It’s been well documented that data science has grown and developed quite rapidly. It’s hard to believe that it’s been five years since Harvard Business Report proclaimed that Data Scientist was the sexiest job out there. Since that headline, the field has undoubtedly come a long way.
With this being said, I would argue that we still know very little for sure.
It’s important to understand that, at its core, data science is a still science. In science, very little is concrete. There is always new findings, publications, and research coming out that improve on previous work. Science is extremely iterative in this way.
“One important idea is that science is a means whereby learning is achieved, not by mere theoretical speculation on the one hand, nor by the undirected accumulation of practical facts on the other, but rather by a motivated iteration between theory and practice” — George E. P. Box
Data science is no different. Iteration is essential to our day-to-day work. Whether we are improving on previous machine learning models, exploring trends of key metrics, or investigating a complex question, it all comes down to iteration.
As Data Scientists, there is very little that is black and white. We do our work in a world of grey.
This comes off to many as consequence of the profession — as a bad thing. However, I beg to differ. At least to me, all the greys out there represent infinite possibilities.
They represent limitless ways to analyze and solve problems. They’re what makes data science so challenging and interesting to such a large population. Most importantly, they are what empowers smart people out there to make an impact in data science and beyond.
Make an Impact
At least in industry, Data Scientists are hired to inform decisions and make an impact on the business. All of the behind-the-scenes exploratory data analysis and visualization that you do will often manifest itself in a report or a well-informed decision of some kind.
You also may have noticed that whether you’re at a smaller operation or a bigger company, time is your most precious resource. There will always be another question to be answered, another feature to be analyzed, and another model to be built.
It’s imperative to the success of you and your team that you are ruthless with your time and priorities.
So with this in mind, let’s get back to the question at hand.
When Do I Ship?
Deciding when to finish up a project and move onto the next thing isn’t easy. Like many things in data science, there will normally never be one correct answer to this question.
However, this doesn’t mean that we can’t make this decision easier with the help of a simple heuristic. At the end of the day, it all comes down to the impact of your work. This post was inspired by a great article from Locally Optimistic which states:
“If they can’t make a decision with our output, then your job isn’t complete.”
As with most heuristics, this is a bit of a generalization, however I believe it’s an extremely useful one at that.
As Data Scientists, we need to do a better job of consistently reminding ourselves that our primary focus should be to drive impact.
If you’ve produced what you need to result in a well-informed decision, then ship the work and move on to the next task on your backlog. A common mistake let yourself get caught up in the trees and spend too much time on complexities with no actionable insights.
It would be easy to sit here and close things on ‘make an impact’, but I think that understates what I’m trying to say. Creating impactful work should be at the forefront of your mind at all times. It should be number one on your priority list no matter what project you’re working on. It should be a borderline obsession for Data Scientists in industry.
It seems fairly simple, but you may find this is much harder to implement in practice than it sounds. It may take some deliberate work and stumbles on your part, but with this advice by your side, I’m confident that you’ll get through it.
Thanks for reading! If you enjoyed this post and you’re feeling generous, perhaps follow me on Twitter. You can also subscribe in the form below to get future posts like this one straight to your inbox. 🔥