Axis My India, a market research organizations known for its exit poll analyses, is the first consumer data intelligence firm from the country to have become a part of the Harvard Business School curriculum. The Ivy League business school has built a case study based on its election forecasting model.
The case study, now a part of HBS’s classroom course on elections, highlights the complexity associated with successfully predicting elections in the world’s largest democracy, with varied geographies, shared borders with six nations, sprawling rural populations and 23 different languages spoken across the country.
This case study written by Professor Ananth Raman, Senior Associate Ann Winslow and Research Associate Kairavi Dey discusses the process which goes behind forecasting elections in India and the methods AMI uses to reach such accurate forecasts – selection of field surveyors, hiring & training, use of technology for data collection, quality auditing, data analysis & final forecasting.
Professor Ananth Raman, UPS Foundation Professor of Business Logistics Chair, OPM, HBS said in a release: “This case study illustrates numerous operational details, including those associated with training surveyors and moving them across different locations based on their linguistic and socio-economic identities to get a feel of the electorate’s pulse. Predicting elections accurately in an extremely complex country like India is difficult.”
According to Professor Raman, while we have seen numerous examples of experienced organizations struggling to predict election outcomes in the recent past, this case illustrates how an entrepreneurial organization like Axis My India has devised a process to predict outcomes accurately.
Some of the key details compiled in the report that sets Axis My India apart from all of its peers include presence across length and breadth of country (700+ districts), an average survey size of over 5 lakh people for national surveys, GPS-enabled tablets to maintain geographical sanctity, as well as computer aided questionnaires backed by social intelligence to garner maximum data veracity.
