If Business Intelligence (BI) projects made difficult because business doesn’t understand technology – it is worse still, according to Gartner because "business doesn’t understand itself".
In a nutshell, BI technology overlays skills, processes and practice to support decision making using historical and predictive data and models. It is decribed by Gartner as the Rosetta Stone of IT.
Gartner analysts said at the Business Intelligence and information Management Summit in Sydney today that BI requires talk, rather than technology, and advised IT managers to set up a competency centre to promote BI projects.
But Australian IT managers are failing to to push for the critical cultural shift, according to analyst Ian Bertram who delivered the keynote presentation with US counterpart Bill Hostmann.
“Australians use BI as a crutch,” Bertram told Computerworld Australia. “They are up there with the best insofar as technology deployments but fall behind in achieving cultural change.”
IT engineers don’t typically make good evangelists, according to Hostmann who is a former tech engineer. He says BI project leaders should enlist an ambassador to promote the project throughout the organisation, while IT focuses on understanding business operations and requirements.
However, Bertram said it is not easy to obtain a consistent language from the business, an essential requirement which forms the bricks and mortar of BI projects.
“Not only is there an age old problem between IT and the business, there is a problem between business and the business – they don’t even speak the same language,” Bertram said.
“Supply uses different language to finance, which uses different language to HR.”
To illustrate the effects of slack BI, Hostmann pointed to companies that post letters addressed to multiple names but ultimately meant for one customer, which occurs due to a lack of agreement as to what constitutes a customer.
Bertram said one Australian hospital that deployed BI had multiple definitions for a received shipment date, which varied according logistics, doctors and finance. It took two weeks to resolve the discrepancies.
“You need to get a glossary of terms that must be transparent across the organisation,” Bertram said.
Many business decisions, such as recovering outstanding debt, are rational and can be automated and drilled into applications, Bertram said. Others are more difficult and require each arm of the business to be canvassed about its decision-making processes, which Bertram in jest said is akin in difficultly as his experience in planning his home renovation with his wife.
Businesses should look to off-the-shelf technology to implement a holistic predictive model, Bertram said, to avoid departmental silos created as each business unit maps its own models. He told the summit audience to look to successful local cases studies to explain the benefits of BI to the business.