(1. The Technology Trap)
by Ian Rowlands
I know it will be shocking to people who know me as the Product Manager for a market leading metadata management technology that I should be discussing this topic. It’s been bugging me for a long time though. I think there are great metadata management technologies available – but for some reason, there are not many long-term successful metadata management implementations. So I’ve put my bullet-proof vest on. I suspect I’m going to take some shots, or at least have to duck some rocks, but I’m going to open the conversation. I’ll probably have several “chapters” to write … I think there are issues of technology, process, culture and economics to consider. I’m going to start with the technology.
It would be pointless for me to claim that the metadata management platforms are perfect. In truth, as a survey the field, I could produce a list of improvements that could (and should, and eventually will) be made, even to the “best in class” products. Feel free to provide your list in comments to this blog. At the same time, some organizations seem to get good value for money, and to be very happy with what they’ve got. That makes me think that it’s not just a matter of waiting for silver bullet technology. There has to be something else – and there is. In fact there are quite a few “something else’s”. I’m going to address them over several posts.
I want to start with the software purchase. The enterprise software purchasing process is broken! Briefly, most of the people we deal with are buying features – bells, whistles, and widgets – and not capabilities. Many business drive purchases of enterprise metadata technology through a “request for proposal” process. It is rare indeed for such a document to define business objectives and specify use cases. All too often, there is an assumption that a defined set of features will support use cases and business objectives that are never clearly articulated.
It’s not just what is purchased that causes the problem. It’s also how the process is operated. Some of you who know will me will know I’ve been in the software business for more than 30 years. When I started, buying enterprise software was a relationship intensive process, a dance in which each party got to know the other very well before a deal could be consummated. Now, the process is much more “hands-off”. It is not unusual for a buyer to research on the internet, issue an RFP, and select BEFORE getting to know the vendor. There’s an interesting transition going on – we may be in a “post solution-selling era” where highly informed organizations are making complex purchases without needing vendors to educate them … but as yet I haven’t seen a model that replaces the value of relationships in the metadata management arena.
The action item is pretty simple. Make sure you know what you are trying to achieve – in detail. Share your aspirations with vendors. Come to an agreement about what is reasonable, and then be prepared to insist on getting what you pay for. Recognize that investing in a metadata management infrastructure should be a long-term proposition – so invest in the vendor relationship. If you want to say that there’s nothing special about metadata management in this respect, I won’t argue. But getting the technology acquisition process right is one step in establishing sustainable metadata management.
First, I think there is a lot of confusion about what we are trying to manage. Are we trying to manage the business or are we trying to manage the data.
If the business, then the business needs a common language, a taxonomy.
If the data, then the solution needs to be connected to the physical assets. The business taxonomy can be connected to the real data assets but it needs to be considered as a distinct issue.
“The business” and IT are practiced at talking past each other. The business tends to talk about ideal concepts, “oughtness”. This is what the data “ought to mean.” While IT talks in terms what the data does mean “is-ness.” There has to be real colaboration in this area and I don’t see it happening much. The business just wants IT to solve the problem and metadata feels too much like documentation to grab much IT focus – especially if IT perceives that it will not be used.
Second, the task is overwhelming. Have you ever told a five year old to clean up his room because it is a disaster? The poor guy looks at the mess and can’t figure out where to start, so he moves a few things around and starts playing with something. Now imagine that it is siblings who share a room. We are doing an adult version of that.
We talk about capturing definitions for entities and attributes. But, someone needs to take a collection of attributes and write a white paper about how they relate: Standard price, discount price, promotion price, manager discount, bundled price, actual selling price. What are the business rules and proceses that govern all those terms. How many varieties of cost are there? Now where can I find that data?
These are issues that transcend any tool.
Improve your google rank today check this out! http://goo.gl/YzNKa
Thus, our shelves land up filled with issues that we experience, but that individuals don¡¯t have got time meant for.