Data-informed, not data-driven
People bang-on about product managers needing to be data-driven. But like many things in building products, it’s more nuanced than that.
Being data-driven is simplistic. A machine can make decisions based on data. Algorithms can be made to choose the best path forward based on data. If it’s simple for a machine to do it, where’s your added value?
The truth is that to be a great product manager, you need to be driven by strong instincts more than anything else. The best product managers, the founders of companies, are guided by strong instincts for what customers want. Most of the time when going from 0 to 1 there is no reliable data to go on, and you must rely only on instincts.
Data is useful, it can help lead to insights, sometimes really powerful ones. But you want to be data-informed, not data-driven. It’s a subtle but important difference.
The mental model
As a PM, you need to build a mental model of your customers, taking all kinds of inputs, from quantitative and qualitative sources. The best product managers tend to have an extraordinary level of empathy for their customers, they can see things the way the customer sees it. Data alone doesn’t allow you to empathise enough to create this kind of rich model.
Relying on data as your primary source is lazy — we have lots of tools to collect and organise data now, it’s doing the work for you. The hard stuff is qualitative and experience related. You need to talk to customers, see their world, see their needs. Read about the problem space, talk to experts, see what other solutions exist, look at adjacent spaces. It’s total immersion to create that mental model which gives you confidence to make good decisions.
The problem with focusing so much on data is that you tend to focus more on things which are easily measurable. You end up knowing a lot about the numbers, but not a lot about the why.
Some things are hard to measure, but it’s those things which often lead to breakthrough insights about unmet needs.
But why is data driven so popular?
Because the big companies talk about it so much.
And if you work for one of the giant tech companies, then it does make sense that you are strongly led by data, for two reasons:
These companies have an insane amount of data about their product usage and users. It’s so far ahead of what most companies have in reach and scale. When looking at the practices of these companies, you need to take their unique context into account. They are not like you. But yet their practices get widely adopted by others who have an utterly different environment.
These companies have so many employees, that they need a way to have some control over their PM’s decision making. If they operate on a strong data-driven principle, then it adds a layer of control over what can change. It would be hard to say to 1,000 product managers: go learn about your customers, develop incredible instincts for what they want, then go build it!
It’s also true that if you aren’t a very good product manager, and you haven’t immersed yourself in the field to cultivate that deep understanding, then you definitely better cling to the data, it’s your only hope.
Instinct first, data second
The reality is, most products get created by people with a strong instinct for what to build. Most successful products make it through their initial phases of growth based on this instinct. In later phases when it’s more incremental in nature, then data becomes a useful tool for deciding what to iterate on, what to optimise.
But recognise the difference here: instinct is greater than data. Instinct is what the better builders develop and use, and data is what the lesser builders need.
Google’s initial search product was born from the strong instincts of Larry and Sergey. They had a deep insight about a better way to do a search engine. But since then, the overwhelming majority of Google’s homegrown new products have been failures — probably built by data-driven PMs. Their successes have come from acquisitions — YouTube, AdMob, DoubleClick, XL2Web (became Google Sheets), Writely (became Google Docs), Tonic Systems (became Google Slides). These were all built by founders with a strong instinct for solving a problem.
I’ve spent my career working for founders of product companies. In every case, the founder created something based on gut feel. They were immersed in a field, it provided an insight, they took that and ran hard with it. They trusted their instincts. Only when they hired people who didn’t have that same level of instinct did the question of data suddenly become relevant.
Instinct first, data second.