By Jeffrey Tarter

Five years ago, Microsoft's support organization was under-funded, under-managed, and notoriously unhelpful. But then chairman Bill Gates opened up the Microsoft checkbook, determined to build a world-class service operation. Gates put some of the company's best managers in charge of engineering a turnaround; they went to work building state-of-the-art systems and training literally thousands of new employees. A half-billion dollars later, Gates has largely succeeded: By almost any measure, Microsoft now ranks among the most responsive and best-managed support providers in the industry.

There's only one glitch: So far, Microsoft's reputation for service still hasn't budged much--if at all--from the bad old days. In terms of improved customer perceptions about Microsoft, the money spent on service enhancements probably ranks as one of the worst investments Gates has ever made. (To be sure, there are other intangible paybacks to service excellence which Gates understands perfectly well. But Microsoft likes to get credit for its good deeds, and that clearly isn't happening.)

For the software industry at large, there's a discouraging message here about the value of good customer service. According to research firms that survey end users, service performance (which, in the shrink-wrapped software world, usually means tech support) now ranks as one of the top two or three checklist items for comparing rival vendors. Classical business school theory points to essentially the same trend: When everyone ends up selling more or less the same features at roughly the same (low) price, customers usually end up treating service as a key competitive distinction.

In fact, PC hardware buyers already make a big deal out of support performance: Experienced customers know who does a good job, and they avoid vendors with screw-the-customer reputations. But shrink-wrapped software support reputations are much fuzzier. Even though buyers claim that service is a hot button, in reality they seem to be clueless about actual support performance. End-user surveys consistently report that most publishers are believed to deliver "average," commodity-like support. The best performers pull ahead of the pack by perhaps 10%; the worst lag behind by about the same 10%--a difference that's usually too small to have any impact on buying decisions. So, despite lots of wishful thinking to the contrary, spending money to upgrade a company's service reputation remains a lousy investment.

In some ways, it's easy to see why customers don't always perceive differences in support performance. Unlike high-end software developers (whose employees show up in person for installation and provide an ongoing trouble-shooting and maintenance presence), desktop software companies deliver support almost entirely by telephone. And, on average, customers typically place only one or two support calls during their entire experience with the product. The trade and business press provide relatively sketchy coverage of support topics; product reviewers have never managed to create meaningful benchmarks for support performance. Like obscure political candidates, support reputations simply don't inspire more than a "no opinion" vote from most users.

And if there's so little payback from support investments for individual software companies, there's probably even less hope that the whole software industry's lackluster service reputation will ever improve. Correctly or not, potential buyers tend to believe that publishers ship buggy products, then refuse to provide more than grudging assistance over perpetually-busy phone lines. That's not a reputation that helps build confidence among the new customers the industry needs to keep growing the market.

Ultimately, software companies are businesses that have to survive, turn profits, and make rational calculations about where to invest their resources. Most of the time, the investments that produce better products and customer service also happen to yield the highest payoffs. But tech support may be the exception: Until customers begin to notice who's doing a superior job, it looks like most developers will spend just enough to sustain "average" support quality--and nothing more.

Jeffrey Tarter is editor & publisher of Softletter and executive director of the Association of Support Professionals, jtarter@softletter.com.