Business Intelligence - the "what?", the "why?", the "how?" and the "what next?"
In my last blog I asked a basic, yet I believe, a very important question. Is it better to speed up or simplify? The context was core ERP business processing, and cutting to the chase the answer I gave was simplify. The rationale - simplification delivers speed, and agility, and speed, and efficiency, and speed, and better affordability, and speed, and … you get the picture. I now want to explore this same concept within the Business Intelligence space.
When it comes to Business Intelligence we typically think about this across four dimensions – three of which are physically measurable, one is derived.
The first dimension is “what is happening right now”. This information, generally termed “operational reporting” resides in your transactional systems recording important information like “orders taken in the last minute”, “invoices posted in the last hour”, “support calls on hold”, and so on.
The second dimension is “what has happened”. This information resides in your analytical systems, allowing you ask more complex questions such as “who bought what product, in what market, over what time frame”. Often this analysis requires data in one system to be joined to data in another system. Subsequently data is replicated to a dedicated reporting platform we call the “Enterprise Data Warehouse”.
The first two dimensions tell us what we sold and to whom, but they don’t always tell us why. “Why did it happen” therefore is the third dimension of Business Intelligence. Enter contextual data. Contextual data reveals insights such as “there is a strong correlation between weather and sales” and “the product on arrival was over ripened due to excessive heat exposure during transportation”. In other words we are talking about information bound up in semi and unstructured data sources such as social media, sensor data, web log information, weather data, spatial data, etc.
The fourth dimension is "what will happen"? What will happen type questions take as an input all of the data from the first three dimensions and use it to simulate what will happen next – essentially working with various predictive models to work out the next best action.
The Business Intelligence challenge lies with the fact that these data sets are more often than not residing in loosely coupled, if not disconnected silos. Subsequently it is very difficult to answer these questions in a timely and consistent manner, especially the last one – i.e. what will happen?
The solution to the 360o Business Intelligence challenge is to break down the walls that currently exist between each silo, thereby enabling an uninterrupted view of these disparate data sets as though they were part of the same physical system. If you succeed in doing this you …
- Know what has happened in the past – and why
- Know at all times what is happening right now – and why
- Know what will happen – and why
- Know what the best thing is that could happen – and how
In my next blog (or three) I will explain how SAP is eliminating the silos through simplification within each layer. I will explore what it means to deliver true operational reporting, how you can create real agility in the core EDW space, how you can unlock and join Big Data to operational data, and finally how you can employ predictive, taking in all three data sets, to enable next best action.