Who you gonna call if your data science is lost in translation?

There was once a business executive who kept coming home troubled and distracted. His partner asked him one evening, “What’s the matter? Something’s worrying you.”. “There’s so much data”, he replied, “and I can’t understand it or get my head around it.” “You could try hiring a data scientist”, suggested his partner. So he did just that. Six months later he was still coming home troubled, and when his partner inquired again what was up, he replied: “There’s so much data, and I can’t understand what my data scientist is telling me about it”.

Make no mistake, this is a growing problem. As data science becomes more of a staple in organizations and businesses, and as more and more data scientists have been hired in recent years, the perception of a gap in communication between data scientists and operational business leaders is growing. This gap is a serious one. It prevents the data scientist from having impact with their work, and it prevents the business leader from realizing the value of their data scientist. This virtuous cycle can lead to frustration and demotivation on the part of both, as well as cynicism about whether data science is all it’s cracked up to be.

The solution to this problem is the analytics translator — a breed of individual who understands the business and its practical decision-making needs, but also has an interest in analytics, strong communication skills and knows enough about what a data scientist does — and how they do it — to be dangerous. This is not just me drinking the latest Kool-Aid! I’m speaking from experience — I work closely with analytics translators all the time, and I don’t know what I’d do without them. I confidently predict that this role is the next big growth role in the field of analytics.

Why are translators needed?

First, if you are a data scientist, you should not take this as a criticism of your skills. They are rare and important, but the risk is that they are not having impact and so are not as valued as they should be. Here are some common reasons why I think data scientists are not fulfilling their potential impact in businesses and organizations. I call these challenges the three Is:

1. Data scientists often are new to the environment they work in and don’t have the instinct for the business problem they are being asked to solve — an instinct that can only come from being embedded in the business for a long period of time. This means that they can’t fully understand how their output might be useful to their stakeholders, and they often have nobody to help them think about this. This can lead to outputs that are not practically understandable or implementable to stakeholders. 
2. Data scientists often don’t have a lot of interest in spending large amounts of time thinking about business implementation. Their passion is methodology and technology, and they get their kick out of getting things working and generating results. If they had to devote half their time to developing Powerpoint decks and attending meetings, many of them would be pretty disappointed in their job, and their progress would be a lot slower.
3. Data scientists are often not provided sufficient insulation from the day to day demands of the business to have time to practice their skills effectively. Without appropriate protection, they can quickly become overwhelmed with tasks or questions, many of which could be addressed by a capable reporting analyst. Often new to the business, with limited accrued personal capital, and with no support in prioritizing their work, they just spend their days trying to stem the flow of random work from people who don’t really understand what they do.

What makes a great translator?

Addressing the three Is tells us a lot about the characteristics of a great analytics translator:

1. They should be sourced from the business — or from a very similar business — so that they come with instinct honed from experience. Much of the conversation between data scientists and their translator partners involve discussing options for methodology and output, with the data scientist providing information on what’s possible, and the translator providing judgment on what’s useful.
2. They should enjoy the cut and thrust and the politics of getting things done in a business, and the communication and influencing processes that that entails. They’ve likely previously been in a consulting or operations role. But they should have an interest and passion for data and how it can drive better decision-making, and a clear enthusiasm for methodology and science. 
3. They should have a strong ability to formulate priorities and agree them with stakeholders, and be proactive in ensuring insulation against other distractions. They should be able to quickly determine the unique skills of their data scientist partners and help defend against work demands that are not utilizing those skills.

So, whether you’re an executive who doesn’t speak data science or a data scientist that doesn’t speak executive, plugging the three I’s can only be achieved by bringing the right analytics translator on board. You might want to get in on the act now before they are all snapped up!

I'm a Data Science, Digital and Human Capital leader at McKinsey. Originally I was a Pure Mathematician, then I became a Psychometrician. I am passionate about applying the rigor of both those disciplines to complex people and talent questions. I'm also a coding geek and a massive fan of Japanese RPGs. You can also find me on Twitter.

All opinions expressed are my own and not to be associated with my employer or any other organization I am connected with.