Content Copyright © 2007 Bloor. All Rights Reserved.
After IBM’s Barcelona conference on MDM (master data management) last year, I criticised the company for getting its MDM messaging wrong. At the company’s Information on Demand conference in Anaheim last October the company presented a new pitch, which was much improved. However, IBM has recently been running an Information on Demand roadshow in various cities across the United States at which I have been presenting on the status of the data warehousing market, while various IBM representatives have been presenting on the company’s balanced warehouse, its Information Server and its MDM offerings. Based on the presentations on MDM that I saw, which all consisted of the same presentations but were presented by different people, I still don’t think the company has got its messaging right around MDM.
IBM’s main marketing message around MDM is that it is, uniquely, “multiform”. This is an excellent message but it doesn’t go far enough and is improperly understood. Let me explain.
Multiform MDM, as explained by IBM, is that it is multi-usage and multi-domain. In my view it also ought to be multi-deployment but I’ll come back to that. Let me start with multi-domain: what IBM means by this is that it has solutions for customers, suppliers, products, inventory, contracts or whatever. This is true, it does, although there are separate products for inventory and products as opposed to customers and suppliers. So, here the multi-domain message is valid. More importantly, my spies tell me that IBM is working on a common platform that will support all domains and, once this is announced, there needs to be an additional emphasis on cross-domain capability. However, it is not with its multi-domain message that I have an issue with IBM.
Where I do have an issue is with multi-usage. What IBM means by this is that its software supports collaborative, operational and analytical MDM. Whilst I would tend to talk about synchronisation as opposed to collaborative usage I have no quarrel with this three-fold definition. Or, at least, I wouldn’t if IBM hadn’t redefined “analytical” to mean something rather different.
To most people analytical MDM means the ability to support analytics by means of multiple hierarchical views across data sets that are kept consistent by the MDM software. This, for example, is what Microsoft has said that it is going to use its acquisition of Stratature for. IBM, on the other hand, at least in so far as its presentations are concerned, refers to the direct provision of analytics, specifically with respect to EAS (entity analytic solutions) and GNR (global name recognition, formerly LAS). Now, these are both great products but it is arguable that they are more about data quality than analytics in the first place and, even if they are considered to be in the analytic category then they are purely that, and are not about supporting analytic MDM more generally. I think part of the problem with EAS and GNR is that IBM didn’t know quite where to put them and so they were lumped in with MDM, which meant that the company had to come up with a marketing message that provided a raison d’être for putting these together with MDM.
Finally, I don’t think the multi-form message goes far enough. I think it should also include multiple deployment styles. For example, if you want true analytical MDM and nothing else, all you really need to support this is a lightweight registry with some federated capability behind it. If you want consistency across data stores (synchronisation) then you need a somewhat more sophisticated repository to store master records in (again with federated capability) but you still don’t need the full-blown hub you would require to support composite applications.
My understanding of the IBM software is that it does support registry and repository type deployments as well as hubs. But it doesn’t talk about it. And then it fudges the analytic issue when it is perfectly capable of supporting it. It seems unlikely but I think IBM is underselling its MDM solutions: its message could be a lot clearer and a lot stronger.