SMARTER STRATEGIES

Five Tips for IT Managers: Boosting the Effectiveness of Your Business Intelligence Solutions

Dennis McCafferty | Date: 11-05-10 | Comments

Business intelligence projects can put an organization on the fast track to growth and increased profitability. But there are steps�and missteps�that often determine whether the efforts are a success or failure.

In the book "The New Era of Enterprise Business Intelligence: Using Analytics to Achieve a Global Competitive Advantage" (IBM Press), author Mike Biere provides detailed guidance for IT managers who are looking to launch a business intelligence (BI) solution for their enterprise. Biere has specialized in data warehousing and BI global solutions for IBM as a tech software specialist, manager and consultant. Here are five critical steps from Biere to consider for a successful BI initiative:

1. Consider SAAS options. They'll avoid the internal effort to "unbuckle and rebuild" from the inside, and allow external experts to take over. This approach works well for IT managers who realize that a complete BI makeover is needed, especially if the organization itself doesn't have a unified vision as to what BI tools need to do. If the project were maintained internally, such a lack of vision will often lead to constant experimenting, failure and re-experimenting. An outside party with proven experience in SAAS BI tools and services can provide a fresh, grounded perspective.

2. Centralize the effort. At least one person within your team�if not you�needs to be designated as the BI guru, the one in the organization who runs the BI Competency Center (BICC). If you have multiple BI deployments in place without such a hub, then the system will be set up for potential failure. No one will be accountable for making sure that the correct tools are being accessed by the right teams and that those results can be effectively measured for value delivery. This function can also ensure there aren't two BI systems being used to essentially do the same thing. Those with this responsibility must keep in mind that the end users may or may not have a high degree of tech skills�and that those without the IT pedigree must be trained to use the tools effectively.

3. Ensure the data is good. A great BI solution applied to bad data presents no value. The data must be clean and accurate�and timely. The mechanisms must already be in place for the data to be updated rapidly before the BI solution can effectively exploit it.

4. Identify the key end-user "influencers" within your organization. You can't build a BI system that pleases every end user. But you can implement one that the most influential end users like. Early in the process, engage them to get a sense of their needs and expectations.

5. Assess costs/returns. It's difficult to get a sense of the value of BI solutions without knowing how much you're investing into them. Beyond testing the BI solution throughout various stages of implementation, a post-launch assessment should take into consideration the number of users, deployment patterns, associated costs, business value and ROI. You also need to get a handle on nonquantitative costs, such as human skills invested. In the end, a product that was considered of lower value may come out on top because it serves more enterprise users at a lower cost.