Adding Data Science to Supply Chain’s List of Job Specs

Data Scientists

How does a data scientist relate to the supply chain discipline? Also, is supply chain management an attractive career option for these professionals, or a novel stepping stone to a future in a more lucrative sphere such as investment banking?

These are among the questions explored by attendees at the Data Management: Preparing for Machine Learning roundtable on October 15 to 16, 2019, at the MIT Center for Transportation & Logistics.

The event brought together supply chain and data management practitioners from a broad cross-section of enterprises to talk about the challenges of managing data in a fast-changing competitive environment.

Tell the right story

Data scientists derive insights from huge volumes of structured and unstructured data, to help organizations meet their competitive goals. They organize and analyze the data and convey the findings to business leaders.

The enterprises at the roundtable employ these specialists in different ways. For example, some data scientists are part of in-house analytics centers of excellence that various functions, including supply chain, draw on to help them meet their analytical needs. In other organizations at the roundtable, data scientists work more closely with supply chain teams.

Data scientists are in demand in most industries — how can supply chain attract these sought-after professionals?

Attendee:

“It’s the responsibility of the leadership to tell the story that supply chain is exciting, and there is a real-world payback.”

Telling such a story has become significantly easier over recent years as supply chain has shed its image as a relatively low-level operational function confined to the tactical weeds.

Keep them interested

Having inspired and recruited a candidate, how does the organization keep him or her in such a competitive market for talent?

One supply chain leader at the roundtable is resigned to enduring a high turnover rate in this department. The company recruits and trains data scientists and assumes that they will leave in two years or so to explore greener pastures.

Other attendees were more sanguine about the retention challenge. Senior manager advised:

“You have to work hard to find intellectually stimulating problems that keep them motivated.”

Such a task is well within the remit of the supply chain function. For example, a company explained how its data scientists had the opportunity to work on an $8 billion e-commerce project. Another plus is that supply chains are fast-moving, and individuals are unlikely to get locked into endless projects that inch forward.

In addition to being engaging and challenging, projects should be worthwhile. Several attendees emphasized how young data scientists are motivated by work that delivers real value, as opposed to work that doesn’t have a clear ROI or appears to have little impact in the real world. Again, this aligns with the hands-on responsibilities of the supply chain function.

A clear pathway to career advancement also helps. This can be a tricky for challenge for organizations — notably small companies — that are inexperienced at employing these professionals, but career ladders are emerging.

For example, a roundtable participant has created senior roles for data scientists, such as the Data Scientist Principal role that is equivalent to Vice President. Setting the expectation at the recruitment stage that individuals are expected to climb the career ladder reinforces the notion that there are ample opportunities for advancement.

Several participants at the roundtable underscored the importance of corporate culture in attracting and retaining data science talent. Organizations that require new employees to navigate or even change deeply entrenched, rigid processes and managerial hierarchies — especially those that are resistant to new ideas — may struggle to keep data scientist hires.

It’s also essential to have the right tools in place before initiating the hiring process. In addition to causing frustration, not having the tools to do the job wastes the time of these highly paid specialists.

Roundtable team leader:

“We have our engineering resource work with data scientists to give them the right tools.”

Skills sets

Of course, hiring is a two-way street, so what should data scientists bring to the table aside from high-level technical skills?

Communication skills are important. Ideally, a data scientist should be able to explain complex algorithms in terms that mere mortals outside of their area of specialty can understand.

One company hired a young data scientist partly because he was comfortable fulfilling a consultancy-type role in the organization.

A team leader suggested that data scientists should be able to explain the business value of a project. Curiosity and an inclination to ask questions are important attributes too, although the same observation applies in most functional areas.

One participant looks for an element of “humility” in data scientists. More specifically, a willingness to learn, “and will they get their hands dirty and cleanse data?”

The task of cleaning data was discussed extensively at the roundtable. Although artificial intelligence and machine learning are now embedded in the supply chain function, applications are still relatively new. Generating the data that machine learning models need is also at a relatively early stage of development, which means that masses of existing data has to be reformatted and cleaned.

It can be an arduous and time-consuming task, and deciding who gets the short straw to do the work is not straightforward. One roundtable company nips the problem in the bud by making it clear at the job interview stage that data scientists are expected to clean the data they need to work their magic.

Adapting to a new landscape

Just as digital transformation is an unfamiliar journey for supply chain professionals, so the supply chain is unfamiliar territory for data scientists. This will change as the journey progresses. Meanwhile, supply chain leaders must find ways to integrate data science into their functional map. The task will require some lateral thinking. For example, one roundtable participant suggested that when it comes to recruiting data science talent, “retention starts before you hire.”