Sometimes life throws you a curve ball. I'm not sure this has been truer for me than recently when last month I ended up with no job and a house full of puppies.
This unexpected situation gave me a lot of holes in my diary/socks/shoelaces/furniture and the opportunity to clean up more excrement than when we potty trained our last child. It also presented me with a valuable exercise in making sure that, before you measure something, you know what that measure will be used for...
The Challenge
I thought having four children made my life pretty hectic, but I realised how wrong I was when my wife recently offered to care for a new mother dog and her 7 puppies. The owner was a dog breeder who had sadly had to go into hospital unable to look after her heavily pregnant dog. My wife valiantly stepped in and offered to care for the mum dog and find homes for the puppies, which meant a few things: our full house got a whole lot fuller overnight, my wife's Facebook feed suddenly got dramatically busier, and I started getting a lot of attention from my children discussing the merits of family pets and how lovely it would be to keep a puppy.
As part of preparing the puppies to go to their new home my wife and I were responsible for giving them some important medicine. We established that the medicine was administered based on the weight of each dog so we endeavoured to weigh each puppy...
...this was not easy.
Plan A
We started by putting them on the bathroom scales. This presented two major problems
- The puppies wouldn't sit still on the scales and holding them down wasn't an option
- The puppies weren't heavy enough to trigger the scales to turn on anyway
On to plan B.
Plan B
The next option was to put the puppies on the kitchen scales. This seemed to work better although still presented some challenges
- The bowl for the scales would not easily contain a puppy so we had to replace it with a large tupperware box and reset the scales appropriately
- the box containing a moving puppy wouldn't stay put and risked a tumbling pup - so we had to try to hold it in place whilst not affecting the actual registered weight
- A couple of the puppies,who were clearly keener on their milk, were heavy enough to go off the scale
On to Plan C.
Plan C
The third option we tried was for me to stand on the scales, record my own weight and then be handed each puppy in turn to record the increase. This worked pretty well with only a couple of issues
- We quickly realised that the kitchen scales were inaccurate over 2 kilos so all the puppies had to be re-weighed
- Once the bathroom scales had registered a weight they would not detect further changes so I had to repeatedly step on and off the scales with each puppy
So after much puppy-juggling we had established a weight for each pup.
Success?
Finally, flushed with the success of our efforts we had all of the puppy weights written down ranging between 2.4 and 3.2 kilos. It was only at this point that my wife opened the medicine and read the exact dosage instructions
"Between 2 and 4 kilos - dosage 5ml".
Not only was this range so broad that all of the puppies needed exactly the same dose, but all the puppies were so obviously within this range we could have guessed their dosage without even touching the scales. I furnished my wife with a suitably scathing look when she read this out loud, however my heart wasn't in making fun of her over it as I clearly should have checked beforehand.
We'd fallen into a classic trap. We had not maintained a connection between the process of measurement and the purpose of that measurement in the first place. The task at hand was not to weigh the puppies. The task was to support a decision around the dosage of medicine to give them. Yes, the dosage was based on weight, but only at a very high-grained level which could have been easily estimated just by looking at them. As Hubbard reminds us in his classic book How to measure anything every measurement should be targeted at informing a decision, and we had not identified the decision that we needed to make from the measurement which was simply which one of a broad range of dosage categories to give a puppy.
Are you weighing puppies?
It's unlikely that anyone reading this will find themselves thinking that they made exactly the same mistake when they weighed their puppies. I'm pretty sure that there are a few people out there who may be actively measuring things without a clear understanding of the purpose of that measurement.
- Taking fine grained measurements that are only looked at at an aggregate level
- Capturing data into dashboards from which no decisions are ever made
- Tracking vanity metrics that don't give a genuine understanding of a situation or predict the need for action
- Tracking hundreds of KPIs when you can only realistically understand at a handful
- Capturing data where you don't know what downstream processes do with it
In discussing this subject earlier this year with a former colleague he recounted how in a previous role he maintained a daily data load with an unknown purpose. On questioning the reason for this daily load he never got a satisfactory response, who this data was loaded for and what they were doing with it was a mystery to him. So in the end he tried an experiment - he stopped loading the data, and waited...
...no one shouted, nothing stopped working and no-one ever asked for the data load to be turned back on, he simply got an extra half-hour back in his working day. He was essentially weighing puppies every morning.
A Dog's Life
The short 12 weeks that the puppies been alive I've also seen something of another curve-ball (though this one partly self-prompted) and a huge amount of learning on my part that culminated in the shutting down of both my product and role. I'll be sharing the key lessons in the next few weeks here on the 1 year journey I've just been on, which in itself has included one or two examples of 'weighing puppies' - I look forward to sharing these here as well as what my next role is, whatever that may be.
Oh and in case you are wondering, yes I inevitably caved in, and we are keeping one of the puppies.
References
The highest recommendation that I have on the subject of measurement is the classic book I mention in the post - "How to measure anything" by Hubbard who provides some great examples on effective measurement and estimation:
Post a Comment