From Ancient Greece to the Metric System; a Warning from History.

by Didier Marlier on Tuesday November 20th, 2012

In Greek mythology, Procrustes was a blacksmith who lived next to a busy pilgrimage route. In his house he had an iron bed, in which he invited every passer-by to spend the night. If they were too short for the bed, he would work on them with his blacksmith’s hammer, stretching them to fit. If the guest was too tall, Procrustes would amputate the excess length. You would probably not find this approach to hospitality on the syllabus of the prestigious Ecole hôtelière de Lausanne. However, what if Procrustes was being closely managed using a simple performance measure? If, as a maker of beds, he has been told that customer satisfaction is measured by how well the customer fits the bed, surely he would be in line for a healthy year-end bonus?

This disconnection between performance measures and results has similarly unfortunate consequences today, notoriously in public sector organizations. The ‘NHS Plan’, the UK Government’s blueprint for the future of the National Health Service, insisted that “by December 2004 all patients will have access to a primary health care professional within 24 hours and a GP (family doctor) within two days.” The intention was to reduce waiting times for patients.

However one Member of Parliament, when trying to book an appointment was told by a receptionist: “In order to meet Government guidelines we have changed our system. You are not able to book more than 24 hours in advance.” It seems that to meet the targets, many doctors were not accepting bookings more than 24 or 48 hours in advance, thus ensuring that no patient had to wait more than the target time between booking an appointment and seeing a doctor. The MP also reported “I have had elderly patients complain that they want to book appointments a week in advance so they can get a lift to the surgery, but instead they are being forced to ring up on the day. Also, because of sheer patient numbers, every appointment is usually taken by 8.30am. Patients are told to ring the next morning. Doctors fear the system is causing urgent cases to lose out to those with routine complaints who might happily wait a few days.”

In many cases, this well-intentioned initiative to improve care made things worse for patients as doctors found creative ways to reach their targets. Predictably there was an incentive attached to the targets: The 24/48 hour target was part of a wider Quality and Outcomes Framework (QOF), which in its first year saw doctors across the country average 90% QOF points, triggering bonuses of £1 billion.

Measures that target the wrong thing are not confined to the public sector. Bob Lutz, in his entertaining book about how he personally saved General Motors, tells a story about paint quality. He observed GM cars had dull, almost grainy, paint. Manufacturing told him “You’re totally wrong. We have the best paint in the entire industry. In JD Power surveys we have the lowest number of paint defects per car of any company, Toyota included”. Lutz makes the sardonic point that a restaurant that advertises the lowest incidence of food poisoning doesn’t necessarily serve the best food; the absence of complaints does not equal excellence. On further digging he found that GM was deliberately reducing the sheen of their paint, so that minor defects were much less visible. So whilst complaints were few, the paint was just ‘OK’, bringing little joy to the customer.

It was not just GM; Ford of Europe was at it too. In the early 80’s they nearly killed the Escort Convertible. Whilst it was achieving the highest overall satisfaction level of any Ford of Europe product, it had a higher than average number of complaints for water leakage and wind noise (not entirely surprising for a car with a canvas roof). The logic went; if we kill the Convertible, those complaints will drop and the average quality score for Escorts will rise to meet our target. So, a target could be met even though no single Escort had actually improved, whilst customer choice would be reduced and many satisfied customers would need to look elsewhere next time. Genius!

Obviously single metrics can cause problems; hence the need for multiple objectives. In GM the VLEs (Vehicle Line Executives) were very powerful people, usually engineers, who were responsible for individual vehicle programmes through from gestation to sketch to design to production. Many of their objectives resulted from benchmarking Toyota; cost, investment, quality, warranty cost, assembly hours per vehicle, percentage of parts reused from previous model and, especially important, time to launch. Interestingly, ‘being appealing to customers’ was not on the list.

Bob Lutz tells a story of a meeting with a VLE who had bought his scorecard to a project review. “He made sure I understood that he had met or beaten every target; it was solid green with no yellow or red. He deserved, in his opinion, a big juicy chunk of that special VLE compensation fund. “How’s it selling?” I asked. “Well, really not that well; the press on it was lousy and the public unenthused. But, I can’t be held accountable for that. I was handed my numerical goals, which were signed off by everybody, and if I make them all, that’s success”. Lutz described this as the tyranny of process over results, and committed thereafter to hold VLEs accountable for the success of the vehicle.

It is not just customers that are at risk from process dysfunction; suppliers and partners can suffer too. There are purchasing departments that are measured, and rewarded, on the reduction they can negotiate on an initial offer from a supplier. Again, the intention is fine[1]; reduce costs and engage people in enhancing the bottom line. The reality is rather different. When the supplier knows what is happening, they can adjust the initial offer upwards in the expectation of having to take a cut. It is possible that the end result is actually a higher price, whilst the purchasing manager gets a bonus for a solid percentage reduction. Not only is value destroyed but trust and the relationship can be tarnished by the game playing.

This blog is not meant to be a polemic against measurement; metrics are a necessary and constructive tool if used carefully. The graveyard is littered with companies that would do anything for the customer, but had no internal discipline and made a loss on every widget they produced. Many of them were taken by surprise when they went bust because they were not measuring cash flow. However the overuse of performance measures seems to be an unfortunate characteristic of large organisations; small fast moving ones just can’t find the time to build complex systems.

Performance measures are necessary, but they must be the right ones. There are some key questions to ask when setting objectives or devising metrics:

  • How does this objective relate to the creation of value for the customer?
  • What is really important to the customer? How exactly do they measure value?[2]
  • If I measure something, what am I going to do with the result? What decisions will I make when I have the data? If the answer is ‘don’t know’ or ‘nothing’, then ask yourself if you really need to divert energy or resources to measurement. Just because you can measure it; it doesn’t mean you should.
  • If I look at all the goals and targets I have; can I immediately tell what is most important and what can be relegated to a secondary role? To borrow a powerful military concept; have I got a schwerpunkt[3]; a point of main effort?
  • If I want to influence my people’s behaviour, is there an alternative to the obvious measure and reward system?[4]

[1] Although we have heard about one executive who said, “We introduced targets for purchasing to make the buggers work for a living”.

[2] There is a great story, possibly apocryphal, of a major mobile phone company who asked their customers what they wanted in their next phone. The answer was “a better quality camera”. The engineers set to work integrating the best Carl Zeiss lens they could find and launched it with great pride and fanfare. Customers were not enthusiastic. To them ‘better quality’ meant 8 megapixel rather than the current 4. Most had no clue what or who Carl Zeiss was and they certainly didn’t want to pay the premium that had to be charged to cover the extra technology.

[3] The Germans, and now NATO, refer to a Schwerpunkt (focal point and also known as Schwerpunktprinzip or concentration principle) in the planning of operations; it is a center of gravity or point of maximum effort, where a decisive action could be achieved. It is summarized by Guderian as “Klotzen, nicht kleckern!” (literally “boulders, not blots” and means “act powerfully, not superficially”).

[4] There will be a follow-up to this blog putting the spotlight on rewards and motivation.

Share and Enjoy:
  • Print
  • Twitter
  • Facebook
  • StumbleUpon
  • Digg
  • Google Bookmarks
  • LinkedIn

Leave a Reply