Amazon started as an online bookstore, and now, it has revolutionized the retail ecosystem completely. Netflix started as a DVD rental store, and now, it is giving giants in the media industry a run for their money. When was the last time you witnessed a new neighbourhood bookstore or movies-for-rent store run so successfully? What makes Amazon continuously evolve, adapt, and grow? Do you think it is the fact that it is a digital platform? Well, a local bookstore can have a website too. Is it the first-mover advantage then? Not entirely, because we were still consuming media content online before the likes of Netflix came into the picture.

What then has been the most predominant factor in the Darwinian upshot of these platforms in the past decade? The answer clearly, is Data. It is the ever-evolving use of data to learn and create differentiating capabilities before everyone else, that has given these platforms their edge. Digital, in-fact, is just an idea. The revolutionary change lies in the continuous and selfsustaining use of analytics to generate new revenue streams, gain insights into consumers, impart the ability to scale, and among other things, make supply chains more efficient.

Measurement and Attribution Can't Measure, Can't Manage.

While this aspect may seem self-explanatory and primarily significant in theory, its importance is often diluted in practice. Resultantly, when the impact of analytics is not accurately measured, further investments in analytics seem unjustified. An effective performance measurement framework that covers all levels of the organization-tying matrices directly to tangible business value, and indirectly to frontline adoption and other operational goals, is vital to benefit from analytics practically. These assessment scales can be of varying types, including:

  • Operational Measures including a reduction in turnaround time and improvement in productivity
  • Analytic Measures including accuracy, robustness, coverage, and predictive power of models
  • Process Feedback including ease of implementation, scalability, user-friendliness, customer satisfaction, and competitive edge
  • Financial Measures including cost management, revenue growth, profit margins, and return on assets

Root-Cause Analysis The Blame Game.

With the advent of machine and deep learning, analytics has become more of a black-box for business managers. Hence, the tradeoff between accuracy and interpretability has become very important. While for some use-cases, e.g., credit-risk modeling and autonomous driving, accuracy is of utmost importance, but whenever decisions are linked with underlying operational processes, e.g., inventory planning or sales forecasting, it is important to understand the key factors or influencers driving the predictions. It not only helps business managers in understanding such models and validating business hypotheses but also goes a long way in establishing the credibility of analytics as a key tool in decision making going forward.

Unfortunately, most of the analytic focus lies in "what" did/ does/will happen? Often, it's convenient to mistake correlation for causation and to find misleading patterns in the data. For analytics to be able to add value to a business, 'why' is most important. Business stakeholders must be in constant touch with analytic teams to understand key underlying factors and identify the elements that can be changed, and to what extent, along with the outcomes of such changes.

For example, while a Machine Learning model might attribute low car sales to unusually low footfalls or low inquiries in car showrooms, there is a need to dig deeper and understand factors causing this change in the market and consumer patterns.

Continuous Improvement The Virtuous Cycle

While it is crucial to get the wheels rolling by deploying solutions on-hand, it is equally important to fine-tune existing analytics solutions continuously. RoI of an implemented solution could degrade with time. Hence it is imperative to assess RoI and adapt existing analytics models continually and BI solutions to adjust for changes like reduced predictive power and shift in underlying variables. While scaling up analytics prototypes, analytics models need to be personalized or tweaked to account for variations across regions, markets, product portfolio and consumer segments. There is also a need for adopting new analytics initiatives in resonance with the changing business environment and goals.

Analytics needs to be conceived as a continuous journey of value discovery through the exploration of new business problems, analytical tools, and frameworks. Each analytics initiative must be ranked objectively, and its impact should be directly evaluated basis its potential value to the business. The effect of such an action can also be judged through derivate parameters like the magnitude of leadership interests, operational feasibility, and alignment with business strategy. For best results, organizations need to continually gauge what is working through fastpaced experimentation and optimize their analytics approach accordingly.

Setting up analytics is only half the battle won. Implementing analytics in any organization is an incremental journey comprising more of a mindset and strategic change rather than a one-time technical or team set-up. Analytics needs to be ingrained in organization DNA through an increase in adoption and deployment across strategic decisions and operational processes. Bringing in an external partner who can handhold you through this paradigm shift can accelerate the transformation process, and help an organization harness the revolutionary power of data, more significantly than ever before.

Originally published July 7, 2020.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.