Thoughts on DQ Consolidation
David Loshin blogs his thoughts on the Pitney Bowes acquisition of FirstLogic. We examine his assumptions and conclusions, and add another factor in the equation: Google.
He chronicles some of the recent history of consolidation in the space:
With regards to the latter, I'm not sure that the viability of these companies was ever in question. Certainly Ascential, Trillium and FirstLogic had nearly - or already, depending on your criteria - achieved critical mass. No, I think this move was strategic in every case as larger companies seeking to establish a foothold in the customer data lifecycle processes deep inside corporate IT departments moved on these companies while they were still small, and therefore affordable.
Something else may be a factor as well. There's another predator lurking in the tall grass these days, and he's hiring. His name is Google. Just look at the skill sets they are seeking:
Looks kind of like the same skill set required for advanced Data Quality, doesn't it? That giant sucking sound you hear is all the programming talent being vacuumed up by the G. Even companies like IBM and Pitney Bowes, who have serious bench strength in their engineering groups, have been feeling the pinch in these specialty areas. I have several friends in these and related fields, and the salaries have been climbing rapidly - life is good. Just a suspicion, but perhaps one to keep in mind as you do your due diligence with vendors on the DQ portion of CDI projects.
He chronicles some of the recent history of consolidation in the space:
Trillium, which is owned by information provider Harte-Hanks, recently purchased data profiling tool vendor Avellino. Ascential, full on a dinner of Metagenix, Vality, and Torrent, found itself on IBM's dinner plate. Evoke, which had sold its remaining assets to Conversion Services about a year ago, just was passed along to European Data Quality company Similarity Systems. SAS bought DataFlux a few years back.David concludes that this trend reflects two market determinations:
The first thing is that data quality technology has matured to the point where it is deemed critical to the information movement process... The second thing might suggest that data quality tools vendors are not able to establish a healthy market standing on their own two feet, or else there might not be such a preponderance of acquisitions.The former is clearly true, as company after company embarking on CDI implementation has discovered that the data cleanup and data loading processes can be large and time-consuming projects in and of themselves. And an ongoing DQ process is critical in operational CDI to ensure that such problems don't rematerialize later.
With regards to the latter, I'm not sure that the viability of these companies was ever in question. Certainly Ascential, Trillium and FirstLogic had nearly - or already, depending on your criteria - achieved critical mass. No, I think this move was strategic in every case as larger companies seeking to establish a foothold in the customer data lifecycle processes deep inside corporate IT departments moved on these companies while they were still small, and therefore affordable.
Something else may be a factor as well. There's another predator lurking in the tall grass these days, and he's hiring. His name is Google. Just look at the skill sets they are seeking:
|
Looks kind of like the same skill set required for advanced Data Quality, doesn't it? That giant sucking sound you hear is all the programming talent being vacuumed up by the G. Even companies like IBM and Pitney Bowes, who have serious bench strength in their engineering groups, have been feeling the pinch in these specialty areas. I have several friends in these and related fields, and the salaries have been climbing rapidly - life is good. Just a suspicion, but perhaps one to keep in mind as you do your due diligence with vendors on the DQ portion of CDI projects.
2 Comments:
John:
You have made some good points, although I still think that data quality tools vendors are still focused at selling data quality tools, instead of seeing how a data quality strategy integrates with the rest of an organization's information/knowledge strategy. Firstlogic is trying to break out of that mold, and Ascential's acquisitions demonstrate an understanding of this, although I am curious about how their customers view the integration of the various purchased tools.
On the other hand, while Google's list of desired skills may subsume those required for data quality, that laundry list could fit any of a number of knowledge discovery/knowledge exploitation tasks. Yet, there is a great need for metadata, semantic analysis, and quality assurance for content accessible via the web, and cataloguing it will require some intellectual input. And so far, Google's approaches have been fresh and innovative - hey, maybe I'll go polish up my own resume ;-)
David,
You’re not going to get any argument from me. A thought on the possible explanation is the best I can do. I think part of it has to do with one of the basic dilemmas of being a solutions vendor. The broader the scope of your proposed solution, the fewer the clients for whom it will be a fit. As a pure data quality tools provider, FirstLogic or Ascential can say, “If you’re interested in data quality, you are going to need a data cleansing tool. We offer the best data cleansing tools because of …a, b, c.”
The moment you move the functional performance specs up the chain toward source data (data architecture, knowledge management, etc.), or down the chain toward the end users (process management, integration architecture), you are going to start ruling out potential clients due to incompatibility with existing systems. The answer, of course, is in SOA design for the application framework. Unfortunately, most modern applications have not been designed with SOA principals in mind, so that means replatforming the application, not a small task. Nevertheless, it is one many vendors are undertaking. In the mean time, they are at least ensuring that common services can be exposed as web services, allowing construction of SOA logic in the design phase of IT implementation.
I’d be interested to know what you see “data quality strategy” encompassing.
Post a Comment
<< Home