Business Analytics Yearbook

Download Business Analytics Yearbook

Business analytics technologies can conveniently be placed under the umbrella of Enterprise Decision Management (EDM). These systems exploit methods and technologies to improve the efficacy and efficiency of decision making through the organization. This is in contrast to traditional systems which have almost exclusively been concerned with process automation and improved efficiency through labour displacement. The technologies employed in EDM include predictive analytics, business rule management, optimization, business intelligence, and in fact any technology which reduces the uncertainty involved in decision making, and increases decision making efficiency.

The traditional business use of information technology can be seen as an extension of the filing cabinet and desktop calculator. Computers have been used to store and process ever larger amounts of data and automate the calculation of quantities used in a business environment. These systems are also fairly inflexible, requiring the hard-coding of business rules and the use of complex programming languages to implement them. This is a deterministic world with no room for nuance or subtlety. Process automation peaked in the first decade of this century with ERP, CRM and other systems aimed at integrating and formalizing the transactional activity in an organization (with the unfortunate side effect of making the organization less flexible). This is now a domain of diminishing returns, and certainly much inferior returns compared with EDM.


In contrast EDM uses the computer as an uncertainty reduction machine, employing statistics, machine learning, data mining, optimization and business rules engines to fine tune decisions and massively increase the speed at which they are made. In fact the current surge of interest in business intelligence (BI) tools and techniques is a testament to the urgent need to have technology help in the decision making process, although BI is labour intensive and prone to misinterpretation. As always the leaders in the use of EDM can be found in financial services with decision systems employed in loan approval, the detection of fraud, customer targeting and so on. The ‘digitization’ of business processes, an era that has persisted for fifty years, is now being complemented by the ‘digitization’ of decisions, and this new use for information technology will dwarf what has gone before it.

Any technology capable of reducing decision uncertainty, and reducing decision latency qualifies as an EDM enabler. Predictive analytics technologies scour historical data, looking for patterns that might be reliable enough to employ in future activities. Typical applications include reduction of customer churn, better sales targeting and other applications such as prediction of machine failure, or even the likelihood of hospital readmission. Technology startups are providing SaaS types services where business users, with little technical skill, can upload data and create their own predictive models. There are dangers associated with a ‘black box’ approach of this type, but it does at least indicate the way things will go. Larger organizations can afford to employ a team of data scientists and analysts to create bespoke predictive models and methods.

Optimization is another technology usually bundled in with EDM. This is primarily concerned with determining how resources should be deployed once the question of what will happen in the future is determined (the province of predictive analytics and statistics). Given a set of resources and constraints, optimization will work to maximize a given objective – usually profit and/or revenue. It answers the question ‘how’, given we know ‘what’.

Finally the use of business rules engines complements both predictive analytics and optimization by saying what is, and is not permissible. A predictive model may suggest selling a given item at a certain price for example. However if the product has already been offered at a lower price to a subset of customers, it simply cannot be used in their cases. And optimization may suggest working practices that are unpopular or even illegal.

EDM is a sea-change in the way businesses use information technology, and the benefits that might be derived from it. Its effective use will distinguish the winners from the losers in a way we haven’t seen before. Needless to say this all requires awareness at the very top of the organization, and there are profound organizational and cultural implications. We will after all be increasingly handing over the decision making process to machines – so we really do need to know what we are doing. Greater reward always implies greater risk, and EDM is no different. The risk mitigator is skill and knowledge – in a world of change some things never change.