer right now. On the last point, as my mentor Bob Kahn taught me, intervention by a greater power is one of the most effective ways to get people, groups, or businesses to stop fighting and start cooperating.
Let's see what happens, as Bartz has many challenges facing her, and even if this is exactly the right move, it may not be enough — but I hope so. I would also appreciate other perspectives on Yahoo! and Bartz, as I realize mine is limited to just a few of their problems.
P.S. The weird thing about centralization is that cures some problems, but creates others. It makes it harder for people to innovate and makes them less responsive to "local" quirks of markets and specialized customers. Often the only cure for the problems caused by centralization is to decentralize for awhile. That is one reason why, if you look at the history of companies like HP, they swing back and forth between periods of centralization and decentralization — which looks like wishy-washy confusion to insiders and outsiders (and may be at times). But it is also pretty rational solution over the long haul, as the best way to cure the problems created by decentralized is to centralize, and vice-a-versa.
If Bartz is successful, one prediction is that four or five years down the road, it will be time for decentralization at Yahoo! I hope they last that long!
