Subscribe
About

Towards a new vision of operational BI

There is a clear need for BI to be made available to the broader base of business.
Corey Springett
By Corey Springett, Strategic business manager at Progress Software South Africa.
Johannesburg, 07 Aug 2008

"Reports from ad hoc queries are now generated in less than five minutes versus six to nine months."

"Our users and analysts are more productive, focusing on our most complex and demanding BI applications."

"The IT department spent hours each day on behalf of users simply pulling data from disparate systems."

These are three actual customer endorsements of a new way of delivering business intelligence (BI). It is an approach that merges the best of two separate but interlinked disciplines and brings BI to the masses.

These two disciplines are:

* Traditional BI, which delivers results, but unevenly and only to power users. It tends to occupy the time, attention and resources of IT, causing resentment among stretched IT specialists, and among users, who desperately need access to organisational information which is locked up in transactional and other corporate applications and databases.

* Search technology, which while ubiquitous and easy to use, is limited to information which has been pre-aggregated and sorted by people who cannot conceivably have specific users' interests and considerations in mind. This means that any traditional search query, as per Google, will return nonsense relative to the requirements of the business.

However, there is a clearly identified and often articulated need for BI to be made available to the broader base of business - for it to be "democratised", or taken to the masses.

Operational BI is where the needs of workers at the coalface are satisfied, without fuss and bother, and definitely without the intervention of IT.

Corey Springett is strategic business manager at Progress Software South Africa.

Depending on whom you speak or listen to, BI adoption stands locked at between 7% and 20%. Whichever way you look at it, this means that 80%, and more, of people in organisations are not deriving the promised value of BI.

Does BI need to be taken to the corporate masses? It can be argued that it doesn't belong with everybody. But let's look at a real-life scenario involving an inbound and outbound call centre involving a vehicle tracking company with hundreds of agents who need to know details of which cars have most recently fitted tracking devices.

According to the traditional hierarchy of BI, these users would be classified as operational users, and their requirements would be subsumed to those of the strategic and analytical users.

The biggest problem in this scenario is that customer-facing staff, among them salespeople, have no single, united, comprehensive view of customers. Yet they are required to compete, on an even footing, with their opposition who might have superior information and insight.

Operate on gutfeel

It is in such a situation that users start to step outside the corporate reporting and BI standard, or make decisions without supporting facts. This is what is known as operating on gutfeel.

Neither situation is desirable, but this is the reality when users find themselves unable to obtain the information they need to do their job.

The solution, and the one that has been adopted by the companies quoted above, is the adoption of what is now known as operational, or pervasive BI.

This is a model of BI that ensures BI is both easily available to the broader base of business users, and that they can ask meaningful questions, at a decent level of granularity, and with some drilldown and slice-and-dice capability.

If current experience and wisdom are that 20% of users can actually use BI, then the inverse of the Pareto Principle applies in that 80% of requisite functionality can be given to users through operational BI.

Operational BI is where the needs of workers at the coalface are satisfied, without fuss and bother, and definitely without the intervention of IT.

It is a requirement best solved today through the twin approaches of natural language queries and underpinning conventional BI infrastructure.

In such an approach, a user types in his query in the most basic and understandable English, and an interpretation, parsing layer makes sense of this to the underpinning BI infrastructure, which interprets the query and returns the answer within two seconds.

This is the new world of operational BI, and it should be the one gaining industry-wide adoption in the next few years, as it has already in South Africa.

* Corey Springett is strategic business manager at Progress Software South Africa.

Share