There was a time in the early ’90s when “expert systems” were heralded as a breakthrough in knowledge management. The goal was to capture and forever reap the rewards of a company’s best and brightest. The systems took various forms, but all in essence sought to extract, through interviews and other means, a talented employee’s genius for, say, picking stocks (or, at Campbell’s, making a perfect batch of soup), and embed it in software. Software, after all, doesn’t call in sick, or sign on with a competitor.
It was an appealing vision, but it remains mostly that. While the larger category of artificial intelligence still garners interest at universities and research firms, it finished dead last in a recent poll of technologies of interest to corporate executives. But there is still plenty of expertise walking the halls at most companies, and few efforts to make the most of it. McKinsey & Co. consultant Michael Idinopulos and principal Lee Kempler offer a new approach: don’t try to bottle the expertise, just be content to find it when you need it.
Many companies use databases or document repositories of some sort to track employees’ expertise, but that information is often captured in a few high-level bullet points that miss key data. Rolling out a new salty snack in Ohio? You’ll find several people who cite “product launches” as competencies, but you probably won’t know that one of them went to school in Columbus and worked on a nearly identical launch early in her career; thus, the ideal candidate risks going untapped. Experience is not the same as expertise, of course, but Idinopulos and Kempler say it is a “frequent companion” and should be tracked more closely.
Now, using a combination of search-engine and database-integration techniques, it can be. Companies can buy or build new, more-sophisticated systems that truly capture an employee’s relevant experiences, interests, and other data. Software can scan documents, forms, and even E-mail messages to create a rich profile of every worker. Interfaces can be customized so that seekers of expertise can slice and dice competing experts’ backgrounds, or select by geography, time zone, or other criteria. Such an effort could take several months and cost as much as $500,000, although the cost could be less if a company already has a data warehouse that can act as a repository for this deeper view of employee experience. Compared with what companies have spent on other forms of expert systems, that’s a small price to pay for a legitimate brain gain.
It’s been several years since you could find the words Internet, initial public offering, and billions of dollars in the same sentence (barring those written in the past tense, of course). But as Nasdaq has climbed, so too has activity in the high-tech IPO space. For example, RedEnvelope.com raised about $30 million in late September, and days earlier the parent company of proFlowers.com filed for an IPO, seeking funds for new online gift ventures. True, RedEnvelope still operates in the red, but many analysts took the IPO as a sign of life not only for E-tailers but for the dot-com world in general.