Discussion on:

Consider these points when making the build vs. buy decision

I've been struggling with the 'buy or build' problem for a couple of months now. I know that building a solution is going be expensive and so will the maintenance. On the other hand, the software that claims to do what I need is pretty pricey, too and comes with and 15-18% annual maintenance fee. My two biggest concerns are 1) the commercial packages look like the actually try to do too much and we'd likely get buried in data entry and maintenance--possibly requiring additional staff and 2) there is some evidence that each of the commercial packages have deficiencies that just shouldn't exist in what is supposedly a fairly mature field (MRP).

What I think I need is some kind of template (or better a database system) that I can use to identify the major requirements, complete with critical data fields, and record the effectiveness with which the commercial packages meet those requirements. In the end, analyzing the results of the 'survey' will give me a much better idea of whether to build or buy, and if I decide that I have to build, I'll have a requirements overview suitable for presentation to a contractor.

Just to be clear, don't worry, I'm not an open-source envangelist about to go off on a tangent =). However, I do think it should be noted that open-source projects do seem to be having an increasingly significant role in the buy-vs-build decision. As the previous commenter noted, there are a lot of commercial packages that don't meet up to the benefits listed in the article. Despite their high price tags, these packages have deficiencies that are caused by possibly rushed coding and testing, sothat the final product can be on shelves by the time the marketing department promised. They may also be overkill in both complexity and price, actually slowing down the development process. Finally, buying software means you're tied to the company that makes it, for better or for worse. You could get a reliable company with great support and stable upgrades, or you could get a nightmare.
In any case, open-source is a good choice to be aware of, because it bridges the buy and build camps. It's usually less expensive than commercial software (or even free!), and has often been refined to a very mature level. Also, the only reason you're tied to a vendor is their service and support, since you have the source code. Besides giving you more leverage and hopefully better service, it also means you won't be totally screwed if they go belly up. I've also noticed a lot of very well written open-source base libraries lately, for many languages. This seems to be what a lot of developers are looking for; a flexible, modular foundation on which to build a customized project.
Of course, there's a lot of badly written, badly (if at all) supported open-source projects out there, so there will still require a lot of investigation and decision-making. However, open-source is definitely a choice that deserves consideration. Besides, if there wasn't any catch, it would be an easy choice wouldn't it =)

Wear the shoes of User and see the application from his role.

Ensure that sucess / failure is determined by user and not by developer or PM.