“We know how to measure things”
It wasn’t always like this…. Agriculture is maybe one of the oldest disciplines to use systematic measurements. But even 10,000 years ago, measuring in agriculture was always tied to control: farmers measured the rain and the soil conditions.
There’s one thing these measurements have in common: farmers can do something about them AND when they do, they can increase the crop yield! In the past, measurement was born out of necessity, out of the necessity to control the outputs of a process. And farmers sure worked hard through years of experimentation to find out, what inputs worked best or at all.
But today, things are different.
Today, measurement means simply that
Most people say “we know how to measure things”, and mean it. They got numbers. They have dashboards. They make sure new product features have a numerical goal attached to it. They make sure they have SLAs, and clear numbers to focus on. They have a north star metric, something everything should work towards raising.
You can find one phrasing of this in one way or the other in most product manager job descriptions.
We all got measurements, dashboards, numbers but, we lack control. They threat most aspects of their business an product the same way they threat their height. We know definitely how to measure it. And we can measure our food intake. But it’s cristal clear, that our height is just that, a number and something we cannot control. Similarly, many business metrics are just numbers without actionable control.
In reality, you don’t want a product manager who merely measures things; you want one who has demonstrated the ability to deliver results.
Some companies and people are different - they would never say “we know how to measure things.” At Amazon, people say “We deliver results. Leaders focus on the key inputs for their business and deliver them with the right quality and in a timely fashion. […]“. To deliver results, somewhere in that process, Amazonians will measure lots, but measuring is neither a goal, nor a primary tool towards delivering results.
Capital One one of the largest financial companies in the USA was born as a spinoff of a small consumer bank called Signet Bank from Virginia. Today it’s a power house of big data. But it wasn’t born with a vision to “measure everything, know everything about the consumer.” Instead, “the core idea behind the company’s formation was “information-based strategy” that important operational and financial decisions should be made on the basis of analytics and data” (Woxsen U).
Bridgewater Capital became the biggest hedge fund in the world by having a founder that realized, the combination of machines and humans can deliver better results, remove emotion and crunch more data together and thus make the best decisions possible.
Control is at the center
Amazonians knows to follow the Six Sigma DMAIC process to tackle problems. The first step of the DMAIC process is simply called “Define” but it is by no means simple. Defining means doing the hard work of uncovering what inputs control your process.
And that’s the step that almost everyone skips. Ask yourself “Do I really know what inputs I need to push up in order to influence my output?” I rarely meet people who can provide a good answer to that question.
Ask yourself “If I had unlimited resources and time, what would be the features that would give me 100% market share?” If you can’t answer such questions, you don’t have control.
For Amazon the answer 20 years ago was “Convenience, Price, Selection.” Every item in the world, at zero price (or lower than anyone else), delivered to you at the most convenience - that’s how you get 100% market share.
All of those inputs are controllable.
Dive deep - work hard
There’s no magic recipe to derive controllable inputs of a system. Well, there is one, but there’s a reason most companies are not Bridgewater Capital or Amazon. The recipe is:
Dive deep,
work hard,
don’t give up.
Ray Dalio, the founder of Bridgewater Capital, spent decades refining his trading systems. The team behind Capital One spent years gathering different kinds of data until they were able to find the right controllable inputs to deliver the right products and services.
Amazonians know, if you think you found the controllable inputs for a process within 24 hours, you probably didn’t.
Finding controllable inputs means running experiements. Amazon ran year long test marketing campaigns only to realize, Ad spent alone isn’t controlling the process.
Find controllable inputs
This might be the most unsexy call to action ever written. But it’s the one I got: Stop measuring and start identifying for controllable inputs.
You know you’ve found a controllable input if you push it, and your output reacts. It is a simple as that. But this simple effect is hard to discover.
As Marcus Aurelius advised, "Waste no more time arguing about what a good man should be. Be one.”