Tech choices loom large on MDM strategy, despite process focus

Despite the heavy focus on people, policy and process in MDM programs, there are still many technological challenges that teams must deal with.

Master data management isn't necessarily a technology-heavy initiative. For example, Gartner Inc. analyst Ted Friedman says a typical MDM project team spends about 85% of its time on people, policy and process issues and just 15% on the technology side of a deployment.

But there are IT choices to be made in planning an MDM strategy -- and lots of tools to choose from, including a mix of products from vendors of data management suites and MDM-only software providers. There are offerings that target specific data domains, primarily customer and product information; on the other hand, companies can buy multidomain MDM software for use across different domains. In addition, tools tuned to the needs of individual industries, such as financial services and pharmaceuticals, are available for the taking.

One of the primary considerations in evaluating and selecting software is deciding how to structure the underlying MDM architecture for storing master data and distributing it to systems throughout an organization.

A centralized MDM hub moves master data off of source systems, which then retrieve the data from the hub after it has been consolidated, cleansed and matched to remove errors and inconsistencies. Another option is a registry-style hub, which handles data cleansing and matching but creates an index pointing to the master data in source systems instead of storing the information itself. Hybrid approaches combine the two styles, maintaining reference links to master data entities in source systems but also serving as the primary source of the data for new applications.

Sidestepping the politics of an MDM program

Consultants say the registry approach might be preferable when there are a large number of source systems spread around a company; it can avoid some of the MDM political wrangling that is bound to result from efforts to relocate master data to a centralized hub and overwrite it in the source systems. A centralized architecture also tends to take much longer to implement, but it offers advantages when the quality of master data is questionable in source systems or an organization wants to have a single, authoritative repository for all of the data.

More on MDM strategy

Read about the connection between enterprise MDM and big data

Learn how the business capability model can help get funding for MDM

Find out about the risks of choosing the wrong MDM software

Ultimately, there's no right or wrong way to go -- the choice depends on the goals and requirements of an MDM program, according to Anne Buff, a thought leader on software vendor SAS Institute Inc.'s customer best-practices advisory team. "Don't get stuck on one concept or another, because it will attract you to specific vendors and you'll be sold on terminology rather than business need," she said.

In addition to the hub, other technology components that have roles to play in MDM projects include data integration tools for moving master and reference data between systems and data quality software for automating the process of cleansing and consolidating the information.

Depending on an organization's overall IT infrastructure, the data integration layer can encompass anything from traditional extract, transform and load tools to messaging middleware and complex event-processing platforms. On the data quality side, there typically is a need for data cleansing tools and a specialized matching engine for synchronizing data elements, such as customer names and addresses. Data deduplication and merge/purge capabilities are also common checklist items.

Too many tools in the MDM box?

But an organization implementing an MDM strategy might be able to utilize existing data quality and integration technologies instead of investing in separate tools, said Rick Sherman, founder of consultancy Athena IT Solutions in Maynard, Mass. "Too many times, companies have integration tool X in-house and then buy another one for MDM," Sherman said. "And then they're stuck with two data integration tools when they only needed one."

The process-oriented nature of MDM programs does raise the question of whether a hub system and related technology are really necessary to get the job done. While automating MDM processes isn't an absolute requirement, Friedman sees it as a must-do item in almost all cases. "In theory, you can do MDM without technology," he said. "But in practice, given the scale and scope of these moving parts, it's naïve to think you can be successful without it."

Aaron Zornes, chief research officer at The MDM Institute consultancy in Burlingame, Calif., said the accumulated evidence in the historical record of MDM deployments points to a similar conclusion. "Can we do MDM without technology? The answer is yes, because we've been doing it that way for 30 or 40 years," he said. "We just haven't been very good at it."

Beth Stackpole is a freelance writer who has been covering the intersection of technology and business for more than 25 years.

Email us at editor@searchdatamanagement.com and follow us on Twitter: @sDataManagement.

This was first published in August 2013

Dig deeper on MDM best practices

Pro+

Features

Enjoy the benefits of Pro+ membership, learn more and join.

Related Discussions

Beth Stackpole asks:

Is your organization using commercial tools to help automate the management of its MDM program?

0  Responses So Far

Join the Discussion

2 comments

Oldest 

Forgot Password?

No problem! Submit your e-mail address below. We'll send you an email containing your password.

Your password has been sent to:

SearchBusinessAnalytics

SearchAWS

SearchContentManagement

SearchOracle

SearchSAP

SearchSOA

SearchSQLServer

Close