The logic is simple: business intelligence software is employed to make companies perform better. Therefore, the vendors say that the products – if used well – naturally pay for themselves. However, backing up these claims and getting management to buy into that logic is more of a challenge.
Because a BI implementation may have numerous effects across an organisation calculating a financial return can be difficult, says Richard Kellett, head of intelligent architecture at BI vendor SAS.
"Because it can be such a huge catch-all, one needs to focus on identifying the costs and savings associated with achieving specific results." Vendors are under increasing pressure to provide hard numbers to businesses wary of profligate IT investment.
This has been reflected in buyer behaviour, says Rob Ashe, CEO of BI and corporate performance management software company Cognos. "The new IT buyer, the new business buyer, is not buying infrastructure. They're not buying feel-good stuff; they're buying tangible results," says Ashe. "There's a fundamental ‘show me the hard ROI' calculation."
But there are obstacles to getting that message across within the enterprise, says Godfrey Sullivan, CEO of Hyperion. "BI increases efficiency, but you can't go into the finance department and say ‘this'll save you 10 financial managers' – they don't want to hear that."
According to SAS's Kellett there are three main considerations in determining the return of a BI investment.
"Companies want to know how quickly they will recover from the initial investment and generate positive cash flow," he says. "They want to know the internal rate of return [i.e. for every £1 spent how much does the business generate]. And, lastly, they want to ascertain, at the actual time of spend, what its net profit or loss is going to be."
While it is possible to demonstrate that BI tools meet functional requirements, producing bottom line figures for improvement needs some lateral thinking. Sometimes these can be estimated: Cognos claims if it improves the spare parts tracking process for a car manufacturer by reducing downtime by 1% it can save the company between $5 million and $10 million annually for an initial capital investment of around $1.5 million.
BI has also been used successfully in many call centres, where businesses can quickly see the impact on sales volumes. By presenting call centre operators with vital information about the customer they are better able to identify sales opportunities – and act upon those. By monitoring the volume of sales before and after a BI investment it is possible to measure and quantify a return.
Analysts at IT industry advisor Gartner certainly believe that, in most cases, BI can be shown to be a sound buy.
"Many BI technologies are proven and represent generally low-risk investments," its analysts report. Additional benefits are derived from BI investment as more operational systems, such as enterprise resource planning and customer relationship management systems, feed into it, says Gartner's Frank Buytendijk. He estimates that companies that take this kind of holistic approach to implementations will generate two to three times greater ROI than those without.
The problem facing IT management is that much of this analysis is retrospective. They can point to examples where BI has delivered organisational benefits, but that is a far cry from producing a business case that will convince board members to invest.
To achieve that, IT decision-makers need to understand where the BI project is intended to add value, and spell out what measurements will be made to assess improvements in performance.
But any executive charged with delivering a successful BI project needs to be wary of how the process is approached, warns Keith Gile, principal analyst at analyst group Forrester Research. Failures are frequently a result of solving the ‘problem' from the wrong side, he warns.
According to Forrester's Gile a common mistake is to assume that more data will result in better intelligence. According to Forrester's estimates, typically 15% of any organisation produces data for the other 85% to consume.
However, instead of assessing what information is really needed, companies get locked into the idea of producing more and more data. "BI is still seen and marketed as a producer tool. Companies need to be able to consume results and make better decisions – to be able to use BI without even realising it. Consumption rather than production should be the focus," says Gile. Organisations do not need more data, he says, they simply need to work out how to derive the most benefit from the data they already have.
To deliver a successful BI project that delivers demonstrable bottom-line results, CIOs need to look carefully at which parts of the organisation need what types of analysis and reports. This requires a considered approach to presentation of data – call centre staff require entirely different information and intelligence, for example, than that demanded by C-level executives.
The solution to this problem has been the corporate dashboard. Individual users are presented with, and have access to, reporting tools as befits their position within the organisation.
This is a radical change from the way in which most companies have historically used BI: often it has been a tactical tool for discrete projects, such as campaign management, or improving customer satisfaction. Convincing senior executives that BI should be more than that will require a track record of success.