What does Measures and Metrics have to do with cars? Or for that matter vegetables?
Shh. Silence, please. A master chef is at work. There he is, dicing tomatoes. He’s efficient and yet relaxed. There’s an aesthetic sense in the air. He’s softly whistling a tune while he watches the food boil. Every few minutes, he uses his spoon to test the consistency or taste. He is continuously iterating – simmering the heat down or adding an extra dash of pepper. Watching a supreme dinner taking shape is a feast for the eyes.
Measures and metrics: A chef’s story
You sneak up on him and ask him how many teaspoons of mustard and oil he’s using. What temperature is ideal? His face goes ashen. You’ve unsettled him with all these “metrics” questions. He’s like the guy who’s been asked not to think of Monkeys. All along, he was in a flow, conversing with the food, creating a master dish. Now, we’ve disturbed that connection and the monkeys have taken over. Put him in front of the stove, and he intuits what the right proportions are. He cooks from “feeling,” not by thinking.
The chef’s “handful of salt” is the measure. He iterates it based on taste. Converting the handful into milligrams so that a novice can prepare the dish converts the measure into a formalized metric. Measures are for artists; formalized metrics are for the assembly lines.
Throughout my professional life, metrics and measures have played an integral part. They have guided crucial decisions and helped avoid catastrophes. But they work best – and indeed only – in context. Sometimes they take center-stage; other times, they only validate. And sometimes it’s best not to use them at all. This distinction between contextual measures and formal metrics, I believe, is becoming more and more relevant today.
And on that note, let’s head to Christmas Day.
Measures and Metrics: Car troubles!
It’s 7 am, the climate is pleasant, and the traffic reasonable. People are heading out for their first long weekend since the pandemic started. My car suddenly coughs up smoke, and on the dashboard, the dreaded warning temperature light pops up. It has developed some trouble in its internals 80 km from home. As my Roman friends would say, “O tempura, o mores!”
And the savior arrives – or does he?
I maneuver the car to the side. Once parked, I begin putting my phone to work. My car manufacturer is proud of their service, so I hope their emergency team will come to my rescue soon. Alas, it’s not easy. It is Christmas, and their service team is thin. The request for help bounces from one service engineer to another as they look to identify someone who can help me at the earliest. The earliest seems to be not so early. It’s two hours, and the SWAT team is still not here. We’ve made too many calls, but there’s no action yet.
It’s time to explore other options. My nephew knows a person who knows a local mechanic in the area – and we reach out to him. And precisely three hours after being stranded, two teams – the company’s SWAT team and an independent mechanic turn up simultaneously. Over the next few hours, I get a master lesson on measures and metrics. So now you know where I was leading you!
Do you own the measures, or do the metrics own you?
The SWAT team has three folks. There’s a young, bright guy who looks fresh out of college. Accompanying him is a senior looking guy who looks like he’s carrying the world on his shoulders. And there’s a third guy we’ll call Mr.Silent. Or if you prefer the third umpire. We never figure why he’s there.
Mr.BurdenedShoulders comes up and asks a set of questions – what happened, how did I figure something was wrong, etc. I have shared this information a hundred times during the day, so I rattle off the answers. When I tell him the warning light had come on (and without any warning!) along with smoke, his shoulders sag even further. He looks like he is mentally reading the manual for smoking cars.
“A warning signal, along with smoke, could mean many things. Let’s tow the car to the nearby station,” he finally says.
The younger guy looks like he wants to poke around a bit; this seems like fun lab work. But Mr.Shoulders is firm.
“We can’t take risks,” he says. I mention that I have to be at a place two-hundred kilometers away. Is there anything he could do to fix it today?
“We need to run the electronic diagnostics on the car,” he says. “And the equipment for that is at the Workshop 20 kilometers away. The Diagnostics will tell us what is wrong. And, today is Christmas Day, so the chances of getting parts are bleak. I think you’ll have to postpone your trip.”
Measures and metrics: Hands-on work!
I don’t like what I am hearing. I recall the experience of a premium automobile owner who had warning lights suddenly pop up on his car. His service advisor asks him to drive down to the Workshop carefully. Our man ends up driving this super powerful car at a very sedate pace for over a hundred kilometers. He suffers smirks from the scooter drivers as they overtake him. He braves it all and travels at snail speed to the Workshop to find – of all things – a faulty sensor! A broken sensor has triggered a wrong warning resulting in a rocket ship driven at rickshaw speeds. Can there be a sadder story?
While I am musing on such things and thinking dark thoughts about Mr.Shoulders and his company, Mr.IndependentMechanic is busy poking away at the engine. He is listening to engine sounds, sniffing for burn smells, and feeling the engine parts’ temperatures. His colleague is beneath the car, looking for leaks. I remember Sherlock Holmes searching for the stubs!
All this poking around doesn’t amuse Mr.Shoulders. How can an outsider be allowed to work on his company’s precious cars? Shouldn’t the computer have a first go at the diagnostics?
We solve the people side of things by ordering tea all around. Mr.Shoulders even smiles a little. The tea is that good.
Soon, Mr.Mechanic is sure he has found the problem. And interestingly, he’s done it is by reversing the process we use to solve problems.
Measures and metrics: Don’t forget to look at what is going well.
We usually are tuned to look for “broken” things. So our weekly dashboards tend to focus only on the reds. Anything that goes from red to green is a hero – we reward the project and project manager. To be precise, if something needs our attention, it better be clothed in red!
Sometimes, people also look for “watermelon” situations. These are situations where things appear green on the outside but are red. They are problems waiting to happen. It is a useful technique, but we are again only looking for reds.
The mechanic was looking for the green. He had narrowed down the potential failure points and had been examining how they were performing. He was looking not just at how they were performing now, but whether there were signs of wear and tear. His intimate knowledge of automobile architecture gave him an understanding of second-order effects. If these parts were in good order, other unseen parts should be working too – you didn’t have to open the engine head to examine those. In essence, he was looking for the greens to guide him. And when he had this list ready, he was able to isolate one module that was both performing below standard and was showing signs of wear and tear. He was confident that fixing that would solve the problem.
The junior SWAT guy is enjoying this analysis. He puts a series of “but what about this,” to which the mechanic patiently explains his rationale. A spirited lecturer-student type interaction is in play.
A half-hour later, we part as friends. There’s an unanimous agreement on having the mechanic fixing the car. We all feel more comfortable with this. Even Mr.Shoulders agrees. As I mentioned earlier, the tea was that good.
Garage time – every problem has many parts.
Back at the garage, we wait for the car to cool down. Our mechanic wants to do a final round of tests to reconfirm his analysis and determine which subparts need replacement. For this, he needs the car to be cooled entirely down. When you measure, you need to have the right context. He had performed his last analysis with the engine running and water being poured in the radiator – now the investigation requires the reverse.
The context of the situation demands how we should collect metrics, not the other way round.
Our mechanic meanwhile goes to arrange some more tea. And so we’ll break this section with a nice picture for you.
The great search and configurable solutions
Our mechanic is back fiddling with the parts, identifying ones that are in bad shape. He confirms the fan motor as the villain. The surrounding bits are all in decent shape. He starts calling the dealers now, beginning with the car company dealer network. Their inventory management system shows the product is out of stock. It is available only in Coimbatore and would take a week to deliver. Not acceptable to me.
So he puts in plan B. He knows he needs only one part, not the whole assembly. So, he puts on his helmet and sets out to the city to start talking to the OEM guys and other large dealers if they have the part in question.
Later that evening, he is back, part in hand. The car is soon ready. He reviews it thoroughly one more time and swaps out a wire he feels is too short. A couple of test runs, and he hands it over. He hands me his WhatsApp number and tells me to call him anytime. He’s super confident it will work.
Measures and Metrics – a summary for you
As I continue my ride, I remember I had promised some learnings on metrics and measures. We covered a lot of them in this post. If the story was so engaging, and you missed them, here they are summarised just for you.
Own the measures!
Measures are necessary because they direct our search. Metrics are formalized measures – they are proxies that work most of the time. They can inform us but should not be the sole decision inputs. We should own our measures and formal metrics – always!
Tailor the measurement approach to the outcome you desire
The outcome you are looking for determines the measures you will use. The Independent mechanic wanted me to reach my destination. So he looked at what was working right and thereby identified the culprit. The company emergency unit doesn’t want the car to break down again – so they look for what is possibly wrong. Both are practical ways – you need to decide which one works for you.
Experts “sense and iterate” all the time
Experts use contextual measures. They “talk” to their product. And hence identifying any gaps and iterating is part of the normal process. They don’t need benchmark data; they have an intuitive feel for what’s working. And they employ all their senses at work.
Formal metrics have constraints we should be aware of
Formal metrics are structured to ensure they “work” in most cases and can be interpreted similarly by people with varied skills. It’s, therefore, essential to remember that the metric is only a guide. A “no” metric need not mean a no go. We need to peel the onion a little more.
Technology and community work great together.
Every problem gets fixed by people. So expertise with the right connections will set you up for success. When you need solutions, go for people who can break the problem into smaller parts and then innovate from there.
And finally, the mustard question!
Don’t ask great cooks for how many grams of mustard they used. They wouldn’t know. And when they start thinking about it – your food will suffer!