Master Data Management (MDM) is changing to reflect some of the more influential technologies in the Data Management space today, which includes Big Data, graph databases, Cloud Computing, social media, mobile, and others.
Collectively, these technologies reflect the overall trend of MDM’s evolution from specific domains and lines of business, to enterprise wide integration. The ideal outcome of this transformation is a single copy of data that can support disparate uses.
Individually, these technologies are shaping various strategies for the deployment of MDM, which are largely based on those technologies themselves.
Graph databases are increasingly finding their way into Master Data Management solutions from a variety of angles. In some ways, graph databases are better suited as the databases for certain MDM domains (such as those for customers) because they take a semantic approach to the relationships and classifications of data elements, which traditional relational models cannot duplicate. Forrester noted that the influence of graphs on MDM solutions includes vendors utilizing graph databases as central repositories, organizations using graphs as databases in their own internally developed MDM hubs, and vendors adding features in which companies can make graphs of their data. According to the research company, the utility of this approach is that: “It immediately changes the mindset and strategy of MDM from systems to views” and is “much more intuitive, analytic, and intelligent about… master data.”
Although utilizing graph databases for MDM options may expose issues of scalability and a skills shortage in semantics to operate these solutions, the benefits of addressing these situations (by utilizing additional data stores or retaining/hiring qualified personnel) are demonstrated in the more “human” views of data these options offer.
Big Data’s impact on MDM platforms is evinced in a multitude of ways, the most important of which might be in the incorporation of sentiment analysis with customer and product data. There are several different ways of facilitating this aggregation of data, which effectively provides a comprehensive view of a customer or a product. There are many CRM for social media applications which are frequently used by line of business users to interact with customers through popular social media sites. Augmenting this data with Master Data can help these users to understand a customer’s entire history or interaction with the organization—rather than just his or her social media history with it.
Additionally, integration options between Big Data initiatives for line of business purposes and MDM repositories can provide a comprehensive overview of a customer’s history with an organization in close to real time, which includes whatever sentiment data or social media interaction that has recently taken place. Myriad advanced analytics options can leverage this composite data to influence immediate or near-future purchasing habits. According to Gartner, the incorporation of sentiment analysis and social media interactions with Master Data ultimately benefits the respective applications and MDM hubs by providing context to the former and “the obvious example of MDM having the opportunity to incur a reciprocal benefit by adding social for CRM data to its data sources for inclusion in the 360 degree view of the customer” (O’Kane and Sussin, 2014).
The Cloud’s impact on Master Data Management is two-fold. On the one hand, it provides another environment (in addition to on-premise) with which to integrate sources from; for example, several popular CRM solutions are provisioned through the Cloud. Additionally, there are numerous vendors that offer MDM products hosted through the Cloud either as SaaS, PaaS, or Iaas offerings. Adoption rates for these products—especially when compared to on-premise versions—are relatively low, and are due to a number of inhibitors, including security concerns, about hosting proprietary organizational data outside of enterprise firewalls and ineffective governance that Cloud access would exacerbate.
However, many of these concerns can be mitigated by the deployment of private Clouds and there is no mistaking the chief benefit of cost reductions that these options provide. According to Gartner, competitive on-premise Master Data Management solutions can require upwards of $1 million dollars of capital costs (which does not include additional hardware and software costs for integration with existing systems), whereas the subscription rates for Cloud offerings can be obtained for approximately $100,000 of what can be listed as operational expenses (O’Kane and Thoo, 2014). This option is attractive for small to mid-size business and those looking to test MDM products before shifting to the full scale on-premise variety, as well as those looking to exploit the flexible architecture and infrastructure requirements of the Cloud.
Some of the key facets of integration that MDM is increasingly being used to account for are addressed by multi-domain MDM solutions, which can manage both customer and product data; some platforms are designed for particular industries and use cases. In certain instances, however, the tendency towards specialization limits the overall efficacy of such an option for more than one domain in addition to its utility for various functions. According to Gartner:
“…for customer data you most often need a data quality capability such as entity resolution. For product data the more common data quality capability you need is semantic and/or text string parsing…Thus specialization leads to a solution fitted for one data domain and/or industry that is less competitive in another.”
Vendors will need to address these points of inconsistency for multi-domain solutions to provide the sort of integrative support that is increasingly required of Master Data Management with Big Data, Cloud and on-premise sources. Competitive vendors offer platforms compatible with a suite of products to address these different needs.
MDM’s relationship to data integration is somewhat paradoxical. By definition, an MDM hub does not include all of an enterprise’s data—merely those which are essential to carrying out specific, critical business processes. Simultaneously, it is of extreme benefit to integrate as many sources as possible to help organizations with those functions, or at least as many sources as are necessary to inform those processes. The integration capabilities of MDM become all the more necessary when involving Big Data sources. Vendors are increasingly addressing the centralized capacity of MDM to accommodate a variety of sources in a way that can even automate critical business functions. Marketers, for instance, can make specific pitches to customers on social media based on their latest updates in close to real time by utilizing data integration and event triggering features of MDM. Gartner noted:
“MDM software vendors commonly provide a low-level data integration service layer and a facility to construct composite business transactions from those services…these composition facilities can be made to include calls to business applications…the result can be deployed to customer service representatives, often in a single user interface that includes the 360-degree view of the customer” (O’Kane and Sussin, 2014).
Developments in this aspect of MDM will include greater functionality in the event-triggering mechanisms, delivering the capability to respond to more complex events.
Elimination of Silos
These contemporary developments in Master Data Management are all focused on reducing silo cultures and processes that, when properly implemented, MDM can eradicate. Not all of these effects are direct; some of the essential boons of MDM in a more integrated Data Management landscape are implied. Utilizing Big Data with MDM ensures that an organization’s most valuable data are represented in a uniform fashion to account for external unstructured or semi-structured data. The incorporation of Cloud sources or Cloud MDM forces (prudent) organizations to structure their governance rules and roles accordingly. Graphs can expand how data is stored and provide enhanced ways to conceptualize and utilize data. The overall effect is that organizations can incorporate more sources in a reliable fashion that can expand the utility of their data across multiple lines of business.
O’Kane, B., Sussin, J. (2014). “What master data management leaders need to know about social for CRM.” www.gartner.com
O’Kane, B., Thoo, E. (2014). “The impact of cloud-based master data management solutions.” www.gartner.com