How To Keep Your Data Scientists

How Can You Stop Them Being Poached?


You have gone through months of searching, interviewing and testing for the best possible data scientist and now that you have them on-board, the last thing you want to do is go through this process again in nine months when they have been poached from you.

Good data scientists are like gold dust today, with every company looking for them and few being available to fill these spaces. This means that when they see that somebody else has one, they will do almost anything to get them, meaning that once you have hired one, you need to create a better place to work than all of the other companies who want them.

But how do you do this? It is not as simple as keeping other employees happy, because their skill sets are normally found in others and can be trained to adapt to their exact position relatively simply compared to the highly technical data science roles. With a data scientist this is not the case, their skills are rare and training up somebody in-house or finding somebody else from outside the company is incredibly difficult.

We believe there are four key rules to keeping a data scientist on board:

Pay Well

This seems obvious, but paying well is the key to keeping a data scientist happy and working at your company. You need to be able to offer them a salary above what others are offering. Essentially, it is allowing them freedom to get the things they want outside of work rather than just whilst in the office.

If you are offering a sum that is significantly less than another company then the chances are that regardless of how good it is working for you, they will leave. After all, the only reason that most people work in the first place is because they need money to survive and the more money they can make the better life they can lead.

Look at what the going rate for data scientists is and offer more at the start and maintain this above average salary throughout their time at the company.

Allow Flexibility, But Promote Office Work

Often the ability to work when and where you want is key to a happy working life. It may be that the person you want to employ is based far away from the office and would rather work from home or they may work better by starting and finishing early or late. Giving them the flexibility to work in these ways will create a far more forgiving and appealing work relationship.

However, it is much easier to keep hold of an employee if they are working in-house and the way to do this is by creating an amazing working space. The argument for working at home may be that they have an hour’s commute to work every day, but if you can create an office environment that makes them want to take that extra hour just to be in that environment then you are going to be in a stronger position.

It could be done by making an office environment that is catered to them or that is designed to promote team working, creating a strong team identity and work ethic. This is about more than just what gadgets you buy for an office, but the community that you foster and the working environment that is created through personality and teamwork.

Appreciate Their Work

One of the things about being a data scientist is that the work done is often under appreciated by others within the company. The result of their work is to inform rather than necessarily react, meaning that the credit for their work can easily be given to the division who uses their insight for success.

For instance, if a data scientist finds that you should be talking to a specific demographic about a certain product, then when this is a success it will be the marketing team who sent the message who are given the credit rather than the data scientist who created the insight that allowed it to happen.

Making sure that you are always appreciative of the work being done and also making it well known throughout the rest of the company about the work that your data scientists are doing will make sure that they feel appreciated and valued by the company.

Let Them Talk To Other Companies

This one is controversial, but is something that has been proven to be successful at companies like Netflix.

Having an organization that is mature enough to let your data scientists talk to other companies is important, if not it can make them feel chained to the company. If you allow it, it also means that the meetings themselves can be used to benefit your company.

If a data scientist gets an offer from another company, they can then come back to yours and tell you what was offered, what they liked and what they were unsure about. This then gives you the opportunity to make a counteroffer, rather than them simply going to talk to another company in secret them hand in their notice when a better offer is put on the table.

Having this approach means that your company becomes more open and accepting, whilst also more rewarding and friendly for those working within it. Not needing to sneak about if somebody comes along with a job offer means that the company benefits from the open interaction and knowledge of what the other company can offer, and the employee benefits from a potential increase in job satisfaction or a new job from the other company.

Having this kind of relationship with your data scientists is going to be key to keeping them on board because as they gain more experience and the need increases for data scientists from other companies, they will get job offers. It is far better to know when they receive these rather than having to constantly ask if they are genuinely taking a sick day or if their holiday day is actually an interview day.

However, the most important thing to remember when you are trying to keep a data scientist happy is that they are not just a data scientist, they are a human being with different feelings and needs to anybody else. This means that although the above are rules of thumb about how to keep them happy, the truth is that communication and knowing what they want and need to keep them happy as an individual, will be what makes a happy and productive data scientist. 


Read next:

Working At The Boundaries Of Aesthetics And Inference