Case study
Fighting the flood of data
This case highlights the continuing importance of business intelligence software. It looks at some of the drivers for adoption and possible technical problems with introducing it. It does not confront the people-related issues of managing change when such systems are introduced.
This issue is addressed as one of the questions that follows.
In the face of an economic downturn, one might have expected companies to cut back on IT projects across the board. Spending on business intelligence tools, however, remains healthy. According to Forrester Research, the analyst group, as many as 44 per cent of companies are considering investing in business intelligence software. This puts growth in business intelligence ahead of that in customer relationship management (CRM), enterprise resource planning (ERP) and integration technologies.
Business intelligence software helps companies analyse and manage information across their operations, whether it is sales data produced by an electronic point of sale (EPOS) system, information from distributors, statistics from a website or traffic from a call centre.
As IT departments improve the connections between business systems, the amount of data reaching managers can increase dramatically.
Areas such as supply chain management, ERP and production systems are showing a trend towards real-time data gathering. This holds out the potential for far quicker and more accurate decision-making. But there is also a danger: managers risk being overwhelmed by the sheer volume of information they receive. Rather than better decision-making, the result can be executive paralysis.
‘It can happen that instead of reacting to the information they receive, managers are swamped', says Roy Bernard, strategy director of The Aspect Group, a consultancy that builds information systems. The best way to avoid this, Mr Bernard suggests, is to start with smaller projects, extending their scope as managers become more familiar with the technology. Often, though, companies introduce business intelligence systems without thinking first about the questions they want to answer. They are not helped by the sheer quantity of information that is available.
The number of analysis tools on offer sometimes makes matters worse.
Adopting a single set of business intelligence tools is a good start. Forrester Research found that more than 80 per cent of companies operate multiple business intelligence tools. This can lead to inconsistencies and ‘several versions of the truth'. Instead of making decisions, managers spend time arguing about which set of data to use.
Yet a centralised approach can be equally unhelpful when business intelligence specialists fail to communicate with the managers who ultimately carry out decisions.
‘Companies have fallen down in the past by using too wide a range of tools', says David Metcalfe, Forrester's analyst for business intelligence in Europe. ‘But we also have situations where the statistical whizz-kids use data-mining techniques to find interesting mathematical issues to solve, but which bear no resemblance to the problems the business managers face. That is a huge disconnect.'
First, then, a company needs to make sure that its business intelligence tools are attempting to answer the right questions. This is not only the case where business intelligence is handled by a dedicated team. Systems such as ‘executive dashboards', which display data in one place for senior managers, are little use unless they provide the right information. A business will not become more responsive just because managers have more data at their fingertips. The reverse can even be true.
‘Companies gather information about their customers and accumulate massive amounts of data without thinking about what it enables them to do', cautions Mike Lucas, technology manager at systems integrator Compuware. ‘They can become flooded with information, and that clogs up innovative decision-making.'
And even the right information will not fulfil its promise unless it is put into context. Raw sales data is more meaningful when set alongside information about current marketing campaigns; and data on repair times achieved by a field service force makes more sense if managers know the age of the appliances they are fixing.
Questions
1 What are the typical problems in introducing business intelligence tools that are described in the case study?
2 What additional people-related problems do you think may arise when business intelligence systems are introduced for end-users? Assess these problems, which are not directly considered in the article, based on the concepts of managing change introduced earlier in this chapter.
In the past, companies would set up major data warehousing projects just to gather the raw numbers for analysis. Such projects are both costly and time-consuming. Instead, the latest generation of business intelligence tools draws information directly from a company's IT systems, then presents it in a consistent, usually web-based, format.
Direct feeds of data from operational systems such as customer ordering software or ERP are much more likely to be up-to-date and accurate. This is important, because data cleaning is expensive. Forrester estimates that as much as 60 per cent of the cost of a business intelligence project lies in data cleaning.
The need for a large, up-front investment in data warehousing is also reduced if companies focus business intelligence projects on finding answers to specific questions, such as ‘How can we cut purchasing costs?' or ‘How can we persuade existing customers to buy more?'. Where business intelligence projects are well defined and managed, the rewards can be significant.
Since introducing software from Crystal Decisions, Virgin Retail, the operator of the UK's Virgin Megastore chain, believes it has saved time by removing the need for manual reporting. And improvements in purchasing, it says, now allow it to manage stock levels more efficiently.
Ottakar's, the bookshop chain, uses software from Cognos to gather information from across its network; staff can download reports onto any computer in the business.
‘As it is standard, web-based technology, the training was minimal', says Richard Wakeham, IT director. ‘Staff are used to using web-based systems.' The company also points to the benefits of better stocking decisions and improved information about its promotions. Ottakar's plans to extend access to reports to its tills when it next upgrades its point-of-sale system.
In these cases, then, a clear goal better information flows between branches and head office led to improved stock levels.
‘At Ottakar's, the business intelligence system gathers information from the bottom up', says Roy Bernard of The Aspect Group, which designed the system for the retailer. ‘The system has been successful because it shares information from the grassroots.'
Source: Fighting the flood of data, Stephen Pritchard, FT.com; 7 May 2003.