Thanks to inputs from more than a dozen BI experts at QlikTech, during the last couple of weeks I’ve written several articles about potholes in the road to delivering BI projects. (List is below.) This article is the last in the series, and highlights a couple of final potholes having to do with the right stuff: good data and good analytic tools.
Thanks to inputs from more than a dozen BI experts at QlikTech, during the last couple of weeks I’ve written several articles about potholes in the road to delivering BI projects. (List is below.) This article is the last in the series, and highlights a couple of final potholes having to do with the right stuff: good data and good analytic tools.
Very often, we see organizations face issues with:
- Data consistency. BI projects can take major damage when they hit the data consistency pothole. A company might have on record eight different variations of the name of the same partner, for example. Or one business unit might count an entity as a customer at the end of the first month that entity makes a payment while another business unit defines an entity as a customer the minute the first order is placed, before payment is received.
- BI software that is inflexible and hard to use. Organizations that want to be nimble need motivated and adaptable people. Those people, in turn, need flexible, friendly tools. In many organizations, though, even when people are motivated to move fast, the software tools they have at their disposal have a long, slow learning curve that inhibits their ability to be nimble.
For example, the ETL (extract, transform, and load) process – just one step in a traditional BI project – can be an enormous hurdle. And many BI platforms consist of multiple layers in a stack, each with its own technical skill set requirements. Different people often look after each layer of the stack (e.g., data warehouse, ETL, and report writing) and it is incumbent upon them to communicate effectively.
It is important that the app developer and business requestor validate data by comparing the source to the destination and comparing various occurrences of the same metric within an analytic app and across multiple apps. Part of IT’s role is to ensure that the data IT provides to business users for consumption in analytic apps is prepared, cleansed, and governed – yet granular enough to provide value. And in this age of empowerment, analytic software that is flexible and easy to use should top the priorities list for any BI project – these are two primary characteristics that can make or break user acceptance and adoption.
(This is the fourth and final article in the “Potholes of BI” series featuring insights from the following BI experts at QlikTech: Chaitanya Avasarala, Miguel Angel Baeyens, Gary Beach, David P. Braune, Greg Brooks, Annette Jonker, John Linehan, Brad Peterman, Olaf Rasenberg, Mike Saliter, Chris Sault, Matthew Stephen, Christof Schwarz, and Mark Wine.)
See these related posts:
• “Lined Up and Ready to Go: Episode 1 in the ‘Potholes of BI’ Series”
• “The Elephant and the Cheetah: Episode 2 in the ‘Potholes of BI’ Series”
• “The Wrong Trousers: Episode 3 in the ‘Potholes of BI’ Series”