A colleague recently requested that I comment on his LinkedIn post about building a roadmap for getting value out of data. I said the biggest obstacle was the people problems in consulting. For those of you who are Dilbert fans, even he agrees: “I found the root cause of our problems… It’s people. They’re buggy.” (April 24, 2015)

Conflicts and ineffectiveness related to data are rooted in expecting data professionals to fix a non-data problem, most of which are “people” problems. That said, it is hardly a newsflash that “people skills” are not exactly a strength of technical folks. And the idea that problems related to their own professional domain are outside of their expertise is sobering.

What Are “People” Problems?

“People” is hard and messy in general. When it comes to data, even non-technical professionals tend to default to thinking about people in terms of technical capabilities.

The people aspect of data, analytics, and related technologies is much broader than many realize. I would even say that it is far more important than the technical chops. Provided the person has serviceable technical chops, of course.

While I have discussed people topics in data extensively, it recently occurred to me that I have never laid out all that entails in a single spot. If nothing more than for convenience, let’s organize them into three (3) categories:

• Resourcing. This is the if, the what, the when, the how, and the how many. You can think of it as defining and arranging a group of chairs that the hired individuals will eventually sit in.
• Hiring and managing data professionals. You can think of this as identifying and managing who sits in each chair.
• Organization/enterprise as a system of people. This is the people ecosystem with respect to data. It is a whole system of chairs across the enterprise: types, purposes, arrangements, and much more.

Resourcing

The first obvious-but-not-so-obvious question is whether you need a data scientist at all. People rush into hiring too often. Outside of the backfilling context, I have never in my life heard anyone say they hired a data scientist too late. Large organizations are not immune to this.

The questions to ask are vastly different depending on whether you need capability or capacity.

Defining Capability

Some questions about capability are:

• What types of data practitioners do you need to get the best return? Getting a resource that is too technical for the need is a very common error.
• What role do you need? A data role? An analytics role? An IT role? What sub-roles do you need?
• Do you need a human, or do you need a skill set? Do you understand the difference between the two needs?

Defining Capacity

Capacity is about throughput.

• Focusing on capacity before understanding the true capability needs in data is also a very common error. The former is more tangible and easier to quantify. However, that you can quantify does not mean what you are quantifying is correct.
• The resourcing requirements for post-launch needs in data and analytics are chronically grossly underestimated. It is much greater than most people realize. I have never come across anyone who has too many resources to support the post-launch operation and maintenance of data and analytics.

Capacity is much more straightforward than capability when it comes to data and analytics. Of course, it requires that you have a good grasp of the capability side of things. Throughput from capabilities you do not need is simply wasteful.

Contrary to popular belief, resources do not always need to be internal. There are plenty of scenarios that would be much better off with external resources, provided they are engaged correctly.

Hiring and Management

While hiring is only the beginning, it is where things can go massively off course. The hard skills are not only straightforward to evaluate but also less important than most people think.

Unless you are one of the handful of the world’s premier, cutting-edge companies in data, of course. Most organizations are not.

It is overwhelmingly common to hire a technical whiz when you need a big-picture mindset with enough technical skills. This also has more to do with the way the person thinks and less to do with the level of experience. Technical evaluations do not capture this effectively.

Then, there is the human behind the skills. You are hiring, managing, and developing a human, not a robot. Among the topics of interpersonal skills, the following do not get enough attention:

• Effectiveness as an analytics practitioner is not just about explaining the insights. Instead, it is influencing others to learn from data. Speaking the language of others is just one aspect.
• Few technical professionals are willing and able to go out of their way for mutual understanding. To be fair, this is a microcosm of the current world, simply magnified when it comes to technical expertise. But it is a critical characteristic of effective analytics professionals. Proactively reaching out to those who are not one of them is often not in their comfort zone.

You do tend to get away with more imperfections the bigger you are, but it is critical for smaller teams.

As for managing them, the basic principles are not particularly different, only that the resources are technical. Professional development in technical skills is straightforward, and resource management is not specific to technical resources.

Not surprisingly, a key challenge for managers of technical resources is people skills development. Unfortunately, managers who are themselves technical do not always appreciate this.

Organization/Enterprise

To quote Deming: “A bad system will beat a good person every time.” However, this idea often goes out the window at the mention of “data” and “analytics.”

The dysfunction in data and analytics is overwhelmingly attributed to the system, not the individuals in the system. If you work in an organization and have never suffered data-related between-group challenges like silos and conflicts, you have not lived.

The suppliers and consumers of analytics create an ecosystem where the rules of marketing, economics, and supply chain apply, even within an organization. This ecosystem includes the following aspects:

• Functional organizational structure. This is not just for the data and analytics teams, but all other areas they touch directly or indirectly. In practice, it is the entire enterprise. The reporting structure is much less important than the functional structure. Are your roles clearly defined around data and analytics?
• Collaboration methodology. This is alignment and systematization of expectations among stakeholders through mutually agreed-upon processes.
• Decision making, accountability, and ownership (i.e., real governance). This has to do with who makes decisions about data and analytics, rather than who does it. It is often done so poorly that the word “governance” identifies itself with a four-letter word! At one extreme is a lot of policing with no decision making; at the other is anarchy from democratization without proper governance.
• Data policies and compliance. Speaking of which! Challenges here are almost always rooted in something other than the willingness to comply.
• The cultural aspects. What people often call “data culture” is the organizational penchant for leveraging data. Data literacy can help, but rarely is it the driver.
• Data ethics. Everyone from the data practitioner to the downstream beneficiary of data and its derivatives has ethical data obligations. Even those who never analyze data.

The Fundamental Question

Do you have the right roles and the right individuals in the right numbers, the right team, and the right enterprise structure to create the right synergies? Are you sure? How do you know?

This article originally appeared on the authors website and was reprinted with permission of the author.

About the Author, Michiko Wolcott CMC®

Michiko Wolcott is a senior advisor and strategist in data and analytics and a Certified Management Consultant® (CMC®) from IMC USA with additional certification from Pragmatic Marketing. She helps clients figure out the people, process, management, and governance of data and analytics, where the vast majority go wrong before even getting to the technical stuff! She has 20 years of extensive hands-on experience in North America, Latin America, Europe and Asia, with fluency in English, Japanese, Spanish and Portuguese. Michiko received her MS in Statistics from Florida State University and multiple degrees in Music from Florida State and The Peabody Conservatory of Johns Hopkins University. Visit her website and LinkedIn page for more information.