Getting Started with Business Intelligence

Getting a Business Intelligence (BI) program off the ground at any organization can be a daunting challenge, particularly when the Information Technology department is driving the train.

Many of our clients seek to understand their products, customers and organization at increasing levels of granularity, yet they do not have the right people, tools or processes in place to do so successfully.

Over the past several years, Centric has advised many clients on establishing a competency for BI.

Here are the top seven actions we advise our clients, typically CIOs, to take:

1.  Find The Right Partner. IT should not be driving BI, the business should. IT needs to seek out a business partner who has organizational clout and who is willing to put their voice forward to advance business intelligence across the organization. Make sure the business partner has interests that are aligned with corporate strategies. Use this as one of the prioritization criteria as you evaluate the BI work on the table.

2.  Start Small, End Big. Large, monolithic data warehouse implementations are a process of the past. Most of these large efforts fail anyway. Instead, we advise our clients to use an Agile BI approach that leverages business-driven user stories to guide implementation. The time it takes to see value in this type of approach is measured in weeks, not years.

3.  Prioritize, Deliver, Prioritize. Look to establish a set of user stories (small functional sets of scope that lead to specific business decisions) and prioritize them based on their business case. It is also important to evaluate IT’s readiness in implementing identified user stories as some systems are easier to access and translate data from than others. Take the highest priority stories and implement in agile fashion using 2-4 week iterations. Once you have delivered the first few user stories, continue the process of prioritizing and implementing.

4.  Maximize the Features Not Built.  Build only those features that specifically support user stories in the iteration. The development team must be disciplined about handling the minimum required data and analytics. Data warehousing specialists often, in the name of efficiency, load data in which they believe users will have some future interest. This leads to data modeling, ETL development, testing and UAT that could otherwise have been allocated to a prioritized user story. It is imperative that this “gold plating” of the BI solution be prevented or the delivery pace of the team will eventually grind to a halt.

5.  Build Foundations That Bend, Not Break. Throughout the implementation process, the BI architecture needs to be flexible and adapt to the larger scope of business intelligence that is supported. Ensure that the proper amount of information modeling is being done as part of the normal cadence of the agile delivery process. Engage BI service offering leadership in project audits to ensure data modeling and architecture will not be an inhibitor to project success.

6.  Don’t Leave Adoption to Chance. Adoption of a BI solution is measured by how many users actively rely on the system to make decisions. However, it is frequently ignored by managers who believe adoption will occur naturally due to the superiority of the information or richness of the tool. In fact, the adoption must be thought of as an initiative that employs: organizational change management, internal marketing, training and supervision. Breaking old “data habits” of users is achievable with proper incentives and structure; it will otherwise remain elusive.

7.  Self Sustain. It is important as consultants that we proactively work to get our clients in a mode where they can sustain themselves without our assistance. If your consulting team truly has your best interests in mind, they will create plans to help you achieve this goal sooner rather than later. Acting with this mindset creates an unmatched customer experience the Centric team works so hard to achieve for every client.