I agree, but don't forget the flip side of the coin

What about the total cost in $x months/years? Is it worth spending the time and money now, as opposed to dealing with a product that you can't get under the hood of? What is their turn around time for bugs? Do the have a "forced" upgrade path, which you *must* follow in order to maintain your support contract? Do they end up providing Platnium service to a few of their customers, while the smaller shops have to deal with stuff as is? If you follow thier upgrade path, and need a new server, do you need a new licence? so on and so forth

I understand the framework you are coming from, and prefer if people dont really understand a problem space, that they go with a canned product, at least initially. But if the organization is fairly knowledgable about what they want (I can hear the groans), then doing in-house devel works out in the long run. The turn around time should be much much shorter on bugs and feature requests, even if instead of a true in house product, the organization maintains a business relation with a contractor who wrote the program. It really depends on how fast and or frequent changes occur to the core business practices and goals, and how this product ties into the scheme of things.

Just my 2 cents

use perl;