It is not often that I talk about vendors here – you know the trouble that can cause (!) but I had to mention that this week Oracle announced an OEM with Silver Creek Systems (see Oracle Improves Its MDM Product Data Offering With Oracle Product Data Quality Cleansing and Matching Server) This is not a vendor plug – it is a blog to call out what should be obvious. MDM (the discipline) requires a hefty dose of data quality capability; how else are you to physically clean the data? MDM (again, the discipline) is, in some peoples view, a data quality initiative. I would have to agree – up to a point. MDM does focus on data quality, but MDM really uses data quality tools to assure the integrity of a process: the process by which mater data is created and maintained across the business.
Oracle has an MDM offering but had not, until this OEM relationship, really addressed itself how users should “solve” the complex data quality side of product data. MDM of customer data and MDM of product data have their differences; the former focuses more on name/address and account cleansing and relationships, much more than the latter. The latter focused, historically, more on workflow and user involvement in creating the data, much more so than the former.
For MDM of customer data, data quality was intrinsically part of the technology since that was central to achieving “single view” of customer from the start. For MDM of product data, this was not the case. However, users that struggled to clean up their product data prior to being used in what they called MDM systems had been using or applying data quality tools to the problem themselves. The choice had been: apply any old data quality tool to the problem and spend time developing a real solution, or find a data quality tool that had already been “applied” to this problem. Silver Creek Systems is an example of what I call “applied DQ” specifically for product data. Some of the DQ tools (and MDM of Customer Data solutions) offer or support “applied DQ” specifically for the customer data domain, even if they never said that.
Conceptually the move – to align (applied) data quality with an MDM solution, to help make the MDM discipline more effective, is a good one. Hopefully users can take advantage of the “one stop shop” that Oracle is now offering. Time will tell.
PS “applied DQ” does not mean that any other DQ tool cannot help with MDM of Product Data, or indeed any other DQ aspect of MDM. It just means that someone (a vendor, typically) has taken the time themselves to “apply” their tools to the problem and in so doing, has learnt something from the effort. The implication is that the “applied DQ” tool is able to achieve the same that any other DQ tool can do, but perhaps faster or cheaper; or perhaps can even achieve some stuff that the generalists can’t do at a cost you can afford.