THE difference between companies is people. With capital and technology in plentiful supply, the critical resource for companies in the knowledge era will be human talent. Companies full of achievers will, by definition, outperform organisations of plodders. Ergo, compete ferociously for the best people. Poach and pamper stars ruthlessly weed out second-raters.

This in essence has been the recruitment strategy of the ambitious company of the past decade. The ‘talent mindset’ was given definitive form in two reports by the consultancy McKinsey famously entitled The War for Talent. Although the intensity of the warfare subsequently subsided along with the air in the internet bubble, it has been warming up again as the economy tightens: labour shortages, for example, are the reason the government has laid out the welcome mat for immigrants from the new Europe.

Yet while the diagnosis – people are important – is evident to the point of platitude, the apparently logical prescription – hire the best – like so much in management is not only not obvious: it is in fact profoundly wrong.

The first suspicions dawned with the crash to earth of the dotcom meteors, which showed that dumb is dumb whatever the IQ of those who perpetrate it.

The point was illuminated in brilliant relief by Enron, whose leaders, as a New Yorker article called ‘The Talent Myth’ entertainingly related, were so convinced of their own cleverness that they never twigged that collective intelligence is not the sum of a lot of individual intelligences. In fact in a profound sense the two are opposites.

Enron believed in stars, noted author Malcolm Gladwell, because they didn’t believe in systems. But companies don’t just create: ‘they execute and compete and co-ordinate the efforts of many people, and the organisations that are most successful at that task are the ones where the system is the star’.

The truth is that you can’t win the talent wars by hiring stars – only lose it. New light on why this should be so is thrown by an analysis of star behaviour in this month’s Harvard Business Review. In a study of the careers of 1,000 star-stock analysts in the 1990s, the researchers found that when a company recruited a star performer, three things happened.

First, stardom doesn’t easily transfer from one organisation to another. In many cases, performance dropped sharply when high performers switched employers and in some instances never recovered. More of success than commonly supposed is due to the working environment – systems, processes, leadership, accumulated embedded learning that are absent in and can’t be transported to the new firm.

Moreover, precisely because of their past stellar performance, stars were unwilling to learn new tricks and antagonised those (on whom they now unwittingly depended) who could teach them. So they moved, upping their salary as they did – 36 per cent moved on within three years, fast even for Wall Street.

Second, group performance suffered as a result of tensions and resentment within the team. One respondent likened hiring a star to an organ transplant. The new organ can damage others by hogging the blood supply, other organs can start aching or threaten to stop working or the body can reject the transplant altogether, he said. ‘You should think about it very carefully before you do [a transplant] to a healthy body.’

Third, investors punished the offender by selling its stock. This is ironic, since the motive for importing stars was often a suffering share price in the first place. Shareholders evidently believe that the company is overpaying, the hiree is cashing in on a glorious past rather than preparing for a glowing present, and a spending spree is in the offing.

The result of mass star hirings as well as individual ones seem to confirm such doubts. Look at County NatWest and Barclays de Zoete Wedd, both of which hired teams of stars with loud fanfare to do great things in investment banking in the 1990s. Both failed dismally.

Everyone accepts the cliche that peo ple make the organisation – but much more does the organisation make the people. When researchers studied the performance of fund managers in the 1990s, they discovered that just 30 per cent of variation in fund performance was due to the individual, compared to 70 per cent to the company-specific setting.

That will be no surprise to those familiar with systems thinking. W Edwards Deming used to say that there was no point in beating up on people when 90 per cent of performance variation was down to the system within which they worked. Consistent improvement, he said, is a matter not of raising the level of individual intelligence, but of the learning of the organisation as a whole.

So if not by hiring stars, how do you compete in the war for talent? You grow your own. This worked for investment analysts, where some companies were not only better at creating stars but also at retaining them. Because they had a much more sophisticated view of the interdependent relationship between star and system, they kept them longer without resorting to the exorbitant salaries that were so destructive to rivals.

The star system is glamorous – for the few. But it rarely benefits the company that thinks it is working it. And the knock-on consequences indirectly affect everyone else too. As one internet response to Gladwell’s New Yorker article put it: after Enron, ‘the rest of corporate America is stuck with overpaid, arrogant, underachieving, and relatively useless talent.’

The Observer, 9 May 2004