Incentivising data scientists

One of the regular Data Science discussion breakfasts. Thanks to all who attended.

I chaired a breakfast meeting for Women in Data Science recently, and one of the topics for discussion was how to retain talent. While demand is outstripping supply and the market is going crazy, it’s enough of a minefield finding good people  in the first place.

Add to this that even after you’ve made an offer to someone, recruiters will be contacting them regularly to try to tempt them away to other roles. It’s impossible to prevent this. I’m a big believer in not playing games with recruitment – I know what I can afford and won’t get into a bidding war. If I’m paying a fair salary and they go elsewhere for money, then they are more likely to jump when a recruiter calls regardless of how well you incentivise them. This isn’t a big company or small company thing, if you want to keep hold of your team after you’ve done the very hard job of hiring them then you need to understand what motivates them and either make sure that you continue to provide those needs or plan to be hiring again in the next 12-24 months.

If people are happy, then they tend not to be looking for other opportunities. Sometimes I lose people for opportunities that are better for them, and I genuinely don’t mind this I can see when individuals should move on, and consider it my job as a leader to help people be the best they can be. Sometimes, that means supporting them going elsewhere.

So you’ve found someone who’s a great fit and they accepted and have started. They’re doing great things and you want to keep them happy. What surprised me during our discussions was that there was far more emphasis on career than all the things that should normally make staff happy. What is clear is that different people want different things to suit their lifestyle. Subsidised gym membership for one, decent holiday for another. In the UK, private healthcare is a taxable benefit and really not worth it for the amount it’s used by most. Free beer on tap? Sounds great, but there’s something social about leaving work at the office when you socialise.

Our discussion focussed on three key areas that we felt made a difference to us and whether they were true in general or just our own experiences.

– good management
– flexible working
– career progression

One of the first things we discussed was flexible working, and there was a difference of opinion here. One company had withdrawn flexible working as too many people were not coming in to the office. For me this is a sign of other problems. Where I currently work, as long as you don’t have physical meetings organised, you can decide to work from home at any time and with no more notice that a message on a relevant slack channel. What I’ve seen is that people do this when they need a quiet day to get a large amount of work done, or if they have something to be at home for (parcels, workmen etc). On the most part, people choose to come into the office. I believe that this is down to a good working environment, friendliness of colleagues, and the enjoyment of the role. If your employees stay away in droves then you need to find out what’s wrong.

Further to this, I’ve never been a clock watcher when it comes to my team. We are well beyond the factory environment and it’s laughable when I see IT companies giving people warnings for being late when there is no impact on anyone. Sure, if you have an important meeting or customer then you’d expect the contingency time to be added to avoid common problems, but who cares if someone is 10 mins late for a day in the office? This seems to be an issue with larger companies rather than start ups, but is an easy indicator of a terrible environment. I’ve known some amazing developers whose timekeeping was poor – whatever time you needed them somewhere they would be up to 15 mins late. These same people were the ones who you could rely on overnight and at weekends when things get tough. Give some flexibility and freedom and you’ll get it back multiplied.

Be the manager that you would want managing you. From my days in GirlGuiding, I learned that you had to earn leadership and trust and it seemed obvious that being respectful of ideas, even if they weren’t the final choice was a common sense approach. No team member should ever be afraid of bringing ideas.

A few weeks ago, two of my new team very hesitantly approached me to suggest a major change to some code I’d written (initially just to suit me) so that they would find it easier to use. What concerned me was that they seemed to think it was such a big deal to suggest a change to a senior individual. I’m hoping that my response (acceptance of what they needed, replanning and fast implementation so the whole team could benefit) has shown them that this is a different sort of environment.

By far the biggest incentive that we discussed was progression. Everyone wanted to feel like there was some growth in their role. Data scientists have a thirst for knowledge – this isn’t just about career progression, but new projects, skills, freedom to go to conferences and showcase their work. Have a budget for this and support them, including with time. Make sure that there is a plan, that they know what you have in store for them. As soon as day in and day out becomes the same with nothing new on the horizon, or if all the new and interesting work goes to the same people, then you’ll lose them. If they don’t think saying anything will make a difference then you won’t find out until their resignation email comes through. If you can foster an open and honest environment, your team should feel comfortable coming to you with any problems and even warn you about problems before they happen.

When the recruiters call them with what they’ll inevitably pitch as an “exciting opportunity”, you want the answer to be “No, I’m not interested” rather than “That sounds interesting, tell me more…”

Published by


Dr Janet is a Molecular Biochemistry graduate from Oxford University with a doctorate in Computational Neuroscience from Sussex. I’m currently studying for a third degree in Mathematics with Open University. During the day, and sometimes out of hours, I work as a Chief Science Officer. You can read all about that on my LinkedIn page.