Despite recent news that master data management has gone mainstream and ranks among the fastest-growing software markets, MDM is not for the faint of heart or the weak of wallet. It's a time-intensive practice that can cost a cool million-and that's according to 2007 estimates.
So, it's easy to see where Jessie Chimni is coming from when he writes an "an evolutionary approach designed to show clear, incremental value" might be a more appealing way to introduce MDM into cost-wary enterprises.
Granted, Chimni has some interest in persuading you to proceed with MDM. He is the VP of North America services for Bristlecone Inc., a supply chain business and technology consulting firm, that offers, among other things, help migrating to SAP's MDM catalog.
Even so, he makes an interesting and unexpected case for using problems with the spend analysis process as a means of introducing MDM into tightwad companies.
Chimni contends spend analsyis is a manageable MDM pilot project. It doesn't involve too many teams, and yet, the data-complexity problems frequently seen in spend analysis are exactly what MDM was designed to solve. He writes:
While the data may come from multiple catalogs or business applications, this is a very contained, finite set of data that can be readily cleansed by automated tools and subject matter experts. Once the initial cleansing and framework is in place, ongoing data management becomes simply a maintenance issue.
As such, it's a low-hanging fruit for MDM, which will translate into a high-visibility success story for IT, according to Chimni.
Plus, it doesn't hurt that fixing spend analysis first will give you an important future ally-the chief procurement officer.
As a bonus enticement, the article walks you through the three phases of a successful spend analysis campaign.
I suspect the big trap with MDM-as with ERP and many other big IT initiatives - will be proving ROI. Granted, there are a lot of statistical justifications for MDM, but you're not implementing in a statistical enterprise. And statistics don't always translate into real savings for specific deployments.
ROI can be tricky with customer data integration, anyway. And, basically, most MDM amounts to CDI. In fact, The Data Warehouse Institute recognizes MDM as a form of customer data integration in its recent report on CDI. (Although, it should also be noted that some MDM products have their roots in product data integration).
According to the TDWI, the problem that continues to plague CDI is demonstrating the ROI. As TDWI put it:
The catch is that ROI is indirect. For instance, up-to-date customer data leads to more efficient customer service, which yields higher customer satisfaction, such that customers churn less. Since CDI is the first link in the chain (and revenue is the last), it's hard to link CDI to revenue. But the link is there.
I suspect the same will hold true for MDM, particularly where it's being used for master data related to customers.
So, it's worth noting that Chimni reports one company managed to reduce its overall purchasing spend by about 5 percent after using MDM to address spend analysis problems:
After finishing the spend analysis initiative, the company was able to reduce its rate of purchase of similar commodities from multiple suppliers at varying contractual prices. The CPO also eliminated maverick buying. After clearing up price variance issues - a prevalent problem throughout the company - the organization reported significant savings, for example a $42,000 annual savings on the purchase of just one item.
That's a good first step on an ROI.