Data-driven decision making for economic prosperity and good governance – part I

By Dr. Ranga Prabodanie

Just as the industrial revolution transformed the agrarian economies to manufacturing economies during 1970s and 1980s, the “data revolution” is expected to transform the world towards a global digital economy within the 21st century. Unlike the coal-fueled industrial revolution, today’s data revolution has opened a fair game allowing all nations, businesses and individuals to harness the relentless tide of data. Today, the production, storage and consumption of digital data has become a virtually unconscious process. Billions of people and devices connected to internet are generating massive amounts of data on people, products, markets, weather, traffic, etc. and the list goes to infinity. Such data in large volume, velocity, variety and complexity known as “big data” together with advances in technologies such as Internet-of-Things (IOT), Artificial Intelligence, Cloud and Edge Computing, Robots and 5G networks have paved the way for the 21st century data revolution. Cost efficiency, scalability and adaptability of these technologies have created an open competitive environment where anyone who take strategic advantage of data by uncovering the hidden insights can reach the heights of success.

Broadly defined, data-driven decision making is the process of collecting and analyzing data and generating insights from data to inform decision making. It’s the philosophy of making decisions based on real data rather than on intuition. This article attempts to rationalize a nation-wide movement to explore data strategically for economic and social development. To establish the context, let’s first take a look at how successful businesses use data smartly to drive their profits.

Data-Driven Business Models

Amazon, the world’s largest online marketplace, which placed its founder Jeff Bezos among the richest in the world, is a typical success story of data-driven decision making. Amazon’s product recommendation algorithms make intelligent predictions on which products each customer would be interested in, based on what the customer have viewed, bought, ranked and reviewed previously. Highly personalized product recommendations are driving Amazon’s profits and sales. By carefully tracking and analyzing every single aspect of customer behavior from mouse-click to product search and purchase, and by integrating the insights generated from such data into their decision making, the company has realized staggering profits and sales over the years.

Coca Cola recently introduced a new flavour Cherry Sprite based on the data recorded on self-service soft drink machines that dispense drinks according to customer specified mix (Bernard Marr – an article on Forbes magazine). They have utilized extensive analysis of sales data to identify trends such as decline in demand for sugary drinks and to optimally combine several factors including price, taste and packaging to match the expectations of local customers in more than 200 countries throughout the world. The company has developed and promoted healthier options such as Minute Maid in response to market intelligence they have generated not only from sales data but those shared on social media.

Uber, which revolutionized, the transportation industry within few years of its inception, is fuelled by data they collect on bookings, trips, travel patterns and behaviors. You may have enjoyed the luxury of picking up an Uber Taxi within few minutes out of the shopping mall or airport. How do they make the taxis available when and where we need them just-in-time? Uber uses historical data on times, days, and locations where the demand occurred and the trips completed to perform extensive analysis which identify areas where the demand can exceed supply. They then inform the drivers to move to such areas well in time to explore the rising demand. Uber uses data extensively in every business function including pricing, driver rating, traffic monitoring, driver guiding, and fake-rides detection.

Ecommerce in Sri Lanka

If we compare those three multinational business models with similar businesses in Sri Lanka, our retailers, soft-drink businesses and taxi services do use information technology, and are not necessarily under-resourced, at least in terms of digital data; but why cannot they stand out from others, cross the national boundaries and make global brands? Supermarket chains such as Keells, Arpico and Cargills collect massive amounts of sales data just as Amazon does. If they properly analyze the data, they can learn more about consumer behaviors and tastes (e.g. fast-moving brands, products bought together, quantities purchased, shopping frequencies, arrival times) and use those insights to offer a unique shopping experience. There was a boom in online shopping during the pandemic lockdowns, but unfortunately, all our retailers had similar ecommerce platforms with same basic functionality and they lacked data-driven innovation. Shoppers have to search, check availability, select and add each item to the cart coping with annoyingly slow web site performance. Reducing search, and thus the transaction cost, is the fundamental concept which drives the success of ecommerce business. If our leading retailers had ever looked at the sales data seriously, they would have known that people usually buy the same brand of milk powder every time and same set of consumer products every week. And who knows, there may be other surprising patterns hidden underneath the heaps of sales data. Simple data-driven innovations such as suggesting or adding customer favorites to the cart automatically can escalate retail sales and profits while saving the customers’ valuable time.

Understanding the Customer

The taxi services in Sri Lanka also have online and mobile booking facilities and all the data on service requests, bookings and completed trips are recorded. If those records are mined using appropriate methods and technology, they could reveal unimaginable insights that would help improve the service. Success in service sector is hugely dependent on understanding the customer. As Coca Cola has always acknowledged, customer tastes vary across continents and they evolve rapidly over time. Hence, acquisition of market intelligence through continuous data collection and analysis is of utmost importance for the fast-moving service sector.

Within the service domain, banking, finance and insurance sector is usually the forerunner in adopting up-to-date technologies. Almost all banks in Sri Lanka have electronic records of customers, accounts, transactions, savings, etc. but what rarely happens is the analysis of those records to generate insights on specific customer groups and their specific needs. Previously unexplored opportunities may arise with emerging trends such as delayed retirement, overseas higher education and online shopping. In such dynamic business environments, data is the primary source of market intelligence and the perfect recipe for understanding the customer.

Data-Driven Governance

It’s not only the profit-seeking businesses which capitalize on the mass influx of data. Forward-looking nations use data to generate intelligence that drive public policy and delivery of public services ensuring transparency and accountability. In developed countries like US and Australia, data-driven predictive analysis is used to estimate and manage the demand for health and aged care services. Emergency services use historical data on emergency calls, locations, response times, etc., for efficient allocation of resources to improve their preparedness and response. Several countries use crime data analysis (dates, locations, types, victims, suspects, etc.) for risk assessment, to improve crime intelligence, and to fight crime more proactively. Simulations based on weather data, hydro-geographical data, emissions and extractions are used for measuring and predicting environmental impacts and informing government policy towards mitigation and conservation.

Data use in Developing Countries

The world bank report on “Big Data in Action for Government” highlights several examples of developing countries adopting data driven decision making to serve their people better. A state government in Brazil has used a mobile app to collect real time data on health care services which enabled them to identify issues such as bribing and to respond timely and efficiently. In Seoul, South Korea, data from taxi services have been used to optimize the nighttime bus service by matching the origins and destinations of the trips. Kenyan government uses a mapping platform to identify areas where there are shortages of education resources. In Mexico, data from student interaction and feedback are analyzed to identify problems and continuously improve the education process. The municipal government of Shanghai has significantly reduced the maintenance cost and service disruptions in their water supply network by monitoring data obtained from sensors installed at various points of the network to identify issues such as leaks. South African government is using data from satellite images and mobile apps to geo-locate households with a view to digitize their census process. In India, data on nightlights captured from satellite images are used to monitor electricity supply. There are countless other examples of governments using widely available technology to collect data on public services, market performance, investments, and other socio-economic indicators. They analyze the data to identify trends, patterns, issues and developments and to predict future events, outcomes and behaviors. The insights and predictions then inform and guide the policy formulation.

Above examples provide evidence that even the developing countries can use data strategically and innovatively for economic development and social well-being. The next part of this article will focus on where we are today as a nation in the path towards the data revolution and where we can be tomorrow.

(The writer is a Senior Lecturer in Wayamba University of Sri Lanka. However, the views and ideas presented are those of the author and do not reflect the policy or position of any institution.)