Data Science Value and Applications

Hector Ureta

2017-04-25 08:44:01
Reading Time: 3 minutes

Companies in almost every industry are focused on leveraging data and increasing their data-science capabilities for competitive advantage. What exactly is data science and what is its (business) value? Data science is a set of fundamental principles and tools that support and guide the principled extraction of information and knowledge from data. The discipline leans on well-known data-mining techniques, however, it goes far beyond these techniques with successful data-science paradigms that provide specific applications, adding value. This post describes the “value” produced by data science within a business context and its potential “business case”, through a hands-on approach and non-scientific terms.

  1. From a high-level point of view, the very first goal of the “art” of applying data science is to increase business value by improving business decision making. This is based on data analysis, virtually at all business levels. The reality is that more and more businesses are engaged with the adoption of data-driven decision making to guide their business strategies, rather than rely purely on experience and intuition. This is applicable irregardless of the type of decision, whether it be improving company logistics, detecting email response rates, or text mining CVs within the human resources department.
  2. Decisions, metrics and conduct problem solving using a quantifiable, impartial, objective data-led approach. Until fairly recently, executives made decisions based on instinct with undeniable subjectivity. However, that’s no longer necessary with the volume of data that’s available from a breadth of resources for decision makers. Individuals that understand how to use and leverage these data, sometimes called Data Analysts, can help senior leaders in making strategic decisions based on facts, impartiality, and data.
  3. Data science can also be used as a data-to-business visualization. Many companies struggle with communicating and interpreting various results. There are sometimes ample data regarding daily operations, however it is challenging to extract key actionable insights. Data scientists can fill a critical role by helping executives interpret the data and furthermore present the results in a meaningful and visually appealing way for better understanding.
  4. Data science is a powerful tool for two key contexts. First, data science is useful for descriptive-diagnosis exercises that derive knowledge and learning from previous events. There is no longer a need to have an “analyst” reviewing past events to identify and extract conclusions as machines can perform this task automatically and efficiently. In fact, machines can even learn (this point is worth at least one separate blog post!). Second, data science is one of the most powerful forecasting and predicting tools. Surveys and the endless “expert” predictions will be continually replaced by data-driven models and predictive data-science paradigms.
  5. Considering the business, sales and marketing departments, data science can be used as a tool to better understand target audiences, including identifying the customer profile. The more information that companies gather and analyze about their customers (and more importantly, about potential customers), the more they learn about their behaviors, expectations, needs, and preferences. Products and services offered should either satisfy a need they have or uncovering a need (perhaps never imagined but of interest).
  6. Data science can also provide great knowledge about the customer’s “framework conditions” in regards to a customer’s current environment or situation. Identifying these framework conditions can be a valuable tool for potentially up-selling additional products, which in principle, were not envisaged. For instance, a couple at a honeymoon may potentially think of having children in the following years. Or, for example, someone that buys a house may then spend a significant amount of money on furniture. Conversely, someone that does not pay their student loan for one month is likely to also not pay other bills in the following months. Linking different type of information and sources can help companies identify the most appropriate products and services for key customers, and their anticipated behaviours.
Author: Hector Ureta