Blogs

Why you may never need to become a data scientist

Post Comment
Big Data Evangelist, IBM

Let’s not get carried away by the vogue for all things data science. In that regard, I recently saw a headline that screamed “We’re all data scientists now!” And I thought “How wrong could that possibly be?!”

Mind you, we’re all enjoying the fruits of data science in our business and personal lives. But asserting that everybody is now some sort of data scientist is like claiming that the ability to add two numbers together qualifies you as a mathematician.

That’s why I’m as jaded as the next industry observer regarding the rampant job-title inflation that covets the magic phrase “data scientist.” But please don’t interpret this as an elitist railing at perpetrators of “fake data science.” There are plenty such points of view floating around in cyberspace, such as this, but I don’t agree with them, as I stated here.

My perspective, as stated in this blog two years ago, is that the democratization of data science is the best thing that has ever happened to this discipline. It has greatly increased the flow of fresh ideas, enthusiasm and innovation into the big data industry.

However, you personally may never need to upgrade your skills to the level of a professional data scientist. Rather, what you truly need is the ability to tap into continuous, proactive and authoritative insights in all aspects of your life. What any of us truly needs is access to cloud services where data-driven analyses generate these insights and deliver them reliably into every application, be it business or personal in nature. And if your job involves being on the front line to the customer—in other words, if you happen to work in a call center, support function, sales office or the like—then you need the full power of customer analytics, big data, cognitive computing and real-time, next best action inside all of your transactional applications.

The data scientists are the smart people who, behind the scenes, make sure that all of the back-end applications are kept fed with the best data and the sharpest algorithms at all times to help you deliver seamless brilliance in all channel engagements. You yourself, on the front lines, may be a data scientist of sorts if you’re, say, a business-domain analyst in support of marketing and other customer-facing efforts. But it’s highly likely, if you’re serving in a channel function, that your domain expertise is just as important as your data-science skills. It’s also likely that you’re using tools and applications that have been developed and provisioned for you by more specialized data scientists who have competencies in sophisticated areas such as artificial neural networks that are beyond your current skillset.

For enterprise channels, advanced cognition is increasingly in the cloud, and much of it pre-defined searches and analytics powered by IBM Watson solutions such as Engagement Advisor. For channel business analysts who want to go beyond that and do deep cognitive explorations in the cloud, IBM’s recent launch of Watson Explorer is noteworthy. The news release includes this interesting channel-relevant customer testimonial on the power of on-demand cognitive exploration in customer engagement: “’We've put Watson Explorer in the hands of our call center agents to equip them with a 360-degree view of the information they need, including self-performance reports and other analytics,' said Farouk Ferchichi, corporate manager of Toyota Financial Services. 'With over four million customers, we recognize the importance of top-notch service and are committed to actively maintaining customer satisfaction. Watson Explorer gives agents the ability to get detailed metrics on their performance to identify strengths and focus areas to improve.’”

In the act of using tools such as Watson Explorer, are these agents “data scientists”? Is the question even meaningful anymore?

I’m not sure that it is. The practical distinctions data scientists, statistically powered business analysts, and cognitively assisted operational personnel are blurring to the point of irrelevance. We all consume the fruits of data science. And more of the “rest of us,” through crowdsourcing, contribute to the enrichment of the data and algorithms upon which high-quality data science depends.

Does that level of indirect participation in the data-scientific industrial complex qualify you to call yourself a “data scientist”? That’s your call and your career, not mine.