Analytics in Business

About the authors

Giridhar Ramachandran

Giridhar (Giri) is currently a research scholar at the Department of Management Studies at IIT, Madras with research interests in social networks and CRM. In the past, Giri has worked extensively in the analytics industry in both captive units and analytics service providers and provided analytics solutions to clients in the CPG, retail and technology industries.

Pritha Choudhuri

Pritha is a co-founder of Analytics Quotient, a marketing analytics firm that counts leading marketing companies in the world among its clients. Pritha has previously worked in the analytics industry for several years, largely in the CPG space.
--------------------------------------------------------------------------------------------------------------------

• Amazon offers you a list of items you might like based on an analysis of your past purchases and those of customers like you. 
• Your credit card company gives you a “credit score” to estimate your likelihood of making future card payments on time. 
• Retailers, led by Tesco, are increasingly using knowledge about their customers to tailor products in each store to local needs.

However, this is already common knowledge. So is this all there is to analytics? What really is analytics? Are there different types? Which companies work in this area? And what skills does it take to be successful at it?

Powered by increased availability of data, much-improved computing, and huge gains from analytics seen by the early adopters, business analytics is fast moving from being a competitive differentiator to a necessary investment for any company relying on understanding its customers to stay in business. Businesses are looking to either create analytics capability within their organizations or are actively seeking analytics talent to partner with. Large banks, IT behemoths, retailers, are all strengthening their internal analytics groups. This has however not deterred the growth of analytics specialists - companies whose special skill is to look at data and find ways of making this data useful. Also, this has fuelled the analytics surge that we see today from service providers ranging from top-end management consulting companies to large and venerable market research firms. Each of these companies today has a large and growing analytics practice.

A combination of increased credibility of the outsourcing/ offshoring model and availability of large pool of relevant talent has put India prominently on the global analytics map. As mentioned earlier, the analytics providers in India are either captive or external vendors. Amongst external vendors, there are pureplay analytics providers, large BPO and IT companies interested in the analytics pie and management consulting and market research organizations extending into this space to provide a full service offering. The multitude of organizations providing analytics services today means that a wide range of services is bundled under the umbrella term “analytics”. From developing executive reports/ dashboards to predictive analytics to optimization, analytics has become a catch-all for anything to do with rows and columns of data.

                                                          Technique Sophistication

The analytics landscape

What are some of the typical problems that an analytics company might try to solve?

These might range from credit scoring to grouping consumers to generate better insights, to answering specific questions about consumers or products. As analytics practitioners, we often get asked simple business questions by our clients. 

Here are some examples -

• What is the relative importance of shopper attributes in the shopper decision-making process? 
• Is there a range of price my product should stay within to maintain market share? 
• How does this change for each sku? 
• How do competitor prices impact this? 
• What are the right metrics to assess marketing performance? 
• How can a firm realign marketing dollars to ensure optimum returns? 
• What will the customer buy next and when? 
• In which segments does the firm’s product have a right to win with customers? 

How are these questions answered?

While numerical skills and a knowledge of statistical tools is fundamental to answering these questions, an empathetic understanding of the client’s business and the context they operate in, coupled with a realistic assessment of the data available to answer these questions, really make the difference between effective analytics solutions and run-of-the-mill models.

However, is analytics then, a cure-all?

Can digging into data, and applying the insights gleaned from this process, translate straightaway into superior business results? Although we would love to say yes, the truth is that analytics, just like any other business tool, has its limitations. Common misconceptions about analytics, if not addressed, only lead to very ineffective analytic solutions.

Here is a list of some common misconceptions –

Analytics is the ‘cure-all’

Businesses are fraught with examples were analytics is blamed for poor results. A good analytics solution is only a diagnostic. It is not the remedy. The remedy lies in powerful execution. It is important for analytic solutions to be rooted in business impact. However, only a powerful execution can deliver the promised impact.

Analytics practitioners need not know the business

Data and Statistics are Industry agnostic. However, it takes a lot of domain expertise for the solutions to reflect business reality. If we do not get this right, the solutions will not be the powerful drivers of decision making that they should be.

One size fits all

There can be broad similarities across business problems and industries. However each problem has its own unique challenges and data always has a way to humble the practitioner. The authors have seen instances of the same analytical model getting used across diverse groups. And more often than not, the blame of poor results is attributed to the data! A black boxed solution has as much utility as a pharmacist dispensing medicine.

Analytics always uncovers hidden insights

It is very likely this is true. However more often than not, analytics validates what users might already know. Analytics helps connect the dots between events or attitudes and behaviour, thereby providing clarity. So, it is unfair to expect something new and hitherto unknown facts about businesses every time an analytic model is built.

Analytic models can be built (and read) by anyone

It is true that the advent of fantastic analytic software has removed the requirement of ‘experts’. However it has not removed the requirement of rigor in application. The authors have seen many instances where statistical techniques misused and analysis performed on data for which the data is not amenable to. All these have an impact of the reliability and robustness of results. Additionally, not having sufficient understanding of the method also results in incorrect interpretations. The power of analytics can be seen well only when the model building and interpretation are backed by sufficient relevant knowledge.

Notwithstanding all of these issues, the world of storytelling with numbers is a fascinating one and we have a multitude of examples to share of both interesting problems and incisive insights from our journey as practitioners. The analytics space is certainly growing rapidly, and will continue to evolve as our understanding on how to harness data and use it effectively increases. Whichever way this evolution progresses, the space is likely to continue to be at the cutting-edge of new business practices, and will therefore keep drawing top-end talent as we go forward.