Master Data Management (“MDM”) has been billed as a panacea for data quality and data integration issues over the last few years, but many companies have failed to see the results they expected through their technology investments. So is MDM just another red or blue pill peddled to solve information management ills? The reality is that most organizations have undertaken MDM initiatives have focused on the first two letters (MASTER DATA) that imply a technology solution, and sacrificed the third letter which is the real critical success factor: MANAGEMENT!
When we look at a managed process from a Capability Maturity Model (“CMM”) perspective, we are looking at a very high-level of maturity that necessitates a defined, repeatable, and MEASURED process. Thus, to truly realize Master Data Management, a company must implement the oversight necessary to support, measure, and ascertain the success of their efforts and then take appropriate corrective action when necessary.
All of this so far sounds rather ivory tower so let me provide an illustration of what would need to be done to MANAGE a Customer Data Integration (“CDI”) solution . . .
A few years ago, I was architecting a Master Data Management (“MDM”) solution for a consumer focused company that was going through a mainframe retirement program. The replacement was a Service Oriented Architecture (“SOA”) with MDM domains covering product, customer, and pricing. The company was very much caught up in using new technologies but did not always appreciate how to optimize the current state business processes and its technologies to deliver a cohesive result. Just slamming in the technology was not going to address data quality or integration. Instead, the solution which we advocated was to re-engineer the business process, monitor business process impact on data integrity, and establish accountability within the organization.
The main challenge with the CDI solution was that the best identity match algorithms could only achieve a limited confidence level. Without the intervention of a data steward and often additional information from the customer, the goal of a single view of a customer would not be achieved. With over six million accounts, it was not feasible for the client to have dedicated data stewards who examined near matches and decided whether to rationalize or not. Instead the solution was to re-engineer the customer service process so that a Customer Service Representative (“CSR”) could make the appropriate inquiries whether to collapse multiple customer identities.
For instance, if the CDI solution showed that a customer might be located at multiple addresses, through business process automation the CSR could be prompted to ask “Mr. Smith, my records show that you also have an account with us at 123 Main St. is that correct?” By making a slight change in the business process, we improved our data quality and integration for the cost of 10 additional seconds on a phone call. The ROI on this investment of time is quickly justified when downstream uses of this information are considered: If a duplicative quarterly customer mailing is eliminated, the 30 cents of additional time on the call would yield $2 in reduced cost in the first year.
All is well? Not quite! As Ben Franklin observed, “Diligence overcomes difficulties, sloth makes them.” So, unless we had a verification process in place, how could we know whether the CSR was actually doing his or her job? This is where the Management of Master Data comes into play: We established metrics which allowed us to monitor the success of the refined business process. As part of the design process of the Master Data Solution, Key Performance Indicators (“KPI’s”) were identified to monitor the business process impact upon the general health of the data. Progress could be tracked and corrective action taken.
Or can it? This brings us to my next point: That unless responsibility and accountability was established in the business, the investment in a master data solution could be rendered worthless. KPI’s may be helpful in identifying the reticent CSR who refused to follow the script, but who was going to be the enforcer . . . Information Technology? During the design of the master data solution a RACI diagram was established for each business process that impacted the master data. For our embedded rationalization process, the CSR would be Responsible; the CSR Manager, Accountable; Information Technology, Consultative; and the Executive Sponsor, ultimately Informed. By delineating roles and responsibilities we had a clear elevation path when deviations from the process were detected and a path for remediation could be taken.
Without properly establishing ownership for Master Data within and across business domains, a technology organization is setting itself as the scapegoat for another project that over promised, and under-delivered. To succeed in managing master data, a partnership with the business must be established that embeds stewardship into the business process, monitors of successes and failures through KPI’s, and enforces accountability/responsibility for results. Getting started with MDM doesn’t require a high level of process accountability and maturity across the organization but it does require finding the appropriate initial points in the business process where the management of data yields the highest returns. We will explore the best ways to get started with MDM and how to gradually develop these management practices in a future post. A final point – having a crystal clear articulation (and quantification) of the business value “at play” through superior MDM will drive the business and technology partnership and maintain organizational support!