Recently I interviewed Bob Howard, an instructor with business-improvement education and consulting firm Oliver Wight, about a new class the company's offering on master data integrity.

I asked about integration, and to be perfectly honest, there was a bit of confusion about what that meant. I meant technology, and Oliver Wight meant cross-team integration.

Howard figured out the miscommunication and warned me from the get-go that he was not a technology expert. In fact, he hasn't even experimented with a vendor's master data management product.

Still, he agreed to chat with me about master data, and I'm glad he did, because IT leaders need to hear what he had to say about master data as a business initiative.

You can read the full interview here, but here are the lessons I learned that I feel are particularly relevant for IT.

Master data is an ongoing discipline. It's not a project you start, finish and hand over to the business. In fact, master data has a lot in common with a quality initiative. Everybody needs to be on the same page and following an established set of practices to keep the data clean and useful. Howard said this is where most master data programs ultimately fail.

Master data requires a cross-functional team. "It's not just an IT project," Howard said. IT has a role, certainly, which he outlined. However, the actual specifications, such as ensuring the data's quality, must be owned by the appropriate business units.

Master data initiatives can clue you into situations where the technology is failing the business unit. Curiously, many master data initiatives don't start in the executive suite or even with the IT division -- at least in manufacturing. Often, according to Howard, business users push for master data because they have "an inability to extract value from the ERP system tool." The data isn't correct, and therefore, their planning isn't based on accurate information. The business units start to experience problems with their key performance indicators, such as on-time delivery. They have problems staying on budget.

IT has a lot to offer master data initiatives. Howard outlined the high-level keys to success with master data, and guess what -- it's all stuff IT's heard over and over again about how to ensure a project succeeds and meets the needs of the business:

• Make sure senior management signs off on it.
• Pick a low-hanging fruit target area.
• Involve all the stakeholders; use methodologies.
• Use governance to institutionalize the change.
• Perform audits.
• Publicize your wins.

You know this stuff. Now you can take your hard-learned lessons and help guide the business.

That said, there are unique issues with master data. As this article, published on b-eye network points out, master data should not be confused with "just another data integration" project.