The master data management (MDM) market is still barely a teenager in absolute terms, but is showing signs of growing up. After some gentle exploration in the early years accompanied by the odd tantrum, the MDM software market is showing a number of indications of maturing.
To begin with, the merger and acquisition activity in the sector has entered a second phase. Initially the large vendors, spotting a fast-growing new area, went on a spending spree to make sure they didn’t miss out. IBM bought DWL and Trigo, SAP bought A2i, and Oracle used its Siebel acquisition to deploy the latter’s UCM product as an MDM offering. The interaction vendors then got in on the act, with Tibco buying Velosel, Informatica acquiring Siperian and Software AG buying Data Foundations. Recently there has been a second phase of consolidation. Informatica is acquiring Heiler and in June SAP bought Hybris, who have specialised in e-commerce but also have an MDM product offering.
Another development has been the appearance of new MDM products from a variety of sources. Recent MDM technologies have been brought to market by Pitney Bowes Software, its modern offering based on a graph database rather than a relational platform. Dell Boomi has introduced a cloud-based MDM solution, and a startup called Enterprise Technology Solutions has just launched a cloud MDM product aimed at SMEs.
All this vendor activity reflects a change in the way that customers are deploying MDM solutions. Early MDM projects were typically point solutions, usually focused around either customer or product data. It began to dawn on IT architects that these two data domains were not the only ones that needed attention, and that MDM could be applied to other data domains like asset, location and person, and to assorted types of financial data. Newer “multi-domain” MDM technologies evolved to meet this requirement, and existing technologies were swiftly rebadged to meet the new demand. It was also realised that ‘reference data’ such as country codes, currency codes and industry classification codes were essentially a subset of enterprise master data, and also needed to be consistent and managed proactively.
Gradually enterprises have begun to connect the dots of their various MDM initiatives. Deploying a series of separate master data hubs, possibly using different technologies, may achieve local benefits but risks putting in place a new series of data silos, which themselves need to be connected in order to realise the full benefit of consistent master data across the organisation. I have recently seen more ambitious MDM projects, involving an architecture that tries to bring together multiple data domains and geographies, broadening the scope of MDM projects from particular business lines or geographies to enterprise-wide initiatives.
Since such broad initiatives are inevitably multi-year projects due to their sheer scope, it is vital to develop a robust business case and to get business leaders genuinely engaged. Relatively recently there has been a rise in data governance initiatives, which include master data and data quality within their scope, and are usually either business-led or at least driven jointly by business and IT. These initiatives are a way to draw business people into the process, assigning ownership of data where it belongs – in the business as an asset and not in IT.
The vendor community has been slow to react to this, bolting on data steward modules to their products and adding a few dashboards, where what is needed is more far-reaching. In particular the broader MDM initiatives will require managed communication between multiple MDM hubs, possibly from more than one vendor, in order to join the dots. Yet the major vendors seem to have barely recognised this requirement, preferring to try to grab market share and fill in gaps in their technology, rather than providing solutions that can assist in the controlled management of federated hubs of master data. Many of the larger vendors are simply bogged down in digesting their various acquisitions.
As customers get to grips with their enterprise-wide MDM deployments, they need to push vendors to better support their real needs. Left to their own devices, vendors will latch on to whatever seems like the next trend, perhaps social media sentiment analysis using Big Data. I would like to see them doing a better job of managing the data we have now.