The Scream
Sometimes I sit there wanting to scream. Analysis, graphs and data surround me. People are trying to convince me that the right thing to do is X, or sometimes Y or Z. But they have just filled my head with information and speculation. I have lost track of what they are trying to tell me, let alone what the decision is that I am supposed to be making.
My brain wants to explode. Why do we make things so complicated? I can feel a migraine coming on. Then, to make matters worse, people start to debate the numbers and question them:
“is that £1.6 million an over-spend or double counting?”
“How does the 1.6 relate to the 2.3?”
“It doesn’t; the 2.3 comes from the 1.7”
Now my mind is starting to drift; it can’t cope anymore. It is overloading, and random thoughts start to drift into prominence:
Do the accountants/analysts/actuaries know what they are talking about?
Why do these people always choose job titles beginning with an A?
Why don’t their trousers reach all the way to their shoes?
Do they really need to have sticking plaster holding their glasses together?
Then my boss asks me what I think, and it all drifts back into horrible focus.
Making It Easy
The thing is, numbers are easy. They don’t lie, don’t move, and don’t change their minds. It isn’t the numbers themselves that are the issue. It is our interpretation of them that causes all the problems. There are only four things you need to know if you want to cut through the blizzard of data:
- What your customers worry about
- The things you can change
- The bikini principal
- The paranoia of perfection
What Would Your Customer Worry About?
Key performance indicators
Possibly the most overused and disliked phrase in management, KPIs. What are they? They are key performance indicators at the risk of pointing out the obvious.
Key: -adjective Chief; major; important; essential; fundamental; pivotal. a key person in the company
Performance: -noun How well a person or machine, etc. does a piece of work or activity. some athletes take drugs to improve their performance
Indicator: -noun Something that shows what a situation is like. commodity prices can be a useful indicator of inflation
Where we struggle is with the word “key”. Key implies that they are pivotal, so there should be very few key performance indicators. Otherwise, they wouldn’t be “key”. In most businesses, however, there are hundreds of them. Because there are hundreds of them we quickly lose sight of the wood for the trees.
Who decides what is key?
Customers give us money, this pays our mortgages and we should worry about what they think. It isn’t “key” if a customer doesn’t worry about it; it isn’t “key”. Say, for example, you fix washing machines for a living. Your customers only really have three questions:
- Did you fix my washing machine?
- Was it fixed quickly?
- How much will it cost me?
These are your key performance issues. Once you know these working out your key performance indicators is easy.
There is a simple technique called a metric tree. The principle is that you start with the trunk. That represents the key issue. Then you branch off from that. Developing more and more detailed measures to create the tree. The customer’s “Was it fixed quickly?” tree might start like this:
- Time to fix: the length of time a customer has to wait from reporting an issue to having it repaired.
- Time to job allocation: the time it takes from receiving a job to giving it to an engineer.
- Time to visit: the time it takes from an engineer receiving a job to seeing a customer.
- 1st-time fix rate: the proportion of machines that are fixed the first time and don’t need a second visit.
Doing this exercise makes it clear which measures are essential and which are secondary. What to worry about and what is just a distraction. A clear metric tree makes life easier.
Worry About the Things You Can Change
Now you are clear about the measures; it is worth splitting them into two camps:
- Measures that look at outputs (lagging measures)
- Measures that look at inputs (leading measures)
This distinction is important because it is difficult to do something about outputs. The die has been cast. Lagging measures tell you what has happened, whilst leading measures tell you what is likely to happen. There is little point in crying over spilled milk.
Several years ago, there was a terrorist bomb at Heathrow airport. I was unfortunate enough to be flying to Germany the following day. When I got to the airport, an army of police officers was milling about. They were all carrying guns and shepherding people around. I got a bit lippy (a personal failing) and asked one what he was doing. He smiled wearily and replied, “We are conducting operation stable door, Sir.” The horse had bolted; they were wasting their time, worse still they knew it.
It is far more helpful to worry about the input measures; you can do something about them.
- If you are making sweets, worry about the quality of the sugar you buy
- If you are servicing cars, worry about parts availability
- If you are running a hotel, worry about staffing levels
Those are the things you can do something about. The horse has already bolted if the customer starts to complain.
The Bikini Principal
The principle states that “less is more”. Your challenge is to explain all those different ratios and measures simply and coherently. You have to do it so that it doesn’t make people feel as though they are doing mental gymnastics. How can you do that when you are surrounded by so much complexity and confusion?
Only worry about the big things
In the 19th century, there was an Italian economist called Vilfredo Pareto. He noticed that 20% of the people in Italy owned 80% of the wealth. This was the birth of the Pareto Principle or 80:20 rule. It states that not all things are equal.
If you only take one thing away from reading this please let it be that not all things are equal.
This is a hard thing to get your mind around so think about it this way. How many pairs of trousers do you own? 3, 5, 27? And do you have a favorite pair, a pair that you wear much more than all the others? I am sitting here writing this in a rather tatty pair of Levis, which in truth should be thrown away, yet they are the most often worn pair of trousers in my wardrobe.
Here is another example. How many people do you know, 100, 200, or 300? On LinkedIn, I have 2,482 contacts (and climbing). Now I admit to being an electronic philanderer, but how many people are important to me? How many will I make the effort to go and see? My wife, daughters, siblings, mother, and a handful of close friends. Let’s be brutal about it, of the 2,500 plus people I “know” at most; I only care about 40 of them.
Not all things are equal (sorry about the highlighting, but it is important).
So when you are presenting data, ask which information is essential and which is just noise. If I am working in London the time I need to allow to get there is driven by the time the train takes. How long I took to eat my corn flakes doesn’t matter.
The Paranoia of Perfection
We spend weeks trying to get to exact numbers. The precise number of calls or letters we answered and the cost we incurred treating each patient or the time is taken per query we dealt with. We plaster everything with caveats and excuses if we can’t get the exact answer. Measurement always has a degree of error and we spend lots of time trying to minimise it. We don’t stand back and ask ourselves how good the measure has to be. What is the impact of getting it wrong?
If I ask you how far it is to your corner shop, you will estimate, ½ a mile, a 10-minute walk, or a couple of minutes in the car. You won’t tell me that it is 825 meters, 32 centimeters, and 7 millimeters. We won’t debate where precisely in the shop I want to be, and am I going via the front or back gate. You know that nobody is going to die of starvation if you get the number wrong.
Yet when we are measuring things at work we obsess about getting it right. We forget about the decision we are trying to make. Next time you are up to your neck in spreadsheets, ask yourself: “What do I need this number for, and how precise does it need to be?”
It is better to be roughly right than precisely wrong
Carveth Read
Homework
Sit down and have a look at your monthly management information pack. Take out a pencil and paper and start to critique it:
- Pull out the things that you would worry about if you were a customer.
- Decide which the 4 or 5 important things are.
- Draw up some trees to show how the different measures link together.
- Highlight the things that are inputs not outputs.
- Get agreement with your managers and peers that these are the important measures.
- Restructure your management information packs to bring these measures to the front.
In the next lesson, we will discuss how to present the information to provide clarity and promote action.
Thank you for reading.
Data is not information, information is not knowledge, knowledge is not understanding, understanding is not wisdom.
Clifford Stoll
Related posts:
- Measuring employee performance (a shortcut): What would your employees measure?
- Measure anything the Sesame Street way: Linking measures to purpose.
- Is your boss really that stupid?: Are you measuring things to look good or improve performance?
Across the web:
- Measure up blog: Stacey Barr writes about “practical performance measurement”.
Further reading:
Bernie Smith has written a practical guide on how to put a KPI framework in place. According to one Amazon reviewer, it is “Highly practical, packed with reusable materials and tips, and written in good humour this book stands out from the more complex, self-important and ultimately forgettable crowd of pristine management books that litter my shelves.” More importantly, Bernie is a friend of mine. So you should have a look at his book.
Post Script
Did you get here from a link from a friend, Facebook, or Twitter? This lesson is part of a 16-part free e-mail course. Learn more about it and sign up here.