Imagine you have a unique business offering that needs IT delivery and support. There are two approaches you can take to do this:
Bottom Up:
- Determine your business requirements.
- Choose an application which you think best fits those requirements.
- Spend a lot of money customizing the application to better fit your requirements and integrating it to other systems.
- Deploy the application.
- Hope you support the business adequately within time and budget.
Top Down:
- Map out the business processes to support your offering.
- Determine the services required to support the business processes.
- For services that already exist – do nothing.
- For new services – build or buy applications to support those services.
- Implement the processes.
- Hope….(ditto).
Often “bottom up” is the default approach, even though “top down” can be more cost-effective and provide a closer fit to your business requirements. However “top down” comes with some risk, because most off-the-shelf applications are built to support “bottom up” acquisition. You might end up having to do integration anyway.
As the world moves toward service portfolio management rather than application portfolio management, vendors move towards offering services rather than applications and the “top down” approach will become more widespread and more viable. That will be a good thing, I think.
2 comments ↓
Nice, Short
The biggest problem I see with this is that a lot of businesses don’t understand what they already have. As a business gets larger the problem becomes bigger and it gets remarkably hard to change to a “top down” approach. My gut feel, it would take an innovative and driven person to turn a large company around in this way
Well that’s what enterprise architecture is really all about. It’s relatively easy (often tedious) to understand what is already there (as is), and to understand where you need to go to (to be). The part you get paid the big bucks for is the transformation and that usually involves cultural change.
Interesting you mention “big company” – I suspect there is an optimal size for such transformation to be successful.
Leave a Comment