Drowning in Data
“A man who is not afraid of the sea will soon be drowned for he will be going out on a day he shouldn’t. But we do be afraid of the sea and we do only be drowned now and again.” - quote from an Irish fisherman - On Celtic Tides by Chris Duff
I’ve been contemplating this quote as I’ve now seen data initiatives drown again and again.
Drowning is a good metaphor, as data journeys can truly wreck a ship, i.e. the career of a data professional or even set the business back for years.
Drowning is also a good metaphor because the data environment is vast and wild.
You can do many things with data and start your road down a million different ways. Some are more likely to fail than others. And a lot can lead you into a data-driven future. It is not really that you must select the right path, but rather the path into the right environment so that it is less likely to fail.
The question really becomes: How do you choose the right environment to let your project succeed? I find that hard to answer. What I can answer, though, is what tells you you’re not in the right mindset yet to make a good choice!
Here’s my little list of signs that you’re not yet afraid enough of the data sea.
1. You want to try out 2-3 different things with data
What is more likely to get you through the storm, ten tiny boats with a one-man crew? Or one strong boat with a strong crew?
Of course, ironically, the quote above is from a book about a guy who circumvented Ireland in a small sea kayak. Still, the point is this: If you don’t focus, none of your efforts will make it.
If you don’t know what ship you want to bet on, you’re probably unprepared to go on the journey. If you don’t know what data initiative you want to start with, don’t start yet. Stay out of the data sea.
2. You just want to do something with data
While there are many paths that lead to a data-driven company, you shouldn’t draw the conclusion that all do!
In fact, while all roads lead to Rome eventually, the world is still infinitely larger than this one place, and if you choose a random road, you will not end up in Rome any time soon.
If you have one data project but don’t know why you think this is the one, you don’t have a good data project yet. Stay back.
3. You’re about to hire a small data team to do the job
Is a group of tailors all you need to guide a ship? No! You need a captain, a first mate, a navigator, and a complete cross-functional team, so to speak.
LinkedIn became successful with machine learning not when it started to pick the right projects but rather when it moved away from siloed teams. Only after it turned its machine learning team into a cross-functional team that was able to move completely independently was it able to launch the now famous PYMK feature.
You can't start your journey if you don’t yet have a true team.
4. You think about a data project
If I had to choose between a novice captain straight from naval school and an experienced one without official naval training for my own journey across the sea, I know whom I’d choose in a heartbeat.
Nothing beats experience when it comes to uncertain environments.
Data science, machine learning, none of this will be a project; it’s going to be part of your strategy, product, and processes.
If you still think of one-time projects, you’re not ready yet.
5. You’ve calculated your investment will pay off within the year
The book On Celtic Tides starts off with the story of how Chris paddles out into the fierce sea, only to turn back a couple of hours in.
He turns back to wait for another day.
It’s a mistake to think you can calculate the return on investment on your data initiatives precisely because this assumes you know when the right opportunity appears. But you cannot influence an external environment.
All you can do is to be ready.
6. You think you’re an expert
While I sometimes call myself an “expert,” it feels off. I don’t think sailors call themselves “expert sailors.” They simply are sailors or experienced. Because it is not the knowledge that matters; it’s the experience.
Similarly, in the data space, actual hands-on experience matters more than any knowledge you can read up and practice.
If you think you’re an expert, you’re not experienced enough.
7. You’re about to hire an expert
If you want to go into the sea, you should know how to pick a good captain. If you don’t, you shouldn’t go.
Same with data.
8. You’ve done it before
It’s great to have experience, but just like any sea journey, any data journey is different. Don’t let your experience get to your head; keep it level and focus on the uncertain environment ahead.
9. You haven’t written a business case
While the sea is uncertain, and there might be many paths to take, a captain still plans his journey! He just happens to be aware that he might need to change course and act fast.
If you don’t have a business case yet for your data initiative, you’re not ready.
10. You think you don’t need to spend more
A boat isn’t a fixed entity. If you overlook regular maintenance and checks even as you go, it will lead to catastrophic failures. It’s the same with your data journey, you must make sure you got some leeway for maintenance and potential changes of course, it’s the only way you will get to the other shore safely.