We use business intelligence (BI) platforms to understand operational activities and diagnose problems. Charts and dashboards have become a favorite mechanism to do this work, but we have to know what we are looking for, and then dig into the data for answers. There are two significant problems associated with this approach:
- Do we know what we need to know? So sales dropped in a particular region in the previous quarter. Obviously we look at sales rep performance, and maybe which products suffered most – and so on. But there may be other factors that have not been encountered before and so we might miss them altogether, or take a long time unearthing them. Unknown to the person digging into the data was the fact that the regional sales manager resigned unexpectedly, and it was several weeks before a new one could be hired.
- What is just random noise in our data and what is significant? We see an unexpected rise in productivity in several factories for a couple of months. Is this significant, or does this come within the range of variation that might just happen randomly?
Traditional BI tools do an excellent job of allowing users to slice and dice their way through data, but these same users come pre-loaded with past experiences, and so their investigations will be biased. A new generation of BI tools is making an appearance which understand the data, and the relationships between various sources. Qlik, with its associative engine does this to some extent, but Panorama and OnlyBoth take inbuilt intelligence to a new level. Panorama will try to correlate cause and effect at the click of a mouse, whereas OnlyBoth performs complex statistical computations to identify what is normal, and what isn’t. Both of these platforms assist the user identify cause and effect within the business – something that traditional BI tools just do not do.
Panorama supports all the usual visuals – and a few more besides. But what differentiates it is its suggestive engine. As the term suggests, this suggests data that is likely to be of interest, and the more the platform is used, the more it learns.
OnlyBoth uses the term ‘benchmark’ to describe the profiling of entities (customers, products etc) and employs an embedded artificial intelligence engine. In addition it also creates written reports revealing the insights it has found. Right now OnlyBoth provides a free cloud based service with a limited commercial usage license.
Intelligent BI will soon be the only BI, simply because our data are becoming too complex for individuals to manually look for cause and effect relationships. And the relationships themselves are becoming more complex. It is well known that the average person can juggle six or seven variables together at once – max. Of course computers can juggle hundreds if need be, and find candidate cause and effect relationships much more easily.
The problem of interpretation will become more pressing as users create ever more visualizations to reveal meaning in their data. In the end the task will be beyond human capability, and so intelligent BI platforms are not only necessary, but they are also an inevitability.