My 4-year-old daughter is learning to count. She is obsessed by it. She counts everything:
1.. 2.. 3.. 4.. 5.. 6..
Everywhere we go.
Most of all though she counts steps. Particularly in castles and cathedrals (which is fun) and car parks (which when you are laden with bags full of shopping… that you just want to put it in the car… before your arms drop off… is less so).
In truth she isn’t very good at counting. If we are counting the steps to the top floor of a car park her number and mine never agree.
There are two reasons why she is wrong (and I am right):
1. She makes mistakes:
Counting is difficult:
- Fifty seven
- Fifty eight
- Fifty nine
- Twenty
- Twenty one
You get the idea.
2. She isn’t clear
3 steps up and 1 step down makes it 2 steps high (Daddy logic).
3 steps up and 1 step down makes it 4 steps (yes but it is a step Daddy, I just stepped it).
Adults like to count as well
We count a lot, just not steps, we count calls and customers and jobs and money. We call it management information.
Of course management information isn’t information if it is wrong. So we obsess about making sure it is right, that the numbers match, that we have a single source of the truth.
But expecting a single truth is naïve. If two people count the same thing they will count it differently and you will get two different numbers.
1. We make mistakes.
We count complicated things. Imagine working out the number of calls that come into a call centre that gets:
- Sales calls
- Quote calls
- Service calls
- After sales calls
- Out bound calls
- Calls from customers
- Calls from people who aren’t customers
- Calls from staff members
- Transferred calls
- Chase calls
All those calls are recorded on different databases, at different frequencies, using different technologies.
The more convoluted the thing we are counting the more likely we are to get it wrong. And we all love complexity. So we get the numbers wrong all the time.
Nobody is perfect.
2. We aren’t clear
We all define things differently; when is a call a call?
- Are we talking sales calls or service calls?
- What about a service call that lead to a sale?
- Are internal calls calls?
- If a call is a wrong number, should you count it as a call?
- How about ring backs and outbound calls?
- Do you want Monday to Sunday or week commencing Saturday?
- Should you count a call from the second it starts or the second it finishes?
The answer to these and a myriad of other questions often depends more on the reason you are counting calls than the fact that a phone rang.
Please don’t be naïve
Don’t obsess about the precise number. Stand back and look at the bigger picture.
- Look at the pattern over time; is it varying the way you would expect it to?
- Look at the relative sizes of things, work out if the numbers are big or small.
- Apply the Pareto principle, work out which number is the one to worry about.
But most importantly of all, ask yourself why you are counting this thing in the first place?
Do you really need to know how many steps there are in Trinity Square car park?
When the lift is broken the answer is plenty.
It is better to be vaguely right than exactly wrong~ Carveth Read
If you enjoyed this post click here for updates delivered to your in-box
Read another opinion
Image by the_gain_card
Annette Franz says
Love it… great analogy… and great message!
Annette :-)
James Lawther says
Thank you
Adrian Swinscoe says
Hi James,
I get the funny feeling that we could apply this to the whole ‘Big Data’ industry and debate.
Stepping back to see the big picture and to ask questions is where the real value is.
It worries me that with so much number counting and crunching going on, that firms are not seeing the big picture and asking the Qs. Rather, they just getting stuck in the weeds because they don’t have the courage, wit, wisdom, experience or ?? to know the difference or do different?
Or, am I being unkind?
Adrian
James Lawther says
I think you are bang on the money.
Very few people appear to be capable of holding their heads upa and asking “what is the problem”
James
John Hunter says
Nice post, kids seem to make every topic more interesting. As you say the most important thing is to figure out what data you actually need. Then use an operational definition http://curiouscat.com/management/operationaldefinition.cfm so the data collection is sensible.
James Lawther says
Thanks for the link John