How to recruit (and keep) a high-performing data team
6 MINUTE READ
In today’s market, having a strong, high-performing data team is an important driver of growth and success. But why do so many teams falter when it comes to extracting real value from data?
This is a pressing question facing tech leaders across the globe. Hundreds of companies have invested big time in their data team, only to see the department fail to deliver a return after years of presence. An incredibly frustrating scenario to be in.
Often the problem can be traced back to the mindset and attitude of those on the team. Speaking on our podcast, Bruce Pannaman, Senior Data Scientist at busuu, outlined the hard-to-teach traits that separate truly great data professionals from average ones.
“Having a strong academic background and theoretical knowledge isn’t good enough. What you need is drive. Someone who is on board with the vision. Someone who wants to get out there, talk to people, work out what their problems are, and work forward from there. What you don’t want are individuals who demand a spec, churn out the project, and then feel it’s someone else’s problem.
Hiring people with real ambition and curiosity is key. Whenever I meet with people who have those qualities, really cool conversations happen. The technical skills themselves are very accessible and achievable. The thing separating average from great is a problem solving, can-do attitude,” he said.
In this post we’ll cover some best practices data leaders and HR managers can follow to build and keep a data team that delivers results.
Keep a high bar on recruitment
This point can’t be stressed enough. If you get recruitment right, everything from that point forward becomes easy to implement. Get it wrong and you’ll face an uphill struggle every step of the way. In fact, if you fail at this, you’ll never be able to build a high-performance data team no matter how much you excel in other areas.
So how do you make sure you are recruiting the right profile? It’s easy to tell who has the required technical ability. But applying technical know-how in a smart way requires traits that are harder to identify. Expanding on Bruce’s point, you should assess candidates on the following factors.
In many ways, curiosity – that insatiable hunger for knowledge and understanding – is the most prominent trait of a truly great data professional. At a fundamental level their job is to ask questions of data and of people. Most people within an organisation either don’t know or don’t care about the limitations of data. So everyone on your data team should be curious about what other departments are doing and what they want to achieve.
Clarity is essentially an outgrowth of curiosity. Whether your team are writing code, performing and analysis, or just cleaning up messy data, they should understand what they are doing and why they are doing it. Individuals who work with a high degree of clarity constantly ask two questions of themselves: “Why?” and “So what?”. Your data team needs to understand what their actions mean within the context of the specific project they are working on and the larger context of the business as a whole.
The common view of creativity it to see it as binary: either people have it or they don’t. But contrary to popular belief, creativity can be learned and developed in the same way an athlete develops skills for their sport. This can be done through methods such as brainstorming, stream-of-consciousness journaling, and reading articles well outside one’s field of expertise/interest. When hiring, determine whether candidates are engaged in any of these creativity-boosting pursuits. Furthermore, a good interview question to ask is “Tell me about a time you solved a familiar problem in a new way”.
Healthy skepticism keeps creativity in check. It’s important to stay grounded in the real world rather than go down a rabbit hole of weird and wonderful, but completely unrealistic ventures. Maintaining an air of skepticism alongside the optimistic curiosity that drives learning is a balancing act. Those that have mastered it understand that while exploring data is great, it’s only as good as the methods that collected it. You need to hire people who are open to cross-referencing their own assumptions with the assumptions that were used to construct the model. In the words of statistician George E. P. Box: ‘All models are wrong, but some are useful.’
Top performers in all disciplines know that they don’t know everything and are always looking for new learning opportunities. This is what drives their success. In data sometimes it’s necessary to swallow your pride and go back to the drawing board when things aren’t working. A reluctance to do so is one of the reasons many data teams are struggling to meet expectations.
Remember, you can teach technical skills. You can manage performance. What you can’t do is force the five qualities listed above.
Companies with high-performing data teams typically implement case studies, roles plays, and interactions during their hiring process to judge candidates on these traits. While they might sound easy to bring in, it’s not always the case. It means saying no to candidates who may be very strong technically. It also means letting go of any existing team members not clearing the thresholds even if it means productivity will dip temporarily. This is a tough management call to make but it has the potential to pay off in a big way.
Managing with a view to retention
Once you’ve got employees with the right attitude in the door, you need to work to hold on to them. Indeed, high-performing data teams are characterised by high levels of output and very low levels of staff turnover.
Writing for MIT Sloan Management Review, Roger Stein, Senior Lecturer in Finance at MIT, said there are three main jobs leaders need to take on to manage a first-rate data team: 1) Build an engaging environment; 2) Make sure the team has access to necessary resources; 3) Get your hands dirty – but stay out of the way.
Build an engaging environment
Don’t ignore the general outlook of the team and the dynamics of daily interactions. Team members should feel comfortable giving colleagues their best ideas and collaborating to turn them into real world solutions. World-class data teams operate within a culture where team members are excited by what their teammates can do. But in order to be truly effective, they need continuity.
“The longer team members work with each other, the more they get to know the ways in which different colleagues approach problems and communicate results, and the more they learn to trust colleagues in key areas. While it is possible to hire new analysts with excellent skills, there is no quick way to infuse new members with this kind of team “meta knowledge.” It is far cheaper to invest in a stimulating and high-performing environment than it is to replace a seasoned analytics professional,” said Roger.
Make sure the team has access to necessary resources
All departments need resources, and generally there aren’t enough to go around. But alongside obvious hardware and software resources, data teams need additional support. At times they need to sit down face-to-face with senior people within the organisation. In other instances they may need high-performance computing platforms or specialised data sets for modelling projects.
“It can be hard for non-technical managers to evaluate the importance of, say, access to an expensive database or a subscription to an online research library. For this reason, the data team manager must be prepared to advocate for the team’s needs with the rest of the organisation’s business leaders,” said Roger.
Get your hands dirty – but stay out of the way
The best data managers need to be able to give their staff freedom and trust to take control of the agenda for problem solving. Of course, managers should still review proposals before giving them the go-ahead but it should be the team taking the lead and initiative. They need the freedom to implement solutions they believe will work best.
However, Roger highlights this does not mean managers should only get involved in oversight. Hands-off managers have difficulty in maintaining trust and morale. “Ongoing experience in doing the work provides empirical evidence to both the team and, importantly, to the market, that a manager is both credible and able to lead. This means that in addition to vision and communication skills, the most successful data leaders are the ones that are also very good statisticians and programmers,” he said.
Looking to expand your data team for your next project? We can help. Get in touch with our Customer Success Team today: firstname.lastname@example.org 0203 137 7005 / 07946 191 397