Being an expert is a tough job. As a practicing consultant for the last couple of decades, I’m very close to the challenges internal and external consultants have in any organization.

You’re often called upon for your expert advice, though you do not have direct responsibility or control over the outcome; as such, adoption of your advice is largely dependent on how well you can convince the benefactors of your wisdom to follow your recommendations. It’s a great disservice to all involved when sound advice is dismissed purely because the arguments weren’t strong enough to sway the intended audience. This is where data science can greatly help.

SEE: Job description: Specialized technical consultant (Tech Pro Research)

I’m working with a large oil and gas client that’s facing a similar situation. The client has internal experts who are responsible for counseling business unit practitioners (e.g., facilities engineers) on difficult situations that require an advanced level of reservoir engineering. Often, their expert advice is considered, and then ignored in favor of a different idea from an entry-level petroleum engineer. As you can imagine, this is very frustrating.

My advice to the experts is to bolster their data science capability. Data science is a secret weapon for experts who want to exert more influence over the organization.

The power of information

Informational power is an excellent complement to expert power. Experts in the organization who are primarily responsible for offering advice are relying on what social psychologists John French and Bertram Raven called expert power.

In some cases, expert power can successfully influence others to take an expert’s advice or approach, but in many cases, it’s not enough; this is especially true when the receiver of expert advice has some knowledge of the subject area, and then it becomes a contest of opinions, and the expert needs some other base of power to sway the argument to their side. This is where informational power and data science comes in handy.

The expert who understands data science can use the company’s data to produce information that others in the organization cannot, thereby strengthening not only their fundamental position, but also their power base from knowing how to produce information that is otherwise unobtainable. Any argument is stronger when there’s credible data to support it; however, what’s more impressive is when a position is supported by information, knowledge, and wisdom that nobody else has–this is how the expert brings informational power to the table. But they cannot do it alone.

The expert-data scientist relationship

It’s important for the experts to build a relationship with the data science team. What I’ve seen in some organizations is the desire for experts to bring data science capabilities into their function, such that it’s self-contained and self-supporting. This is a very bad idea.

Data science is extremely difficult–it takes a solid background in business intelligence, data warehousing, data visualization, advanced mathematics, artificial intelligence, and computer programming. It’s naive to think that these disciplines can somehow be absorbed into a function that is primarily focused on developing knowledge in a completely different subject area.

SEE: Turning big data into business insights (PDF download) (ZDNet/TechRepublic special report)

It’s better for experts to partner with data scientists who already know what they’re doing. The collective goal of this partnership is to curate the company’s data into information, knowledge, and wisdom that expands and supports the experts’ base of knowledge. For instance, hypotheses can be tested and vetted using the company’s operational data and then developed into an expert system. This becomes a private repository of information that can only be accessed by the experts.

Pitfalls to avoid

Informational power is a great addition to expert power, but there are some pitfalls to avoid.

The first pitfall is a very important one: Don’t practice data science without the help of data scientists. The time spent struggling with the nuances of the database or the analytic engine is not time well spent. It’s easy to spend days or weeks troubleshooting an inefficient query or a complicated database join when that work is better suited for the data professionals.

Also, you must ensure your information doesn’t lose its power by making it so obscure that nobody outside the expert group understands what it means. Informational power only works when the information is perceived as valuable. To avoid this pitfall, make sure your expert system has a layer that translates its results into a language your intended audience understands.

SEE: Data science: Feeding the all-seeing beast (ZDNet)

Finally, you should avoid being competitive and confrontational when using data science, because that defeats the purpose. The end goal is for the intended practitioners to adopt your advice–the goal is not to prove you’re right. This is accomplished by showing others information they’re not aware of in the hopes that they understand the situation in a different way. Steer away from information that clearly shows their ideas are wrong because that is the quickest way to be dismissed, regardless of the merits of your findings.


Experts who use data science bring an elevated level of influence into the organization. Great ideas don’t matter unless they’re adopted, and data science is an expert’s best friend when it comes to swaying hearts and minds.

As an internal expert or consultant, it’s important to build a relationship with your data scientists. Partner with them to create a system that fortifies your expertise with information that cannot be obtained anywhere else. Then, make sure you use this information wisely to educate the rest of the organization and influence them down the right path. After all, that’s what they count on you to do, right?