In the shadow of Big Data, MDM players continue to innovate

News

Our colleagues from SearchDatamanagement.com point out that MDM, although it is suffering from the marketing tidal wave of Big Data, is nonetheless an innovative sector, capable of responding to the real problems of companies. And in particular that of the management of data repositories, essential to the proper functioning of companies.

One might think that Master Data Management (MDM) is no longer the main concern of publishers of data management tools. The concept is not the sexiest with marketing departments, and after being relatively trendy for the past 5 years, the "hype" influx that these products could generate has moved to the field of Big Data and virtualization. data. To this is also added another thing: the fact that MDM remains a vague concept for many companies, which damages its image a little more. "There is a lot of misunderstanding around the concept of MDM", says Clive Longbottom, founder of the British analyst firm Quocirca. “Most companies look at MDM and think it's just about cleaning up their customer data to have a single data repository. This is however not the case. It's about focusing on what matters to the business of the business." "For example, Longbottom points out, customer data may not be the primary master data for a business. This can be the type of object sold, the suppliers or more generally, a combination of different factors - which then requires having one or more data repositories.”Aaron Zornes, director of research at the MDM Institute in the United States , explains, meanwhile, that "MDM is very useful to support companies towards processes that they traditionally had trouble implementing". For example, establishing a cross-selling process between different product lines, between different departments, or having a unified view of customers, or even meeting legal compliance constraints, he adds. According to him, the concept of MDM is also more concrete than Big Data. “I get angry every time I hear that term. It's not about pressing a single button and ending up in the age of Big Data. The growing volume of data is a phenomenon that has existed for several years." Today, MDM is extending its ramifications to embrace other concepts such as BPM (Business Process Management), data integration, data quality and - of course - Big Data. And although MDM has somewhat disappeared from radar screens, it remains a technology borrowed from innovation. MDM: a world that innovates According to Aaron Zornes, many technological advances have been made with MDM, such as identity resolution example, which relies on powerful algorithms capable of verifying whether a person is who they claim to be - a key function in risk management and fraud prevention. "Among the very recent algorithms, some very successful ones manage to determine who is who, by a simple indirect association - for example, via a telephone number or the analysis of marital ties."Through a series of takeovers, IBM succeeded in assembling data resolution functions. Just like Informatica or Tibco, which have made it one of their priorities, comments Aaron Zornes. from publishers. As for example Oracle with the acquisition in January 2010 of Silver Creek Systems, specialized in the quality of product data. But the strongest innovation around MDM is that which aims to analyze the mass of unstructured data generated on the Web. Aaron Zornes explains, “Companies believe they have a 360° view of their customer base just by purchasing customer data from Dun & Bradstreet. This view is actually only 75 degrees. If you want to see the big picture, you also need to sort through the mass of information, present online, on Facebook, Twitter and LinkedIn. MDM tools can also help aggregate data in text format from traditional sources - storing structured data - within a unified master data repository. And that has a definite impact on business. "Being able to inject information like customers' hobbies, which football team they support, for example, provides real value for sales reps who want to be close to their customers," he says.MDM systems are equipped with in-memory technology to offer real-time dynamic data analysis functions, he adds. SAP, for example, places its MDM technology above its Hana in-memory processing engine. "We are going to witness the arrival of real-time MDM, which is certainly expensive, but it will become the norm." Another example is that of Microsoft, which plans to integrate free MDM tools into the next version of Office, name Office 15 code. Which, according to Aaron Zornes, should contribute to opening up the MDM market a little more to SMEs. meet their MDM needs. Program Director Christoph Balduck says that while the group appreciates the value brought by mega-players in the field of data quality, he believes that certain innovations are today in the hands of niche players. For example, we are looking closely at Talend, which has a particularly efficient Open Source offer in data migration,” emphasizes Christoph Balduck. He also says he is impressed by what he has seen at Orchestra Networks: “They are thinking about semantic modeling, which I believe is the future of MDM. It is also one of the only vendors to offer cloud-based MDM.”Orchestra's DataSpace version control feature, which allows users to simulate the live effects of a data change masters, is of particular interest to Balduck. "We can not only control the impact on your data model, your content and attributes, but also how it changes, or not, your workflow and your business rules, in particular," he explains. “If the simulation does not generate incidents, the modifications can be applied and put into production. This ability, that of being so flexible, is very value-creating.” However, there is still a long way to go: Aaron Zornes notes that technology providers have been very slow to address the issues of data governance in particular. Only a small number have laid their groundwork in this segment, and few are those who have really engulfed it. And there still remains the problem of integrating the different master data sets, developed within each company. “There is currently no other solution than to buy an additional hub,” says Aaron Zornes. "Vendors talk about 'über-hub', but currently no one is able to integrate all applications at a desirable level for enterprises".

To learn more about Decision-making and analytical tools