The truth is that hiring for data science is in many ways more of an art than a science. That does sound oxymoronic, but that does not make it any less true. The reason is obvious. Data science is so new that it can be hard to know what you’re actually looking for. What is the set of skills and abilities that will make a data science team fly and which one will make them flounder?
If you don’t know, then you’re certainly not alone. Fortunately, we do some years of company experience to draw on. What’s more, IT has been with us for nearly two decades and there are plenty of valuable lessons there too that we can implement in data science hires.
So let’s see what we’ve learned so far.
It’s not only about the numbers
A lot of companies think that if they just get a couple of people who are incredibly good with numbers, then it will all sort itself out. That couldn’t be further from the truth because the numbers alone won’t get you anywhere.
In the data community, there’s a famous saying: Garbage In Garbage Out. When they say it, they’re mainly talking about the quality of the raw data, but it should really be applied far more broadly. Here’s how:
If your data science team doesn’t understand the business that you’re in or the numbers that matter, then they’re not going to be paying attention to that and so much end up giving you suggestions that are actually counterproductive. And you won’t have a clue as you won’t be able to double check their work, as most of us don’t have any idea what happens in the black box that is big data.
For that reason, you can’t just hire people good with numbers. You also need to hire people who understand your business. In other words, you need people who can bridge the divide between what you do and what they do.
When you hire them, ask them to explain something difficult
Some data scientists will say that they can’t explain something. Others will be reluctant to try and explain it. On the other hand, there’s a third category who will try their absolute best to explain complicated concepts in accessible terms.
You want at least one of this category on your team, as they’ll be able to make sure that you actually understand what’s going on. Even better, because they can explain things in simpler ways, they have a better general oversight than most of the pure data scientists do and can, therefore, act proactively to secure and improve your business.
Variety is the spice of data
It doesn’t end there. Really, there should be people from all the departments that will have some bearing on big data involved in the team. These people don’t necessarily need to be part of the team full-time, but they will need to have enough time available to understand what is going on and make informed decisions.
This will make it far more likely that your data science team will actually deal with the real problem rather than what they (or you) think is the actual problem.
Of course, whatever people you bring into the group do need to appreciate numbers. They actually need to be able to pick up what the data science team is doing. For that reason, find people that have a real interest in these kinds of things and have a thirst for picking up information. Otherwise, the chasm that you’re trying to bridge will instead get transported into the very data science team, which will slow things down considerably.
Of course, these kinds of people that understand both numbers and the business aren’t easy to find. So why not consider training them up? There should be plenty of people who understand the importance of this kind of thinking and willing to learn.
If you have somebody like that on your team, it might be worthwhile to get them to take a number of courses or even attend university part-time. After all, they don’t need to the whole understanding right off the bat. Instead, they need a strategic understanding so that they can communicate with the people who have the deep understanding.
Some positions that will benefit from having a high IQ in terms of the numbers are Chief Executive Officer, Chief Data Officer, Director of IT, Human Resources Manager, Financial Manager, and Marketing Manager.
Data science is a group effort
You know all those stories about lone wolves working in laboratories and coming up with fantastic insights? Well, they’re just that “ stories. They have no basis in reality. Big data is something that you do as a team. For that reason, if you’re serious about big data, then you’re probably going to have to hire a whole team to do it.
Alternatively, give the department a big enough budget to allow them to contract outside help when they run into problems that are too big for them to handle alone. This will also allow you to possibly hire these people somewhere down the line when you realize that one particularly skill is absolutely essential for your team to excel.
There are strategy and function
I’ve already briefly touched upon this above, where I mentioned that the people who understand your business can be of great help if they have a strategic overview of what’s going on. It’s great to have a few of these people.
Note, however, that not your entire team should be made up of these people! You still need the actual number guys who can build the algorithms and do all the other stuff that’s important. If you don’t, then you’ll have some great strategic plans, but nobody with the capability to actually execute them.
Have the right tools in place
You might think that if you give data scientist the right data and lock them in the broom closet long enough, they can do their magic. It doesn’t work like that. Data scientists “ like almost any other profession “ are only ever as good as their tools.
For that reason, make sure that the right systems are available and make sure that the people who are making the overall decisions aren’t just deciding what way the team should head but what tools they should get to get them there. Sure, the data scientists might be able to work all the things out without those tools, but it will be far more time-consuming. And as data scientists aren’t cheap, that will probably put you back far more in the long run than giving them systems that allow them to work faster and more effectively. If you see something new, don’t forget to read reviews of the best websites.
If you’ve got the right team, give them the right environment
If you’ve followed all of the steps above, then there is one more vital step that you’re going to have to take and that is to stop telling these people how to do their jobs. A lot of managers fall at this last hurdle. They get way too involved in the nitty-gritty of the data science team and as a result, end up steering the team in the wrong direction.
And that, obviously, means wasted resources. If you’ve set the team up correctly and you have people who understand the problems that your business faces, then the most important thing for you to do is to give them space to actually find and solve the problems that your company actually has, rather than the ones you think it does.
If the team has the ability to think freely, then they might just come to you with that million dollar insight. If they don’t, then they’ll probably only be able to do nickel and dime work.
Last words
As I’ve hopefully managed to make clear, data science is a big concept. There are a lot of moving parts that go far beyond the actual numbers. Most importantly there’s the ability for two-way communication between the data scientists and everybody else in the company.
If you can get that right, then your company is going to benefit a lot from the data scientists. They’ll be able to make suggestions that are actually implementable and the rest of your company will be able to give feedback in a way that the data scientist can actually use.
If you don’t have that bridge, however, then things will quickly fall apart. The data scientists will end up looking at things that people might think are important, but don’t really matter. Alternatively, the data scientists think they’re exploring something vital when in truth their ideas are interesting but can never be applied. And that, obviously, is a waste of everybody’s time (and will lead to high turnover as well).
So focus on that bridge. Make sure you have the right people in place. And then the rest will happen pretty much as if by magic.
