07.05.2013

BI projects regularly trigger headaches

Business Intelligence (BI) projects are plagued with significant problems for companies. They are even described as more pronounced compared to 2007 in a comparative study by Actinium Consulting. The difficulties are often also related to the fact that a large number of companies recognize wrong decisions in tool selection. According to the current survey of more than 200 BI users, almost two-thirds of the respondents state that they have found previous BI measures more difficult than average. 44 percent even describe the degree of difficulty as “very high” and another 37 percent as “high”. Only one in five companies has so far experienced business intelligence measures largely free of unusual requirements and, according to their own statements, can even look back on comparatively easy project conditions. In 2007, there were still 12 percent more in total who found the implementation of BI projects easy.

Actinium managing director Klaus Hüttl suspects that unchanged strategic weaknesses are the central cause of the multiple project difficulties. “They usually exhibit great complexity. The potential pitfalls are correspondingly diverse, because functional, technical and organizational conditions have to be brought together,” emphasizes the consultant. If a strategic goal is not defined, the result is inadequate planning because the necessary basic orientation and detailed analyses are inevitably lacking. This assessment is also confirmed by the survey results. According to the survey, two-thirds of the users state in retrospect that inadequate requirements analyses were among the biggest mistakes in project implementation. Six years ago, the figure was significantly lower at 56 percent. Process analyses also failed to meet requirements in a similar number of cases.

To prevent such problems, Hüttl recommends that companies submit their draft strategy to an experienced BI expert, who can point out possible weaknesses and optimization potential with his neutral view. This also applies to the selection of BI tools as another problem area. Whereas in 2007, 46 percent self-critically admitted that, in retrospect, they had not made an adequate system selection, dissatisfaction with tool decisions has currently risen by one-fifth to 57 percent. “Instead of deriving the selection criteria from the BI strategy, the opposite approach is still frequently taken, in that the strategy is tailored to the selected tools,” Hüttl problematizes. However, the starting consideration should always be to avoid a technically dominated BI orientation and instead derive the performance profile for the tools from a strategic overall view. “Otherwise, the tools create considerable shackles, prevent the potential benefits from unfolding, and instead limit the ROI of business intelligence investments or even create expensive late effects,” he judges.

Hüttl, however, does not see BI providers free of responsibility in this regard either, which is why they need to be viewed through a critical lens, he says. “One of the main causes for frequent wrong decisions in the selection of BI products is also that the manufacturers often do not put the practice in the foreground, but lure with interesting features and other arguments,” criticizes Hüttl. Many companies would then find clear discrepancies between their needs and the possibilities of the BI software during project planning.