Businesses don’t generally make investments in people, technology, or anything else, unless management believe there will be a return. And it is helpful if the return is larger than the investment. Business analytics technologies are touted as the key to more effective and efficient businesses, but when it comes down to details it all gets a bit hazy. So today we find businesses investing in ‘big data’, business intelligence, predictive analytics and other technologies with the aim of enhancing top and bottom lines, but often without any well defined operational plan. This did not happen with investments in other operational systems (accounts, sales, purchasing, etc). In fact the investment in most operational systems was usually predicated on a good understanding of savings and efficiencies that would be realized (reduced head count usually, but this brutal reality was not usually stated so explicitly). When it comes to investments in business analytics technologies we tend to lose this keen focus and buy in to the notion that having a few hundred people playing with data will somehow deliver a return. Sometimes it does, and sometimes it doesn’t. Either way it’s a pretty sloppy way of going about things.
Let’s be clear about what we want from business analytics. We need more accurate and more timely decisions that cost less to process – and that’s it. It might be called decision automation, or at least assisted decisions. We ponder over reports and dashboards to help us make decisions, and we deploy predictive models to help automate decisions. And so it would be useful if we had some understanding of how business analytics might be operationalized. These can be summarized in the following way:
- More efficient and effective business decisions. This requires an understanding of the decisions that are routinely made within the business (customer credit approval, best upsell offers etc) and the technologies and systems needed to enhance these decisions. The effect of systems deployed in these ways can be easily measured, and it soon becomes apparent how such systems are contributing to top and bottom lines in the business.
- Understanding the decisioning processes of our customers and business partners (distribution network for example). When a business operates in its markets it collects transactional exhaust in the form of data. These data have become a valuable resource in their own right, and most businesses can create data products which become a source of revenue in their own right. These data products need to be addressing the uncertainties of customers, trading partners, suppliers and whoever might find the information useful. The more uncertainties they address, the more people will pay for the information. It just needs a little clear thinking to deliver information based systems that become revenue generators in their own right. So we lose the vague notion of data being a valuable asset and we create something that is operational and profitable. If you want some concrete examples of this then look no further than GoodData. This company provides the technology, infrastructure and skills to convert data assets into profitable production analytics.
- All of this can be taken a step further by using external data sources, and particularly those of businesses in related markets. So it is easy to see how airlines, hotels and car hire firms might merge data to create pictures of customer behavior that are more accurate and more useful than a single business might be able to create.
All of this is a long way from the current notion that charts and dashboards somehow confer an advantage to a business in their own right. In the main they don’t. These are early days for business analytics and we desperately need to move to a production oriented approach, where profitable business analytics can be realized. The alternative is a gradual disillusionment with business analytics technology, and missed business opportunities.
Many thanks to Roman Stanek, CEO of GoodData for sharing his thoughts on these topics.