[bialecki]

This kind of comes out in this post, but the way I always look at the "I'll do it myself" argument is in terms of time. If you have a product that people known is easy to use (more on that in a second), you just say, look a license costs $100/yr or you could do this yourself and spend 10+ hours working through all the issues I've already solved. So if your time is worth $10/hr, go for it, otherwise I'm saving you a lot of your valuable time.

The only wrinkle to this is that you have convince people that your product actually does make their life easier/save them time. If they think they'll end up spending a lot of time understanding how to use it and integrating it, that's when people say, not worth it. For instance, you could be selling an awesome jQuery table plugin, but if it doesn't support some feature I need (say, a find as you type search box), I'm probably not going to use it because the time to build that feature onto your plugin might be longer than building my own.

[HerraBRE]

It's not just about time you can save right now, it is also about risk.

Depending on a 3rd party, proprietary solution may be considered a form of technical debt. If (or when) the 3rd party raises their prices, goes out of business or does something else unhelpful to your business, you'll have to replace them.

Depending on what it is, that kind of work can be far more costly late in the game than it is at the beginning.