This guest post courtesy of The Data Incubator.
Many companies understand that hiring a Data Scientist can help drive their business forward, but knowing how to manage and best utilise their skills can be challenging.
Particularly for founders who have no experience in data or analytics, understanding what Data Scientists do, how long things take, and why they’re important can seem a bit of a mystery. Thus many struggle to realize the full organizational and financial benefits from investing in a Data Scientist or a data science team.
Immerse yourself in their world
In data science it is difficult to precisely plan a project's timing or final outcomes, which makes it hard for founders to know what’s going on, and when they can rely on receiving information from their data. Data science can seem quite messy, and the uncertainty can strain relationships between data experts and the rest of the team.
That is why we recommend immersing yourself in the world of your Data Scientists - actually get your hands dirty, helping on a project. That will help you understand things like why data wrangling can take the time it does.
Take the time to understand
When starting a new project, there are typically three factors to balance out: Time, Cost, and Quality (TCQ). It’s virtually impossible to maximise all three; if you want to do the project quickly and to a high standard, it is likely to cost more; if you want to do the project cheaply but at high quality, it will take longer. If you’re prepared to spend a decent amount of time and resources, it will be of exceptional quality.
Because founders can struggle to understand the data science world, they often put time and cost saving methods in place that considerably affect the quality of the results from the data, without realising it.
It’s up to the founder to be transparent with the Data Scientist about what is needed, and then take the time to really grasp what the Data Scientist can do to achieve that result, so you find the right TCQ balance for your desired outcomes.
The difference between doing this and rushing ahead is the difference between making well-informed decisions, and decisions that have blind spots - as they say, a wrong answer is worse than no answer at all.
Give your data scientist a purpose
Make sure your Data Scientist understands how their work fits into your company as a whole. Knowing the broader context will help them better understand the problems they are facing, and will strengthen their analyses.
We all know that connecting employees to a mission they care about is key to having them produce their best work. Show them the impact of their analyses of decision making, and how they are a valued member of the team. It will lead them to make better products and deliver better services to your customers.
Remember that a simple graph may be the result of days of analysis drawn from petabytes of data. What is presented to you is often just the tip of the iceberg in terms of work and it is appropriate to recognize all the work that went into an analysis.
Keep them stimulated
Data Scientists love a challenge, so encourage them to work on compelling new data sets. Find intellectually stimulating questions for them to work on, or ask them to come up with their own - they may surprise you with how they can help you and your business.
Make sure the tools you have for your Data Scientist are appropriate for the job, in both type and speed. You don’t want your Data Scientist writing de novo.
The data science community is large and open, and a cost effective way to upskill your Data Scientist is provide catering and conference room space for them to hold a “lunch and learns” with fellow Data Scientists from other companies. This way they can learn new skills to help better leverage data to the company’s benefit.
It’s important to recognize that data science and engineering go hand in hand. Make sure you’re providing enough engineering support for your Data Scientist to not just do their jobs but thrive at your workplace. If the servers are slow or capacity constrained, it will be difficult for them to iterate quickly, and the result will be a loss of motivation and creativity.
Implementing these points will show your Data Scientist that your company really does care about them and the work they do. And the more engaged they are, the happier they’ll be to stick with you and provide quality work.