by John Ladley
Part of the EDJ mission is research. This includes surveys as to who is doing what and whether or not they are having any success. Our survey for SOA was sent to our Best Practices Council, a growing body of technology users and beneficiaries. Below is our detailed presentation of the findings, and a recap of responses. However, for the time constrained reader, I’ll present a summary.
The SOA efforts surveyed are predominantly technology projects. We see scant examples of using SOA to change businesses. Most of the efforts are applications oriented and justified as experiments or to address focused areas vis-a-vis holistic enterprise architecture efforts.
Given this is the data strategy journal, we were keen to see evidence of data services, but alas, there was not much evidence of overarching development of services related to data management. While services certainly handle data, we saw little evidence of governed design and deployment of an SOA data strategy.
Our survey also contained some free-form comments. These revealed few, if any, best practices. There were respondents with wide-spread deployment of SOA, as well as a fairly robust application if data management functions; there just weren’t that many. The feeling we get is akin to looking at the vast numbers of shops deploying data warehousing in the mid to late 90’s; flashes of brilliance, but no strong success trend. For example, there was no correlation with the size of team and success. This is consistent with other tech areas in the past. Although respondents with 2 – 4 EA types did not report in production and none were full steam
Our summary findings therefore for SOA can be summed up simply – there is a long way to go. SOA is too focused on one off applications, which frankly, as our contributing authors reveal, is not the smartest way to do this. The intensity of application development without an associated use of data management of governance means one thing – we will be deploying SOA applications that, in the long run, unless methods change, we will just create more of the same data log jams we have currently.
Most shops are working on it. Most are “working on it” but a lot are Not there – only 17% are running, and upon examination we found only 1 shop out of 20 that was running SOA to have a robust enterprise view with associated data governance.
When you add up the categories in terms of applications centric vs. data usage centric, you see System Flexibility and Reuse as the top two benefits being sought. Consistency and Integration come in third. From this view it seems that SOA is applications centric with fewer benefits being sought from data service related or data usage.
Also of note, the leading benefits are soft – uniform in their distribution but little in the way of hard benefits. Since we did not ask a specific hard dollar return question, we made some calls. Our suspicions were confirmed – not a lot of hard dollar measurement going on.
While this question looks promising, our optimism faded a tad when we read some free form comments and correlated this question with others. Yes DM groups are involved. However, less than 20% are involved in every project. This means 80% of projects in some organizations are allowed to skip the use of data management. Worse, it reinforces our perception that too many SOA projects are focused in narrow areas, perhaps by function, or application. And, of course, when you add “None” and “A Little”, you get 38% of our respondents declaring that DM is not involved much at all.
This is the bellwether question in terms of data strategy and SOA. Even when there is a data service plan in place, it is a work in process. No respondents reported wide use. (We double checked with the ones who did not know what we meant by data services, which was… SOA services that are designed as general service for data – vs. specific to an application)
Good news here – data management is pitching in by pre-defining structures and supporting meta data. The strangely low number (3%) for BI integration is a puzzle – usually one of the first data management areas to receive technology benefits is business intelligence. Again, it appears SOA is pointed more at applications and operational functions.
We also asked for a list of tools. Several respondents felt this was confidential information, so we do not have a large sample. Those that did report listed the following vendors. We have not screened these to determine if, for example, the analyst firms feel the list below represents SOA vendors. It is just what we were told. They are not categorized.
TIBCO BusinessWorks, TIBCO EMS and TIBCO Hawk. TIBCO BusinessStudio Datastorm, Casewise, and Agilense, Hybernate, SAP NetWeaver, NET; Biztalk, CA ERWin. Enterprise modeling using Telelogic System Architect.
One notable response for this question ….“Our organization finds it more exciting to build things by hand.” (The reader can supply there own thoughts to this)
There is some governance during deployment. This is encouraging. (Now can we get some while we are planning SOA?)
ABOUT THE AUTHOR
John Ladley is an internationally known information management practitioner and a popular speaker on information and knowledge management. John is widely published and has several regular columns. Until recently, John was a Director with Navigant Consulting. Prior to Navigant, John founded KI Solutions, and John was Senior Program Director of Data Warehouse strategies and a Research Fellow at Meta Group. Mr. Ladley is an authority on information architectures, business performance measurement architectures, knowledge management, collaborative applications, and information resource management. John is currently President of IMCue Solutions, a new firm focused on data governance and information management.