10 things to avoid

Selecting things to measure in business is easy. Selecting the right things to measure is very difficult. We frequently make errors when selecting metrics because we are busy and believe that it is better to measure something than to not measure at all. This of course is our first and biggest mistake. Measuring the wrong things ultimately results in poor decision making and a lack of direction.

Here are 10 things to avoid when developing KPIs:

1. Using measures that do not relate to a company vision or at least to company objectives. You can be forgiven for including metrics that may not relate directly to a company vision, but not relating them to company objectives is pure folly. Every metric must be associated to a company objective.

2. Relying on ‘lag’ measures. Lag measures look back at what has happened in the past. They can tell us if we have achieved our targets, they do not however have any influence on the targets. The classic example: getting on the scales tells us how much we weigh, but does not influence our weight. Deciding on what we should eat and how much exercise we take does. These are ‘lead’ measures, that is, a measure that can be influenced to cause a change.

3. Measuring it because it is there. We like to let management know how much work we are doing and so count everything that can be counted and present the numbers. Often it makes very little difference to the business or to the way individuals perform. Sometimes it has a detrimental effect as we artificially try and force (useless) numbers up. Just because it can be counted does not mean it should be.

4. Creating measures that do not have targets. If a measure does not have a target then it is not a valid measure. Furthermore, a measure should have defined thresholds i.e. what is good, what is bad and what is acceptable during the time it is being measured. This produces a classic red/amber/green status and can provide a strong visual representation as to the current status of the measure.

5. Believing customer satisfaction surveys. Not only are the measures associated with customer satisfaction surveys always ‘lag’ measures they rarely reflect what is actually happening. Let’s face it, they are easy to cheat on, they get sent to our favourite customers and often the questions are written in a way to elicit a positive response. There are better things to measure including numbers of complaints, numbers of unsolicited letters of praise and numbers of customers that turn up at events.

6. Annual measures. The ultimate in ‘lag’ measures and often a snap-shot in time. Clearly there are end-of-year measures that are important e.g. annual profit, but generally these must be measured throughout the year as well.

7. Feel-good measures. Usually put in place to demonstrate that a problem area is getting better without paying attention to the underlying issue. An example is to measure the number of customers rather than measuring the customers worth, i.e. lots of customer must mean that business is good. Lots of customers that don’t pay their bills will not help the business.

8. Measures that are not clearly defined. Some measures can be both good and bad for a business. For example staff-turnover. Staff leaving the company because they are performing badly is generally a good thing. Key staff, without whom the company cannot function, is a bad thing. It is important to have clarity in metric definition.

9. Using measures that are out of your control. If as a result of a measurement you cannot implement a plan of action to make a change to improve that measurement, then it is not a good measure. It may be a very noble thing to measure the CO2 output of your 3rd party data centre provider, but you cannot be held accountable for the failure to meet their targets and nor can you influence their activities to make improvements.

10. Ignoring the human behaviour a measure will cause. All measurement will result in humans acting in a way to best meet the targets. This can often provide surprising results. A concrete fence post company put in place measure and target to reduce ‘concrete wastage’. The result was the production line workers ordered less concrete towards the end of the day and therefore produced less concrete fence posts. Another more well know example is to focus on Sales Revenue to grow a business without putting due emphasis on profitability. Cost soar to successfully generate revenue while profits tumble.

Most of the above is common sense. There is a balance required in any measurement system between the need to measure for operational efficiency and strategic desire, to look carefully as those measures that can be influenced and those that cannot, to be very clear about the purpose of the measure and its definition and finally to ensure enough thought goes into understanding what behaviour will result because of the measure itself.