Predictive Analytics – The Future for Retail - IEDP
  • Customer Focus

Predictive Analytics – The Future for Retail

In a recent article, Indian School of Business’ Professor Galit Shmuéli explains the value of and methodology behind ‘predictive analytics’.


By downloading this resource your information will be shared with its authors. Full privacy statement.

Using ‘predictive analytics’ to exploit customer data is within the reach of more organizations than ever before - whether ‘big data’ or just the data we hold. For retailers particularly, analytics that predict the future could be a dream come true.

In a recent article, Indian School of Business Professor Galit Shmuéli explains the value of and methodology behind ‘predictive analytics’. Considering it from a retailer’s perspective, she also debunks certain myths and suggests how it can be integrated into business processes.

Predictive analytics is a step up from more traditional forms of statistical business intelligence, which are about looking at the past or the present. Predictive analytics uses past data for the purpose of predicting future events. Its focus is on the micro rather than the macro, looking at individual interactions with customers, suppliers, employees, etc., rather than at average behaviour or at high-level aggregate patterns.  Looking to the future based on past behaviour to predict: Will a particular customer engage or churn?  Or will a particular supplier deliver on time?

In our technology driven, data intensive, world most retail businesses have a grip on data generation and collection. But they do not necessarily know how best to leverage this data to drive business growth. Predictive analytics can be one of the keys to this. Applied to relatively large sets of customer data it can enable marketers to predict future behaviour, and based on such predictions, customise the best customer offers, or ways to interact with customers or suppliers. According to Professor Shmuéli “In a diverse country such as India, where variability is at its core, such personalisation to customer preferences carries especially great weightage...”

Perhaps Shmuéli’s key observation is that “To gain useful results, it is critical to have a close connection among predictive analytic techniques, data, and business needs and requirements. The process of predictive analytics implementation, therefore, begins with the most difficult step of translating a business problem or challenge into a predictive analytics problem.”

In the article she uses the case of online retailer to explain the process required to implement a predictive analytics project, under the headings:

Problem Identification: translating the business challenge into a “precise predictive analytics formulation” with measurable outcomes.

Measurement and Data: determining which measurements are of interest and can be applied. Considering how to access sufficient data and integrate data from multiple sources, focusing on the end goal of the analysis and potential insights rather than the quantity of data.

Once assembled, the data set is randomly divided to test the model before deploying the model on completely new data.

Model: the model is built by trying different algorithms and models to search for patterns and correlations in the data, using various ‘data mining’ techniques and methods from artificial intelligence, statistics and other disciplines.

Deployment: adjustments may need to be made as new factors emerge during the deployment, but when the predictive algorithm is deployed, it will generates a probability for new customer interactions etc.

Uses of predictive analytics by retailers includes: developing direct marketing campaigns; personalising customer offers; improving employee training; building customer retention; forecasting demand and optimising stock levels; etc.

Professor Galit Shmuéli goes on to offer several tips for practitioners of predictive analytics, not least that “A key determinant of success is support and even leadership from upper management. Predictive analytics can lead to successes, but there is often a learning period in which there will be failures. Managerial support must extend to allowing failures, as long as they result in lessons learnt.”

And as a concluding tip she returns to her first point that “Business and statistics go together in predictive analytics. When applying findings from the analytics component, one must carefully align outcomes to reality and not just apply them mechanically. Without business knowledge, deploying analytics can be useless at best and disastrous at worst.”

Applying predictive analytics with care and creatively, companies can strengthen their relationships with customers and other stakeholders and reduce the uncertainty behind many of their business decisions.

Read the Full Article from ISB Insight

Google Analytics Alternative