Banner Image

Corporate Themes

Explanatory Models: Extending the Value of Your Predictive Analytics Investment

 
February 14, 2018

Companies that spend on marketing analytics also tend to have higher marketing budgets, according to a survey by Wall Street Journal and CMO Today. The key question though is: are companies able to drive adequate value from their marketing analytics spends? The limitations of using only predictive analytics are becoming obvious – it can only tell you what might happen in certain scenarios. It is incapable of helping you understand why certain outcomes are likely, what correlations drive those outcomes, or how you can prevent adverse outcomes. Using explanatory analytical models along with predictive analytics can help you gain deeper insights by helping you understand the “why” of a predicted outcome, to better serve your customers.

The value of explanatory data analytics in driving growth

The insights of “why” your marketing is driving certain types of results and the correlation between these insights can help you identify solutions to change course, in case the predicted outcomes are undesirable. Using such insights, you can also bridge information silos and enhance the value of marketing spend for your organization.

By using data visualization to present data in various scenarios, explanatory data analytics helps you understand the data better. With predictive analysis, you follow a linear path that assumes only one outcome. Explanatory data analytics, on the other hand, helps you visualize different outcomes based on changes in the scenario. It uses variables that are directly or indirectly under your control to help you zero in on new opportunities or drive superior results on your marketing campaigns. Let’s consider two use cases for explanatory data analytics.

Use case 1: Increasing lunch-time sales for a restaurant chain

Consider a steakhouse chain that wants to increase sales. Let’s say they have detailed sales data on what they are selling the most, in what quantities, and at what time. Using analytics, they come up with the hypothesis that the best way to increase afternoon sales is to target couples during lunch time.  So they tweak the menu options and run a campaign to target couples, but sales do not increase as expected. How can an explanatory data analytics model help in this situation?  Using exploratory analytics on the right data, the company can find the “why?” behind the lack of increase in sales – i.e. most women do not like the presentation of the steaks and the machismo messaging. So, the steakhouse comes to the conclusion that to improve afternoon sales by targeting couples, they need better presentation and messaging that is more gender inclusive.

Use case 2: Targeting millennials to drive growth for a motorbike company

The number of motorcyclists in the US is much smaller than in other parts of the world like Asia. Growing the market in the US can therefore be a challenging proposition for motorbike companies. Let’s say an American motorbike company wants to sell more motorcycles and has found that the millennial target market offers the maximum potential for growth. Assembly line changes take time and are expensive, so the company wants to understand exactly what features are most important to millennial motorcyclists – before they set about building one.

With explanatory analytics, the motorcycle company can pinpoint why millennials are likely to take to motorcycling. For instance, millennials are highly concerned about the environment.  So, the company has to develop the value proposition around fuel economy and emissions as compared to four wheelers. Also, let’s say the company finds that millennials value experiences over possessions, making it essential to market the experience of traveling around the country on a bike – either solo or as a couple. Without explanatory analytics, the company might sell the motorbikes as an advanced toy for adrenaline-driven drives. Explanatory analytics helps the company avoid the ‘advanced toy’ messaging trap. With explanatory analytics, assembly line engineers can be assured that they are choosing the right features for the right reasons. In essence, explanatory analytics lets the company’s marketers become its growth drivers.

The time of explanatory analytics is now

Investment in data analytics isn’t as simple as investing in a tool and deploying it on your data sets. To get the most out of your investment, it’s important to change the way you explore data – in sync with evolving consumer interests and buying habits. With customers routinely demanding personalized experiences, increasing the use of explanatory data analytics can help you extract greater value from your existing predictive analytics investments.

Are you already using or plan to use explanatory data analytics? Do respond in the comments below.

Avinash Pattabhiram is a Marketing Manager with the Business Operations unit at Tata Consultancy Services (TCS). He has seven years of experience in the areas of research, marketing, and consulting. Prior to the current role, Pattabhiram worked as an Energy Analyst, tracking energy trends and alternative energy markets. His research reports and articles have been published in many international journals and magazines. Pattabhiram holds a Masters degree in Thermal Engineering from the College of Engineering, Guindy, India, and a Management degree from the Indian Institute of Management, Ahmedabad, India.