By now, everyone knows master data management is expensive-and by "expensive," I mean as in millions of dollars. But can it actually save you money by reducing IT's costs? Mervyn Mooi, director of BI and data warehousing consulting firm Knowledge Integration Dynamics, seems to think so.
In a recent IT Web article, he outlines seven ways MDM can help lower IT costs. At the top of his list is cutting integration costs, which he says MDM does by reducing point-to-point integration requirements. He also says it can reduce the cost of data cleansing, reduce software fees and hardware costs, and cut how much you spend on delivering information.
I can't help but notice, however, that he doesn't say it will reduce IT costs enough to actually pay for the MDM implementation. He does at least caution that MDM isn't a silver bullet, and lists important steps IT should take and questions to ask before moving ahead with MDM.
Speaking of MDM, last week I found a short, but intriguing piece on Information Management about the implications of semantic technology for MDM. The gist is that semantic Web technologies could be used to add meaning to the data, helping MDM filter out redundant customer data, for instance. The writer seemed to think it also could speed up and simplify integration. The article's not what I would call "essential reading," but if you're curious about semantic technologies or extremely involved with MDM, check it out.
Seth Grimes also wrote an excellent piece on semantic search for Intelligent Enterprise that's worth checking out, though it's not specifically related to integration.
For those of you less interested in cutting-edge tech and more interested in business strategy, JP Morgenthal, aka the Tech Evangelist blogger, published a nice write-up this week called, "The Busy Executive's SOA Reference Guide." It's short and sweet, and nicely cuts through much of the pedantic crap that tends to surround SOA. I got a kick out of his basic definition for figuring out what's a service and what's not a service:
If the business itself would be less scalable or cease operation because the service is not available, then you have a service. If the mulching service doesn't show up when the landscaper is there, the landscaper cannot complete their task.
Business readers might also want to check out my recent interview with Ilan Sehayek, the CTO of the open source data integration company, Jitterbit. Sehayek says business analysts are handling more integration work, and he believes this trend will grow as budget constraints force IT cutbacks. Coincidentally, eWEEK published a review of Jitterbit 3.0, this week, including a lab that shows you, via a slide show, how the data integration suite works.
Finally, there were quite a few announcements from integration companies this week.
MuleSoft released Mule MQ, which the press release describes as "an enterprise-class JMS (Java Message Service) implementation for high performance, low-latency messaging." It's designed for speed-with a low latency and high throughput. It works with Mule's open source ESB, but it can also stand alone as a messaging platform.
Also this week, Boomi announced a new partnership with Coupa Software, which offers online purchasing software. Boomi sells a SaaS solution that can integrate any combination of SaaS, cloud and on-premise applications. This announcement is of interest in part because Boomi traditionally targets the SaaS consumer, but with this move, it's bundling its integration capabilities and making those available with Coupa's solution. It's also interesting if you're a Coupa customer or interested in that type of online service, because Coupa users can link their e-Procurement applications with on-site ERP or accounts payable systems, via Boomi's connectors.