Organisations must adopt an enterprise-wide, non-siloed approach to performance management implementations, and ensure that business intelligence (BI) technology is not confined to a small group of senior decision makers if they are to reap the benefits, according to industry experts at this year’s SAS Forum in Stockholm.

In the opening keynote, SAS senior vice-president Jim Davis argued that BI – which he described as effectively the presentation layer in an enterprise ‘Intelligence Platform’ – should be democratised throughout an organisation.

“One size doesn’t fit all,” Davis argued. “You must serve the data to a broad spectrum of users [through BI tools], so you will need many different interfaces.”

Butler Group analyst Sarah Burnett added that for performance management applications, giving access to data “at all levels" is extremely important for success.

Lars Takla, former chief executive of energy company ConocoPhillips, advised that letting all employees be involved in the system would allow them to "feel ownership”.

He added, “The performance-based management system [we implemented] enabled everyone to have the stats and info they needed to make decisions."

Experts at the show also warned that taking a siloed approach to implementing performance management systems is unlikely to bring the business benefits associated with a more holistic approach.

“There is a different order of magnitude impact if you have enterprise-wide capabilities,” argued Jeanne Harris of the Accenture Institute for High Performance Business. “There’s nothing wrong with st arting off in functional areas or divisions but the broader you can drive it into your business the broader the impact and benefits.”

Butler Group’s Burnett added that firms which implement performance management in silos risk creating conflict between individual departments.

Also in his keynote, SAS’ Davis claimed that analytics is “one of the most used and abused terms in the market”. He argued that it is often misapplied to tools which only have capabilities to analyse historical data, such as querying and reporting. These are of limited value if used in isolation by firms, Davis added, and so the analytics label should only be used to refer to tools that also include predictive capabilities such as forecasting and predictive modelling.