Thimbles & Care

(This is a continuation of the previous post. I will probably have more to say after this one, too.)

One way to avoid developing a vision for administrative IT would be to in effect outsource it by stopping in-house development and migrating to a commercial ERP package like PeopleSoft. There is a strong argument for doing this: Nicholas Carr wrote a whole book (Does IT Matter?) arguing that IT has become a commodity and that there’s no strategic advantage in building your own applications. While I think he overestimates the maturity and stability of the IT industry, he may be right.

On the other hand, we do have some evidence suggesting that building applications ourselves has benefits. In his post for staff appreciation week, Pres. Powers said:

The percentage of UT Austin’s budget spent on administrative costs (5.5%) is about half of the average for Texas public universities.

I’m sure there are many factors that contribute to these lower costs. UT Austin can undoubtedly take advantage of economies of scale that smaller universities can’t. Austin probably attracts more intelligent and creative people than most places, so UT has a better caliber talent pool to recruit staff from. But I also suspect that one of the things driving lower administrative costs is the caliber of applications we’ve developed here—if we used the same software as everyone else, it seems likely that our costs would be more in line with everyone else’s.

Also, we should realize that the cost of moving to a commercial ERP would most likely be similar in scale to the costs outlined in the mainframe migration assessment, as much of the effort would be the same or similar: testing, integrating the migrated pieces with the parts not yet migrated, purchasing new hardware, retraining staff, and so on.

I’m not convinced that migrating the a commercial ERP would benefit the University, but if you believe the argument that IT is a commodity it would make more sense than the proposed mainframe migration.