“Why is it that people start to behave strangely when all of a sudden performance scorecards are introduced into their organisation? They say they want them but the moment they have them they try to wriggle around the data that comes out of them?”

That is one of the critical questions that Frank Buytendijk has been struggling with for years. As the former head of the performance management practice at IT adviser, Gartner, and now as VP of corporate strategy at BI company Hyperion Solutions, Buytendijk draws on scores of examples to explain the subtleties behind maximising business performance – possibly solving his conundrum.

One thing is clear from the start, he says: “Measurement works – in our daily lives as much as in business.”

When trying to lose weight, for example, individuals begin a diet by figuring out how many kilos they need to shed. And then they need constant feedback on how well they are doing. “Unless they get that feedback, the whole experiment of trying to lose weight will end in a very negative way very soon,” he says.

That relates to sociological phenomena known as the Hawthorne Effect: the fact that when some activity is measured in a social setting, the very act of measuring it alters the participants’ behaviour and therefore the independence of the result.

Managers trying to create better performance by measuring employee behaviour often don’t understand very much about the effects of their measurement, he suggests.

An insurance company Buytendijk recently worked with, for example, had defined its most important metric as ‘the average time taken to process a claim’.

“They realised this metric needed some improvement so they came up with an extremely simple solution,” he recalls. They put one graph on the departmental message board that showed production for the week, broken down by the four internal claims processing groups.

The management thought that this graph would make the people more competitive, but it had a completely different effect. After two weeks, one group observed that a ‘rival’ had slipped behind because of staff shortages. The first group offered to take over some of the excess workload, aware that some time in the future they might need similar help.

“The desired effect was achieved: the average overall processing time, the key performance indicator, improved,” says Buytendijk – even though the management view of how it would be achieved was incorrect.

Buytendijk believes there are two ways of better understanding human behaviour in relationship to performance metrics: understand exactly what metrics are needed to drive the right behaviour, and understand the cultural context of the metrics set in place for achieving the other objectives.

Moving targets

But the focus on individual or group targets – where everybody is responsible for a metric – might not provide the optimal performance result.

What organisations need are metrics that have multiple owners. There are interfaces between the domains that different people are responsible for and it is at these business interfaces where most time is lost or gained, quality problems occur or are solved, or costs are being incurred or reduced. “Put metrics in place on the business interfaces,” he says.

IT is a good example. Most organisations have operations and development sides: one measured on running IT, one focused on the quality of software. But, in Buytendijk’s model, there should be metrics that measure where they naturally overlap. “The speed, quality and cost of new developments as they are taken into production – that should be the responsibilities of the managers of both sides of IT,” he says.

But there is a second issue to consider. Organisations need to better understand why people behave as they do when they are being measured. And that requires managers to make a realistic assessment of both the organisation’s positive and negative characteristics. When individuals and groups are measured, they will not take kindly to metrics that are based on some interpretation of the corporate mission statement that is more wishful thinking than reality.

As he concludes: “Performance is not in the numbers, it is in the people.”

Further reading in Information Age

Business Intelligence briefing room