Business intelligence, like military intelligence, is sometimes jokingly described as being a contradiction in terms. But in fact, in the era of big data and the Internet of Things, the potential benefits to be gained from business intelligence (BI) are greater than ever.
BI is about turning raw data into useful information for business analysis in order to aid the decision making process. The technology therefore needs to be able to handle large volumes of structured and unstructured data, turning it into reports that are easily understood and deliver insights that can provide businesses with a competitive advantage in the market and help their long-term stability.
Then idea of business intelligence actually pre-dates computing, the first recorded use of the term was in 1865. It took the information handling power of computers to really unlock its full potential though. The idea was being looked at by IBM in 1958, but it was the decision support systems of the 1960s onwards that began to get close to what we think of as business intelligence today.
It wasn’t until the 1990s that business intelligence became a mainstream concept for the use of information in support of enterprise decision making. Since then it’s been associated with other technologies, primarily data warehousing that provide a pool of data for analysis.
Modern BI can take in a wide range of different applications including data mining, statistical analysis, predictive modelling and data visualisation. The rise of big data means that Hadoop is increasingly being used to support BI infrastructures.
Within a business, BI can be applied to a number of different areas of operation. Collection and measurement of performance metrics is probably the most common, the information produced can then be used by managers to measure progress towards goals. This can be used in a proactive way in order to issue alerts if certain conditions aren’t met.
The next step from this is analysis of the data to produce meaningful insights. This can involve a number of sub-processes including data mining and business process modelling, allowing the business to have access to data to support decision making, but also to unlock hidden knowledge.
This data can also be used in creating strategic reports in order to give management an overview of the entire business. This will usually involve data visualisation techniques and may feed into an executive information dashboard that can highlight changes to key indicators. It can be used for collaboration too, whether between departments of a large enterprise or with external customers and suppliers.
BI can also be used as part of a knowledge management strategy to ensure that information relating to key parts of the business is readily available.
Business intelligence isn’t something you can simply throw a switch and start using. It requires a level of commitment from the top of the organisation downwards. It’s important that everyone involved in implementing BI has a clear idea of the benefits it can bring and the objectives of implementing it. Potential users of the system at all levels need to be consulted as part of the initial planning process.
Organisations turn to BI for a number of different reasons. Often these are driven by competition and the need to gain a competitive advantage against other companies in the market. Another common reason for implementation of BI is the acquisition of or merger with other organisations. When the business is enlarged in this way BI can be used to create more oversight into the larger whole. BI tends to be favoured by larger organisations as it offers visibility into operations that may otherwise be hard to obtain.
As with most IT-based initiatives, BI is only as good as the data it has to work with. Many businesses choose to start the process by undertaking a data profiling exercise. This pinpoints where data is located and looks at its quality and structure. If the data isn’t easily available or is of poor quality then these issues need to be addressed before BI is put in place. Ensuring the data is in a suitable format may involve creating an operational data store to correct and clean up raw data, and to check its value.
To be properly effective BI needs to be properly integrated into a company’s other systems. If intelligence results are hard to access and the reports difficult to interpret, users won’t make effective use of the system and its value to the business will be reduced.
There are a number of companies offering business intelligence solutions. These include big name companies like IBM, Oracle, SAP and Microsoft, but also specialist suppliers such as Birst and GoodData. Offerings range from comprehensive sites that deliver business dashboards, notifications, reports, strategy recommendations and more; to specialist applications that focus on specific areas such as visualisation if data.
As with other business software, increasingly solutions are based in the cloud, Microsoft’s Power BI for example is a software-as-a-service solution aimed at being easy to use for non-technical business people. is designed to operate on public or private clouds and to provide users with easy access to a suite of analytical tools.
As technology extends into more areas of our lives via the Internet of Things, the amount of data business collect will increase exponentially. But data is of no use unless you can do something with it. The role of business intelligence in extracting insights from information is therefore only going to grow.
It’s likely to become increasingly tied to trends such as big data and cloud computing. Research company Gartner sees demand for BI as already having reached a tipping point and predicts that the BI market will grow by more than five percent in 2016 to reach a worldwide value of almost $17 billion. Ian Bertram, managing vice president at Gartner says, "Organisations must transition to easy-to-use, fast and agile modern BI platforms to create business value from deeper insights into diverse data sources."
Businesses will also increasingly demand analysis in real time, delays in processing data and producing reports will become unacceptable. BI software will therefore need to become more responsive and take advantage of predictive analysis techniques to deliver faster results without sacrificing accuracy.