Why Microsoft's Next CEO Will Fail

Steve Ballmer isn't necessarily a bad CEO. After all, Microsoft's on strong financial footing. But Ballmer made enough bad product decisions--Zune, Kin, Vista and perhaps Surface--to suggest that Microsoft employees, swayed by a forced-ranking employee rating system, told him what he wanted to hear, not what he needed to hear. If that culture doesn't change, Ballmer's replacement will fare even worse than he did.

Microsoft, like a lot of companies, adopted an internal employee review process called " forced ranking." The process, a byproduct of General Electric's turnaround, can be appropriate as a short-term measure to address a long pattern of reviews based on entitlement rather than performance.

Over the long term, though, forced rankings tend to pit employees against each other. They focus on assigning and avoiding blame, not finding and correcting problems. In such a system, it can be more lucrative to let a peer make an avoidable mistake than to avoid the mistake altogether - the peer will be fired or graded down, paving the way for a better review and higher raise for you.

In addition, Microsoft has adopted an internal research methodology that gives executives the data they want to see, not the data that reflects the real world. Decisions are often based on false facts.

Steve Ballmer, who graduated at the top of his class at Harvard and had a higher IQ than 98 percent of the folks in Microsoft, failed under these two massive disadvantages. The odds that the next guy will be as smart as Ballmer are low, but the odds that these practices will be retained are high.

Analysis: Missteps Under Ballmer May Have Hit Tipping PointMore: With Ballmer Retiring, Tech World Awaits White Smoke From Redmond

It's even more likely that Microsoft's new CEO will fail. In effect, by not fixing the reasons why Ballmer failed in the first place, Microsoft may have just assured things will get worse instead of better.

Forced Ranking Only Enforces Office Politics, Sets Firms Up to Fail

Forced ranking is a plague on business because it takes the single tool a manager has to motivate employees - the annual review process - and turns it into a problem. Managers rank their employees from top to bottom and must force out the ones at the bottom. It doesn't matter how good you are. What does is how good everyone else is; there are limited slots for rewards every ranking period, and there are also a few slots for employees to effectively get managed out of a job.

It doesn't stop there. Under forced ranking, a truly great manager who's able to choose the best people is at a huge disadvantage to a political manager who instead builds a team that fits his or her bell curve. Granted, this manager still have to get things done, but if you're good with setting expectations - something a politically focused manager is always good at - you can get ranked highly even if you have a poor team, as long as your team doesn't complain and you set your group's goals acceptably low.

The folks who are good at keeping managers happy get the good reviews, while the deadbeats get the bad ones. Meanwhile, one of three things happens to top performers. Some catch on to the forced ranking system and flee to work for your competitor. Others, particularly those who are task-focused and not career-focused, stick around. The politically competent employees stick around, too, and often thrive. Unfortunately, then often kill your firm in the process.

Let's apply this to Microsoft. When you release a disastrous product such as the Zune, Kin, Windows Vista or the Surface RT tablet, you fire some of the people involved and move on to the next disaster. The organization never improves, as a good chunk of it was either intentionally hired to fail or is being encouraged to fail. Potential hit products - such as the Courier Tablet, Chrome Effects or the smartphone that should have been released instead of Zune - are killed off when the risk that the person behind the project would outshine his peers becomes too high.

Analysis: Ballmer's Replacement Must Be a Product Person if Microsoft Wants to ThriveMore: Why Windows 8 Isn't Ballmer's Biggest Regret

Forced rankings also pit departments against each other. Look at Office and Windows. From Windows 95 to Windows XP, Office was the killer app driving Windows adoption. With Windows Vista and Windows 8, though, the app worked better on the prior OS, making it far more likely that Windows would fail but Office would hit its numbers. In the world of forced ranking, a strategy that both moves more of your product and less of a peers' is truly golden.

Bad Data Leads to Bad Decisions, Bad Products

What's funny about this problem being so prevalent in technology companies is that one of the first sayings taught in programming is "garbage in, garbage out." Put bad information into an application and you'll get bad results. So many CEOs fail, though, because their source of information has been intentionally compromised.

This is especially true in blame-based cultures, as there's a tendency to shoot the messenger. Years ago, I created a report highly critical of the vice president of sales. He turned around and gave the report to a competitor, who cancelled a big deal and pointed to my report as the reason. Fortunately, I tracked the leaked report back to the VP - I owned security for the division and am naturally sneaky. The VP was the messenger who got shot, not me, but it was a very close call.

Analysis: Ballmer's Exit Revive Demands for Office on iPad, Android TabletsNews: Microsoft Declares Windows 8.1 Ready for OEMs, But It's Not Finished Yet

Ballmer, as noted above, is incredibly bright and was hand-selected by Bill Gates, who isn't a dim bulb either. Yet Ballmer is known not for his bold moves but his bad ones. (How the heck did Yahoo ever look good?) When you give executives reports showing them that what they want to do is brilliant, you get better promotions and raises, but the company makes more catastrophic mistakes.

I performed the postmortem on John Fellows Akers, the only CEO ever fired from IBM. You know the cause? He was being fed bad information - so much so that he maintained until the very end that IBM was still in great shape. IBM spends more time grooming and selecting CEOs than any other company, but even the best can't succeed if they're separated from reality.

Zune Will Never Kill the iPod, But the Next Microsoft CEO Will Try Anyway

Ballmer couldn't get the support of a blame-oriented culture and couldn't make good product decisions from bad information. However, he was a numbers guy who could make good financial decisions. Throughout his reign, Microsoft turned in strong revenues and profits.

The man or woman who replaces Ballmer will be determined to build the next iPod. He or she will still have the same bad information and blame-based culture but won't have the same focus on financials that Ballmer did. This means the part of Microsoft that's actually working is likely to fail, while we'll get a few more Zunes. As much as I liked the latest version of the Zune, I doubt a higher volume of them will make for greater success.

Therefore, Microsoft's next CEO will be more of a failure than Ballmer - unless Microsoft fixes the cause of Ballmer's failure first. Can Microsoft do that, or is the company doomed to embody Albert Einstein's definition of insanity and keep doing the same thing but expecting a different result?

Rob Enderle is president and principal analyst of the Enderle Group. Previously, he was the Senior Research Fellow for Forrester Research and the Giga Information Group. Prior to that he worked for IBM and held positions in Internal Audit, Competitive Analysis, Marketing, Finance and Security. Currently, Enderle writes on emerging technology, security and Linux for a variety of publications and appears on national news TV shows that include CNBC, FOX, Bloomberg and NPR.

Follow everything from CIO.com on Twitter @CIOonline, Facebook, Google + and LinkedIn.

Read more about ceo in CIO's CEO Drilldown.

This story, "Why Microsoft's Next CEO Will Fail" was originally published by CIO .

Join the discussion
Be the first to comment on this article. Our Commenting Policies