Companies Need to Develop a Data Science Strategy

Unlocking the secrets contained in data is at the center of business success, but the path to progress can be daunting. Today, data comes in an array of forms—both structured and unstructured— and streams in from many sources.

The challenges run deep. There are data mining techniques to consider, statistical methods to tap, and applications and tools to use. Finally, there’s a need to bring business context to the data and ensure that an organization is using the data to solve fundamental business problems.

Enter data science. In recent years, this interdisciplinary approach has emerged as a powerful tool across a variety of industries and situations. “It is being used to drive more and more key decisions,” observes Athina Kanioura, global data science Lead at Accenture. In fact, in a best-case scenario, it can lead to transformative, if not disruptive, results.

However, like other areas of business and IT, the use of data science doesn’t guarantee results. The approach is often expensive, it can consume enormous time and resources, and it requires highly specialized skill sets that are in high demand.

How can organizations navigate this space? How can business and IT leaders put data science to work effectively? And what limitations does it present?

States Gautam Shroff, vice president and chief data scientist at Tata Consultancy Services (TCS): “Some of the biggest challenges revolve around imagining the right business problems and opportunities, while being aware of both the possibilities as well as the limitations of data science.” Amid all the hype and hoopla, he says, there’s a critical need to “balance expectations while delivering and measuring real value.”

Data Science Can Deliver Broad, Deep Insights

In recent years, data science has emerged at the center of digital business. As organizations connect systems and devices, tap social media, plug in legacy and third-party databases, harness web and online data, extract information from apps and machines, and tie in real-time supply chain and POS systems, the volume, variety and velocity of data swells.

Assembling the pieces of the puzzle can deliver broad and deep insights. It can provide answers to questions that redefine the enterprise and lead to new, better ways to navigate the business.

It’s critical to focus on both processes and technology, Shroff says. A starting point for designing a strong data science framework is to understand what an organization hopes to achieve through data and where the enterprise can make the biggest impact.

Broad social or business problems are often at the center of the equation. But data science can address a spectrum of challenges that revolve around cost efficiency, improving productivity, providing more targeted and useful answers and insights, and digitalizing activities that require computational and predictive power, such as real-time contextual marketing.

Once enterprise leaders understand how data science can make a difference—and when and where it’s needed rather than using more narrow business intelligence and analytics tools—an enterprise can formulate a plan for putting data science methodologies into motion. An initiative typically incorporates subject matter experts from fields such as statistics, machine learning, data visualization and business domains.