What has analytics got to do with operational research (OR)?

For businesses and decision-makers seeking to improve their operations, ‘analytics’ is often marketed as a panacea solution to any problem. However, nailing down a definition is quite difficult.

One of the most common definitions is: “the systematic computational analysis of data or statistics.[1]”

But this definition does not account for the many subcategories of analytics that have arisen (advanced analytics, business analytics, data analytics, web analytics…) or their purposes (deductive analytics, predictive analytics or prescriptive analytics).

Operational research has evaded definitions in a similar manner, with attempts ranging from ‘the science of better decision-making’ to ‘the scientific process of transforming data into insights to make better decisions’.

One could be forgiven for struggling to see the difference between these two interconnected fields or toolboxes.

Convergent evolution

Operational research (OR) began emerging as a discipline in first and second world wars and has continued to evolve its tools with the times and with new technologies. The rise of global connectivity and big data have not lessened OR’s relevance – indeed, they have created challenges that OR is uniquely placed to tackle.

Analytics and OR share many similarities, however. Our sister organisation, INFORMS, says that “OR and analytics enable organisations to turn complex challenges into substantial opportunities. They transform data into information, and information into insights for making better decisions and improving results.”[2]

Both OR and analytics generate insights and solve problems, but are developing along different timelines and are at different levels of maturity. It is reasonable to see analytics as having been developed as a response to business problems, unaware of the existence of its more developed cousin, operational research.

So what is the big difference between OR and analytics?

The difference is this: definitions of analytics tend to ring true for operational research, but not the other way around. OR is wider, broader and deeper in its array of problem-structuring and problem-solving tools, thanks to its longer, richer history of gathering data, processing it into intelligence and informing decision-makers.

The OR Society’s own definition is: “Operational research (OR) is a scientific approach to solving problems in the management of complex systems that enables decision-makers to make better decisions.”

The word approach is key. OR is more than a toolset – it is also a mindset and a way of looking at the world. Core operational research disciplines, such as simulation and optimisation, continue to evolve with the times and draw upon the latest analytical technologies, but this does not make them ‘analytics’. Indeed, it makes more sense to argue that analytics is a subset of operational research.

To attend our one-day conference that combines OR, analytics and AI, visit our AS22 page.

To find out more about getting certified as an analytics professional, visit our CAP page.

To find out more about the history of OR, watch our ORigin Story video.

[1] Definition of ‘analytics’, Oxford Languages

[2] INFORMS website, bit.ly/3OuTkqD