Think about the one objective you want your data science team to accomplish this year, and then let go of everything else. Otherwise, you jeopardize team morale and productivity.
For data scientists to properly analyze a root cause or an optional product offering, they must have variables to consider. However, it's dangerous — for you and for them — to pursue multiple objectives at the same time.
As a management consultant who specializes in execution, I've been in a lot of situations where leaders feel they're not accomplishing everything they want. Nine times out of ten, they're trying to accomplish too much at once. The situation is exacerbated when you're working with a data science team. If you want to get the most from your data science team, you need to get them to focus on only one objective at a time.
Save massively parallel for processors
There's significant power in the number one. Legendary CEO Jack Welch suggests that companies come up with only one metric to define success. There's a reason why.
When an individual only has one thing on their mind, the person's focus is clear. When this concept is extended to a small team, the results are exponential. This is how top athletes perform at their peak, and it's how your top data scientists will produce their best results.
Time and time again, I've seen leaders try to eat more than they can chew. When I was consulting for a well-known computer company, leaders would go through a process of determining which projects would be the focus for the year. They would organize projects in order of priority on a spreadsheet, and there was a thick, blue line that would serve as the cutoff point. Anything above the blue line would get worked on — anything below the blue line would not (until somebody screamed and they would make an exception). I don't know how they came up with the math for where the blue line would land; however, it always turned out that people were working on way too many things at once.
Some people can juggle multiple things at once, and some people even thrive in this environment — as a general rule, data scientists are not these people. Data scientists typically have a precision mentality. Precision requires focus. That's why most data scientists are introverts; it helps them focus enough to get the level of precision that makes them feel satisfied. If there are too many distractions, the quality of their work goes down. They know this, and they hate this. No matter how hard you try to enforce the 80/20 rule, a quality standard of 80% is never going to sit well with them.
Find focus and then let go
Filing your objectives down to one is not a difficult exercise, but it does require a great deal of discipline. The most common cause of leaders having too much running simultaneously is rooted in emotional rather than technical issues. It's hard to say you're not going to do something. You might wind up having your data scientists explore the best markets to pursue, analyze fraud patterns to reduce risk, create dashboards to increase organizational capability, and so on.
Instead, you should figure out what your single most important outcome is for the year, and make everything else secondary. With this kind of clarity, you'll be amazed at what your data science team can accomplish. Your data science team can even help you with this exercise. Before anything else, put everything you want out on the table and have them analyze the best objective to pursue. What they come up with will be quantitative, so don't make the decision purely on their analysis. They will, however, provide insights that you may not have considered. You should gather all of this information, apply your best judgment, and then make the final decision.
Next comes the hard part: letting go of everything else. It's not that it won't get done — it just won't get immediate focus from your data science team. During execution of your single objective, you will be haunted by the ghosts of the ideas that didn't make the cut. Resist the temptation to divert focus until your primary objective is done. Even the slightest hint that you're moving off your mark will communicate hesitation to the team; at that point, you lose focus and credibility.
It's tempting and psychologically easier to pursue many objectives at once — it's also damaging to your data science team's morale and your overall ability to execute objectives. You don't need sophisticated project selection algorithms or a blue line — you need the discipline to select the one objective that will be accomplished by the end of the year to move your strategy forward.
Consider this question: If you could only accomplish one thing this year, what would it be? Once your mind is set, rally your data science resources, and let the power of focus take over. You'll be amazed at how quickly things get done with a tightly focused group of data scientists.