A Vampire Division is defined as a business unit that sucks value from a company but doesn't return a comparable level of benefit. In short, it does more damage than good but often conceals that damage because the true costs are not captured by finance. Vampire divisions tend to crop up in complex companies. They are often very hard to spot and eliminate or correct because they appear to be making a marginal contribution even though they are actually doing a substantial amount of damage.
The recently axed Microsoft Entertainment and Devices Division was a classic example of a Vampire division. Another example would be the part of Apple that managed Apple clones, which Steve Jobs eliminated in the 90s; Open Office/Star Office for Sun; and PC Company initially in IBM. The Apple clones took more revenue and profit than they contributed, Sun's anti-Microsoft efforts became a core reason for Sun's decline, and the IBM PC company drove a shift away from areas that IBM dominated, effectively giving large portions of a new market to companies like Compaq and Dell.
Why Vampire Divisions Suck
None of these efforts were intended to be a Vampire division. Apple's efforts were an attempt to switch to what was, at the time, a more successful Microsoft model. Same with Sun's. Neither realized that this model was fatal to them. While Jobs, on his return, saw this immediately, Sun never did seem to figure it out. IBM saw a trend, but from a corporate perspective, didn't seem to understand it even as it tried to adapt to that trend. IBM got caught in the nuances between terminals, PCs, client/server computing and mainframes. The end result was that it crippled its computers, lost a series of expensive standards wars, and prematurely nearly killed off its very lucrative mainframe businesses. Eventually, it too figured it was better to step away from a market it didn't understand as an institution than it was to pour more money into it.
Vampire divisions are like children who never learn to stop nursing. I think the reason is that, generally, these divisions are just too different from the rest of the business the company is in. The symbiosis, or synergy, that is supposed to occur with a healthy division never takes place. In addition, they seem to attract the type of executive that knows how to game the corporate system, and then retain those same execs because that is the only kind of executive that can survive in them for long. The result, however, is a vampire in a company that can suck the life force out of efforts critical to the company's success and survival.
Staking or Turning Microsoft's Vampire Division
Eliminating a vampire division is very difficult. It looks like it is contributing, it is often very high profile, and the executive running it knows how to play his or her superiors and game the system. Sun never did figure it out, Apple required the firing and replacement of its CEO, and IBM's problems went on for decades and were ignored while it was suffering the most damage.
Microsoft's server and Mac divisions were vampire divisions at one time. Clearly, the server division has transitioned into a powerful asset for Microsoft and the Mac division is more of an arm's-length subsidiary now. Either path can be better than letting the effort help kill the company, as Sun did, or eliminating the effort, like Apple and IBM did.
The issues with the Microsoft Entertainment and Hardware division actually lend themselves more to the Mac division arm's-length model then they do the Server division's integration model. This is because the primary economic costs that are hurting the company come from conflicts much like they did with the Mac division. The Mac division validated a competing platform, the MacOS, so it had to be at arm's length. Otherwise, it would either drain too many resources from the more strategic office efforts or be crippled by them. With the Entertainment and Hardware division, the Xbox effectively and increasingly competes with Windows and has done a substantial amount of damage to related partner relationships. The Zune effectively killed off Microsoft's MP3 player licensing effort, and efforts like the Kin could critically damage Microsoft's partnerships with HTC and other phone makers (even though those relationships largely resided in the same division).
Separating the technology from the hardware execution and putting the hardware execution at arm's length could go a long way toward turning what is now a corporate liability, a Vampire division, into valuable assets for the firm.
Wrapping Up: Garlic Wanted
Steve Ballmer is actually rather good at turning things around once he focuses on them, but entertainment products aren't his strong suit. He has placed David Treadwell, who has had responsibility for Microsoft's competitive cloud efforts, in charge of the core technology. That may mean that Microsoft is taking opportunities and threats like OnLive seriously and realizes that the cloud is likely to be where Google will try to be a first mover. In the end, the lesson to be learned here is that divisions need to be evaluated more broadly than just by their P&L because that doesn't fully represent the cost. There are opportunity and economic costs related to every effort and both should be factored into decisions surrounding an organization's existence and organization. This was an expensive lesson for Microsoft to learn. Not learning it contributed greatly to Sun's loss of independence, and it would be wise to learn this lesson from others -- learning it from experience could put a stake in your own career. This knowledge is your virtual garlic against the creation of a Vampire division of your own. (As a side note, I am clearly reading way too many vampire romance novels at the moment.)