The proper implementation of business intelligence and analytics can help bring success to a company. But it’s not hard to take just one wrong misstep that can be costly.
As Gartner found a few years ago that between 70 and 80 percent of corporate business intelligence projects fail, lessons need to be learned.
Let’s take a look at some common ways companies fail at business intelligence (and perhaps add a little bit of advice along the way):
- Fail to Set Goals and Requirements
Before a company even considers what BI platform will be used, its leadership must decide on what KPIs are important and what they want the data to reveal. Data alone will not solve any problems. Context must be given and it must be presented in a manner that eases the analysis that needs to be performed. (Think data visualization with dashboards and a multitude of sunburst, heat map and other styles of visual reports – with animation.)
- Use Bad Data
Data governance can be a prickly subject for the people who set up the various databases, Excel spreadsheets and other data sources in which a company’s data has been stored. A BI solution can’t provide users with actionable data if the data is incorrect, incomplete, outdated or inaccessible. Either the data sources must be adjusted, or when the data is pulled it must be “cleaned.” One database may include a different set of measurements, making a straight comparison impossible. Failure to recognize these types of problems can ruin the integrity of what the BI platform produces. With trust in the data being lost, users will not make decisions based on the numbers. Standards must be set for the data, and data sources must be compared to determine what will skew reports and how to fix them.
- Executives Fail to Champion BI
Before any initiative in a company is adopted someone or some group needs to be its biggest fan. If executives are not sold on this business intelligence platform, user adoption will be close to zero. BI needs to be part of a normal day’s business, not something extra that users might get around to trying out whenever they catch up on “real work.”
- No Steps Taken to Assure Buy In by End Users
Even if the CEO and the rest of the C-Suite love the idea of using a BI platform, it doesn’t matter if the right end users aren’t empowered to use it. Restricting BI to just the C-level executives, or even including the managers, will put a crimp in the benefits and ROI. The staff members who work with data on a daily basis must have access to the BI tools – and they must be empowered to make business decisions on actionable data. Otherwise, what’s the point of implementing BI? Users need to have an understanding of their data either by training in data literacy, or having an easy way to identify the database fields on which they are reporting.
- Information Lags in Dashboards & Visualizations
The right data needs to be available at the right time, in as near to real time as is possible. Those days of waiting on quarterly or annual reports better be over at every business, or competitors will easily drive them out of business. Every time that an end user calls up data via the BI platform, it must include the latest information available. The BI platform should not have to go through data warehouse ETL to access data.
- It Takes a Team of Analysts and IT Staff to Operate
For every time another person or department has to “touch” the data to provide answers, a little bit more of the edge a company has on its competition is lost. If end users have to consult with IT to design every report and also with data analysts to decide what data is important, the process will fail due to its complexity and slow response time. End users need to be able to dive deep into the data when they find a new line of inquiry that no one foresaw.
The BI platform ideally will be embedded within the main business application end users utilize daily. It will not directly act on the data, but will only access it to present it for analysis. To be self-service, its controls should be intuitive or at least as easy to use as possible, following the simplicity of functions such as the Google Search tool.
- Failing to Train End Users on the BI Tool
No matter how easy an embedded, self-service BI platform may be to use, implementation of that solution will fail if end users are not properly trained to use it. They will be unlikely to know how deep they can dive into the data through reports and dashboards, or that they can collaborate in their business intelligence efforts through the BI platform.