Data Science Is More than Data: Cultivate Business Knowledge and Relationships
Data scientist Deb Cooper explains how her business understanding impacts her work and why data scientists and advanced analytics practitioners should cultivate support from business sponsors.
If your career in the hard sciences isn't what you hoped it would be, you might consider switching to data science or advanced analytics. It worked for former molecular geneticist Deb Cooper.
Cooper, a data scientist who helms her own analytics consulting practice, Deborah M. Cooper Consulting, leveraged her analytical and mathematical skills, along with her expertise in statistics, to forge a successful career in advanced analytics -- back before "data science" was even a thing.
From Hard Science to Data Science
She started via a temp agency, of all places. "When I left academic research and moved to Boston, I went to a temp agency and got a job as a circulation analyst at a direct marketing company. I used that experience to move over to [a financial services company]," she says.
In the late 1990s, there wasn't any such thing as a bachelor's or master's degree in data science. There weren't data science boot camps. There wasn't "data science" as we now know it.
Cooper, like so many other early data scientists, was basically thrown into the fire: her employer assigned her to work with an internal proto-data science group. "While I was [working in financial services], I was recruited into a strategic decision support group and that was when I was able to learn SAS. The manager of that group had an operational research background, other people in the group were experts in traditional statistics and heuristics, [as well as] call center statistics," she says.
One of the focus areas of this group was multichannel marketing, specifically the problem of linking customers across different channels or contexts (e.g., print, call center, Web).
Today, multichannel marketing is a fairly well-understood problem. In the late 1990s, it might as well have been rocket science, she explains. In retrospect, many of the problems her team worked on stemmed from the limitations of the then-current technology. This is still true to some extent in data science today. It's one of the most exciting aspects of the job for Cooper and other practitioners.
Business Savviness Is the Thing
As a data scientist, one of the most important things Cooper brings to the table is her ability to understand how the arcana of technology maps to business operations. This same business expertise helps her to anticipate the people and process problems that develop as businesses attempt to optimize or transform their operations with analytics insights. "When people see my resume they know I can help them because of the way I've blended technology with business knowledge and strategy. Companies want people who understand the business and can 'speak' business," she argues.
Data science doesn't happen in a vacuum. Data science innovation can and will change how a business does business. In practice, this tends to provoke pushback from entrenched stakeholders.
It's not unusual to encounter resistance even at the outset of a business transformation project, Cooper reports. "I find that a lot of the toughest problems aren't technical problems but actually political ones. When you think about business information, the way it's captured and described represents the world view of people in a particular line of business or a particular business functional area. When you try to combine that with information from other lines of business or other functional areas, you're talking about two or more world views colliding."
"Reconciling that, it's a question of language; it's [a matter of] how you describe what's happening in the business. That can be very political, just the describing of it [is political], and then add to that the KPIs that execs are held accountable to -- or how you define a metric in general -- all of this also influences how successful you can be with that metric. The implications are far-reaching."
Market Yourself and Your Successes
Right now, data science is hot, hot, hot. The danger is that red-hot demand might lull data scientists into a kind of unsuspecting complacency. As a survivor of two recessions -- including the financial crisis of 2007-2008 -- Cooper recognizes that job security of this kind is illusory.
She believes it's important to cultivate support from business sponsors and to market your analytics successes. Most businesses aren't Silicon Valley start-ups. They're willing to buy into the idea that analytics research and development takes time and tends to result in more failures than successes. Ultimately, however, they want results. If data scientists and advanced analytics practitioners aren't yet measured on the basis of what they produce, they will be, Cooper says.
Cultivating relationships with important business sponsors helps you market what you've done and advance your career. Cooper stresses that sponsorship and mentoring are critical for career advancement -- yes, even for data scientists. What's more, if you have a sympathetic business sponsor, you'll be more effective -- you'll have an advocate who promotes your work -- and you'll have more visibility within your organization. This is why business sponsorship is crucial for success -- especially when it comes to resolving political problems, she maintains.
"If the project is sponsored and has support from the top level of the organization, then you can find solutions to political problems. If the project is sponsored laterally from where these discussions are happening, then it's a lot more difficult to get to these solutions," she says. "One organization I was with, the president told the CFO she owned all of the [financial] numbers. That cut through a lot of political issues. Ultimately, we had his [the CEO's] back in resolving those issues."
In day-to-day business decision making, consensus is nice, Cooper says, but elusive. In large organizations, executive authority is what gets things done.
"I've worked in other organizations where peer groups are just trying to come to a consensus and those projects tend to move sideways for a very long time," she says.
Final Thoughts
Cooper says data scientists and advanced analytics practitioners need to be familiar with (and able to synthesize information across) disparate business domains.
This is a particular challenge because of the way career advancement often works: "Being able to connect the dots and think broadly is really helpful, which is not a typical career path in the corporate environment. Generally you rise up within one functional area or silo [and] there are relatively few opportunities to get the breadth you really need for a broader knowledge and experience base."
Finally, an important (and oft-overlooked) skill is the ability to listen, Cooper argues. "I try to help people in various roles in the organization understand one another and translate between the different functional areas. Some of that is just being curious. Some is being a good listener."