Tilting at Windmills
Quite a long break since my last architectural blog entry. It's not that there's nothing to say - only lack of time to write it down. I started writing this post about different integration worker characteristics as a continuation to my earlier post on architects. I sketched different shades of grey on paper but somehow ended up on writing about couple of organizational characteristics in integration space from organizational perspective. Maybe I'll come back to my original thoughts later on.
There surely are differences in a way of doing IT operations between organizations. I'll try to outline some stereotypes based on my experience in the enterprise business and maybe aggravate a tiny bit to underline my point.
Some want to make it big and outsource everything. However heavy externalization without IT headache is typically not possible. When huge system upgrades are bought with "just do it, we don't care how you do it" mentality I think it's a sure recipe for disaster at least in the long run. There's a lot of talk and many opinions - for and against - about enterprise architecture, service-oriented architecture, event driven architecture or some "best practice" out there. There's not one proper way to think or act when planning overall architecture but if you don't know how your vendors are solving problems and most importantly if you don't care about it maybe you should think twice.
There are a bunch of human factors affecting this scenario - for example when there's only short term goals and division/business unit leaders have incentives based on them; of course it's natural try to reach the specified goals as easy as it's possible. But the goal should really create benefit for the organization in long run. If enterprise architecture and business process integration matters are not paid any attention when doing major environmental changes it will eventually create issues. And I'd strongly argue that it's much much faster and cheaper to react before than after. Antti's warstory about project-oriented integration development fits into this aspect as well.
Other extreme is to have "we'll develop everything from scratch" attitude. I'm not talking about R&D teams but guys in company's IT department. These organizations build noteworthy in-house development team(s) either from own employees and/or by using external consultants. In general nothing wrong here but there's a risk that you'd get blind sighted from the real objectives. It might be hard to see the forest from the trees. Without top level management decisions, prioritization and requirement analysis they'll just develop whatever seems to be the coolest thing to do or whoever is the loudest one asking for some new capability. If your team is doing nothing and waiting for tasks the pressure of getting something under work might lead to poor or misleading requirements and even completely wrong roadmaps.
Both extremes and everything in between eventually ends up to the importance of running and maintaining of a successful competency center; however it is implemented within the company. It's absurdity to outsource key knowledge when it's linked to core business processes. And when there are experts available from vendors to solve specific challenges it's unconsidered - and typically expensive - to ignore the expert knowledge and try to re-invent the wheel. Of course there are a lot of issues which affect on how organizations work and their history by far is not the smallest affecting thing but change is inevitable. Change might be slow in some environments but it will happen.
While it is important to get results, use agile development methodologies, improve the quality processes and in general be more productive. But it's more important to know why you are doing it. Keep the long term goals always in mind and try to avoid shortsighted decisions. Ways of doing daily business operations, business processes and underlying technology will change. So keep re-evaluating and improving your overall plan and architecture. Focus on your core business but don't forget that successful IT is part of the core. Use experts in right spots and choose them wisely.
Back ->
Back ->