Partnering with ITWeb, YoungBlood Consultants conducted an online Business Intelligence (BI) Benchmarking Survey to determine how prepared organisations are to efficiently carry out their business intelligence strategies and to better understand the customer's viewpoint regarding business intelligence.
The survey, which ran online for two weeks during March this year, captured responses from a wide range of businesses across various industries and sizes (a combined 49% were from organisations with more than 1 000 thousand employees).
The results were more or less evenly split when respondents were asked what their business needs focused on, with 33% emphasising operational reporting, 20% focused on providing insights that enable strategy, and 18% concerned with providing decision support to line management.
"Very often operational reporting is the most time-bound of all levels of reporting: If your production line has stalled, you need to find the reasons immediately. If customers are queuing, you need to know immediately and deploy more floor staff. If your systems have a backlog, instant action is necessary," says Patrick Wilson, MD of YoungBlood Consultants.
Wilson believes that BI is best in these contexts when alerts can be generated by close-to-real time observation, and delivered on a mobile platform.
"The traditional tabular operational reporting to identify exceptions is nowhere close to quick enough or granular enough to deliver the call to action that is needed," he notes.
Just over a third of respondents reported that information frequency was generally more than a day old (32%), 22% said it was more than a month old and 20% said it was less than an hour/real time old.
"Many organisations can live quite happily with day-old information. If the strings that steer your business are themselves more than a day long (operational latency), often it makes no difference if information is received early."
Wilson says often the opposite is true - the quantum of truth that a piece of information carries in an organisation is often determined by the number of other data items that agree with it.
"Sometimes this takes time, but the information item itself becomes 'more true' with supporting evidence," he adds.
According to Wilson, older data can be automatically stored in a cheaper, slower-response medium if it is not needed for low-latency trend analysis or comparison. If it is needed, the new wave of storage media and database technology can deal quite well with having big data sets at the fingertips.
It emerged from the survey that the respondents' provision skills are evenly split with departmentalised at 40% and all in IT at 41%. Only 19% indicated a central service organisation.
"There is no silver organisational bullet. If the business changes slowly, then much BI will be predefined and predeveloped and slow to change itself. In this scenario, a central IT development resource may cope fine. However, if new requirements are popping up in fast-moving areas of the business, departmentalised BI (i.e. where the business department has its own BI resource) may be better," Wilson advises.
It's not surprising that 14% of respondents indicated that it can be catastrophic if old information is not being utilised to gain a competitive advantage. Over half of the respondents said the consequences can be serious in areas where they have a competitive advantage.
"We have found that it is mostly a technology problem. People are used to information disappearing when it gets old. BI practitioners need to pro-actively engage with business to explore the value that can be added by the addition of older data to a statistical mix. The role of the data scientist is important in this - the DS will explore the value that new approaches to data and its analysis can add value to the business, and this must include old data," says Wilson.
He also believes that one item that hasn't been addressed in the BI maturity research is that of comfort, and believes that every organisation has a zone of sophistication of data usage with which it is comfortable.
"Staff are used to turning to BI sources for certain decisions, certain processes are known to need data and the role of data is therefore well supported in those places. The natural tendency is for the operators in the comfort space to seek to return to it whenever new developments are started. This is death to migration up the BI curve to the necessary level of maturity," he says.
Wilson concludes by saying that the discomfort caused by remaining at a certain level, with the associated practices, structures and data, needs to be more than the discomfort to be caused by change. Energising the discomfort of standing still is a vital role for management to play in managing the organisations' maturity growth.
Share
