InetSoft Webinar: Create a Shared BI Support Organization

This is the continuation of thetranscript of a Webinar hosted by InetSoft on the topic of "Analytics and Agile BI" The speaker is Abhishek Gupta, product manager at InetSoft.

The other end of the extreme is where larger organizations create a shared BI support organization. That has negative connotations as well, or negative implications. Large centralized organizations inherently are bureaucratic with multiple levels of signoffs, and different conflicting priorities.

Things start moving slower so as a result business users who can’t wait, remember in the age of the customer, customers rule, and customers don’t wait. So if customers don’t wait, your business users, your business analyst are going to go back to their home grown applications.

Very often unfortunately we see that when you are already centralizing your BI support organization, you actually become less agile, and you proliferate the silos. So we offer some kind of a middle of the road approach where you get the best from both worlds. Absolutely empower individual contributors with self service BI tools so that they can produce up to 80% of their own BI content.

You definitely need to co-locate business and technology staff. That’s why off shoring and outsourcing only works very for repeatable, commoditized tasks like software maintenance, maybe communication writing, maybe regression testing etc., but not for this agile development and deployment.

We recommend that you do not have the same teams that support areas such as ETL and data warehouses co-locate with the same team that supports data usage such as creating reports and dashboards. Separate them. Centralize your data prep teams. Empower IT to own those layers of the stack. But decompose and then federate BI usage support, and empower Business Users allow them to be independent.

Do not try to instill the same rules and policies and principles for your customer facing and non-customer facing BI applications. These have completely different requirements and tolerances for risk and accuracy and latency. So if you are a large global business, definitely try to centralize your financial and HR analytics.

Those requirements are pretty similar in your head office, but try to decentralize and federate your customer facing BI support for your customer facing BI apps such as sales and marketing. So basically think in terms of this hub-and-spoke environment for BI support which really has to mimic your business organizational structure.

So if your finance and HR are centralized, by all means centralize BI support for these organizations, but if your marketing is completely federated between North America and Asia Pacific because you market and sell different products to different customer segments then trying to create centralized or shared BI support for those organizations will just produce impediments.

The third pillar of Agile BI is all about a monitoring and learning what business people do. And as I mentioned earlier in the discussion business people tend to say politically correct things at meetings but then they go back, and they follow that path of least resistance.

View a 2-minute demonstration of InetSoft's easy, agile, and robust BI software.

They do what they need to do today to get their job done so monitoring their environments and creating this three tiered BI environment is definitely one of the best practices that we more and more have noticed is really making enterprises very successful. Empower your business users with their own sandboxes and when you create sandboxes literally think in terms of spreadsheets.