Make Less Data-Driven Decisions
If you want to make more data-driven decisions, try making less.
Ideally, just one.
I call it data-driven minimalism. I’m serious. Ignore all the stuff you’ve been underpinning with data, and discard all your dashboards and projects. Choose just a single decision, and go all in. Pour all your data and all your data efforts into this single decision and make it a good one.
I know that sounds weird, but it works. I know it works because I’ve seen it at work dozens of times. I’ve experienced it myself, again and again.
I’ve also seen the alternative fail, almost every time. Amazon has a huge BI team and hundreds of dashboards because they started off by focusing on one decision top decision, not the other way around!
All-In Data-Driven Marketing
John is the Head of Marketing at a hot new startup. The company pivoted around quite a bit, but John has a good set of dashboards to control and optimize the marketing activities. But the company is still searching for a product-market-fit.
A couple of weeks ago, the company started to focus on a new customer segment and John is in charge of building an audience inside this new segment to sell the soon-to-be created product to them.
Since John already has a good data-driven practice in marketing, he was confident he could quickly build out an audience. He discussed the set of 30 metrics every single week and adjusted the marketing activities accordingly.
And yet, he didn’t seem to get anywhere. Yes, he was starting to build an audience, but not as quick as it needed to be.
Common wisdom would tell him to include more metrics, run experiments, and see how to grow. But that’s not what he did.
He tried something different. He threw away all the data and all the metrics because he felt they weren’t the right ones. John realized he truly only had one number to grow: The maximum audience he was reaching. And only one decision to update every week: Which channel should we focus on to grow the maximum audience?
So he started to drill down, redo the entire marketing metrics framework, and focus it on this one top number. He added input and output metrics focusing on the individual components that make the number up. And sure enough, he quickly saw what we needed to do: Whenever he added a new channel, the numbers went up.
So, John realized he simply needed to grow the number of channels and distribute his content more! There wasn’t one optimal one, a fact he could’ve never seen using his old and thorough framework.
It Won’t Work For Me
If John would’ve kept his old metrics, he would’ve never gained this new knowledge. He would’ve kept on doing little optimizations, but never would’ve gotten growth.
I can also guarantee you, that John will need to adjust his framework to gain more knowledge, he’ll need to add more decisions as he goes on. He will need to discuss his metrics every week and think deeply about the key decisions he wants to make.
I know what you’re gonna say, But John is ignoring important numbers! Other decisions need numbers, too! Is he not tracking the Slack community growth at all? (Nope) Isn’t he tracking his email opening rate? (Nope) I make lots of decisions, not just one. We can’t afford to just focus on one decision.
What I’m saying is… if you don’t limit yourself to just one decision, you’re not gonna go anywhere with every one of the others.
All other decisions need depth, too. And you can’t develop the muscle needed to make good data-driven decisions by going broad. It won’t work. You’re not gonna become an expert chef by changing your area of expertise every other month. You’ve gotta build up that muscle; your organization needs to build up the muscle.
That muscle comes only from one thing: Focusing your data on one decision. Focus.
Moving On
Yes, there is a time when you’ll feel the need to move on, add a second decision to your rooster, or add more data. But I haven’t seen a single person so far that has moved on too late.
In fact, the whole point of systems like OKRs is to arrange all decisions into a hierarchy that leads up to one single big decision. Once the higher-up decisions change, you’ll need to change all the data you’re using below.
Airbnb has focused on just one question for years: How can we scale? And you should, too.
The goal isn’t to be good at using data in decisions, the goal is to build up the muscle to be able to change your data base when new decisions arise. The goal is to adapt to change, not to sustain the status quo.
Questions To Start Right Now
This will only take 5 minutes of your time, but I recommend you write this down:
What is the single most important decision you need to make on a daily basis? (Stop! Answer this one first before you continue; think hard about it!)
What % of your data efforts are directed at supporting this decision?
What are the rest of your data efforts supporting?
Related Writing
I detailed the story of how Airbnb turned itself into a data-driven company by using business intelligence. What Airbnb did from day one was to focus on the question of scale, and to my knowledge it still does. Read that to understand how this journey can start.
The book “Working Backwards” about Amazon has a ton on their Weekly Business Review (WBR) practice. It is a fascinating story of data-driven decision-making in action on a large scale and is very much in line with this thought.
Finally, I wrote a post on the story of Signet Bank (now Capital One) and their thorough approach to data-driven decision-making that transformed the company from a little town bank into one of the largest financial powerhouses in the US. They, too, discarded everything and focused on just one decision on offering the right loans to the right people.