Business Intelligence: Make Your Data Work for The Organisation


28 January 2009 Clive Margolis


Clive Margolis, consultant director or Acestar Solutions, explains how managing business intelligence can raise value in any organisation.


Ever feel like your data is just a burden on the organisation, eating up costs but not really giving much back in return? You’re not alone. That’s why more and more organisations are discovering the benefits of a well-implemented business intelligence strategy. Datamonitor’s recent Business Intelligence in Retail Banking report predicts business intelligence spend by retail banking in North America, Europe, the Middle East and Asia-Pacific will increase from $5.6bn in 2006 to $9bn by 2012, an increase of around 60.7%. Business intelligence (BI) is one of the fastest-growing areas of IT.

Business Intelligence is a system

There are no end of self-certified consultants, business analysts and 'techies' all too willing to help you spend your budget on BI, but you can never do justice to that outlay until you understand what it is all for. Then you can start to devise strategies to not just implement BI successfully but prolong that success well into the future.

BI differs from systems that might handle, for example, manufacturing, distribution or finances, as developments within both the organisation and the software industry can make it desirable to add to, or improve on, the outputs you started with. These systems should have enough flexibility to allow this natural development. It should be remembered that these are desirable changes, in other words they are only implemented because the anticipated benefits outweigh the costs.

As BI is an IT concept, it is implemented using software tools, and there are a growing number of companies, large and small, that provide these, such as IBM (Cognos), Oracle, Microsoft and Business Objects.

Collect

"BI is always best achieved by purpose-built BI software and specialists designed to do the job efficiently and effectively."

While all organisations have databases, which at first may appear to dilute the need to collect information, it may well be that the data you need is missing. For example, having several years' worth of data on sales transactions and client names and addresses is no help at all if what you are monitoring is absenteeism in the workforce. You may well have an HR database, but it may not necessarily record the data you need to track absenteeism.

Missing data is just one issue. Even if you have the information, there may be good reasons for the BI system to store it somewhere else. These might be performance reasons, data management reasons, or data processing reasons.

The most popular way to collect the data for BI to process is in a data mart or data warehouse. These are databases designed to organise the data so that it can be quickly and processed for BI purposes. A good data warehouse design is resilient, and can be added without having to make major alterations to the design. This means that as new business reporting requirements come along, a good data mart or data warehouse can cope.

Analyse

In the good old days of computing, reports were printed out a line at a time by onto fanfold computer paper, often printing several copies at a time. The line printers were, and still are in some cases, kept in air-conditioned computer rooms and required weekly maintenance contracts. Business managers would expect to see a huge pile of paper, maybe 12in-18in thick, on their desks (or on the floor keeping the door open) whenever the report was due, and it was not uncommon for these reports to be delivered on a wheeled trolley as they were so heavy. Of course, no one could possibly read such a document!

Now we can get much more information from our data by summarising it in ways that means something to our organisation. For example, margins, year-on-year variance and gross sales are all different measures used to summarise data and measure performance. You can add a lot of value to data by making use of powerful KPIs such as profit per customer and revenue per employee.

But why bother with BI? Why not just get someone to go to the computer system and draw up a report showing these KPI’s? The simple answer is, you probably couldn’t. Even if you had all the information you need, you would probably find crunching it into the right shape simply too taxing on your existing systems. Much of the information would need to be summarised at various levels which, for a large organisation, usually needs to be done overnight.

BI is always best achieved by purpose-built BI software and specialists designed to do the job efficiently and effectively. Development costs can be controlled by good planning, focused priorities and, above all, phased implementation.

Present business information

Presentation is a key issue. Whereas the main method of storing data (in relational databases) has been around since the 1970s, presentation methods and the software packages that deliver them have come along in leaps and bounds, and are constantly evolving. These include:

  • Online and printed reports and queries
  • Graphs
  • Three-dimensional cubes (for drilling up/down on data)
  • Dashboards
  • Scorecards

The presentation of BI metrics is getting constantly improving. All reporting can be online and alerts emailed to you if a key performance figure falls below a preset level.

Supporting business decisions.

There has to be a commercial reason for collecting all this information, because whether you are a public service, private enterprise, or a charitable organization, chances are you have a limited budget.

And whether you are looking at a sales analysis, year-on-year margins or your market position, you are probably gathering this information in order to support the decision-making process.

After all, if you don't know whether your organisation is improving its market share or getting worse, and in which region, or product group, or sales team, how can you make an informed decision on which changes need to be implemented?

That just about sums it up. Why are we collecting, analysing and presenting this information? In order to make better-informed business decisions. Everything has to come together in the right way.

Get a good BI strategy

One last thing: it pays to outline a good BI strategy. In a recent press release, Gartner identified the lack of a BI strategy as the single biggest flaw in BI implementations. Everyone makes mistakes. But some mistakes are more costly than others. And all IT projects, including BI projects, are prone to errors that can be very costly to correct. If a mistakes is going to be made, then the earlier the better. The more advanced a project, the greater the cost in correcting mistakes. Get yourself a good strategy and you’re well on the way to success.

Also, not taking action can be similarly costly in terms of the lost opportunity for benefits to your organisation. BI systems allow organisations to maximise the use of their data by creating performance indicators and presenting them effectively. A well-implemented BI system can cut costs, improve productivity and make an organisation more competitive. An effective BI strategy is vital to success.