Interstingly, I just went through this. We had to replace an collection of bought solutions that were duct-taped together.
I was part of the team proposing that we build it. Competing with some vendors of SaaS tools and a couple outside consulting teams that would go from either a full custom build to creating a collection of bought solutions.
There's a place to buy. There's a place to build. Absolutely. Both depending on the quality of the tool and the quality of your internal teams.
I don't like the over all "if possible" - It's almost always possible. It's almost always not going to do everything you need.
I question the "buy" approach because you MUST start to shape your business and processes around how the TOOL decides things should be done. Internal - You decide, the team makes it happen. (again - quality team matters).
OK; because the quality of the team matters to being able to be successful... I can kinda get behind the "buy" stance. ESPECIALLY for enterprises.
Over all though... it becomes a gate to what you can do. You start to think in the tool, not exploring all options. "Oh the tool doesn't have that" becomes an excuse to shut down ideas.
Buy has more impacts to the origanization and how it can evolve than just solving the immediate need. Buy has costs; ones that may not be noticable until it stops you from doing what you want.
All tools have this. All of them.
Build also has costs. Quality is the biggest. Since producing quality software isn't something the industry at large does; it's really hard for any given team to be able to produce quality software that can quickly adapt to the business requests.
I've been on a few teams like that - it's freakin' magic.
OK - What's the book say...
"Development could cost more"
Really. That's the argument. Evne if it only does 75% of what you want...
Ugg... I can't fully disagree... but you end up with a frakenstein of components trying to get to 100%... and which of components with the overlapping functionality do you use... it's... It's becomes a tribal knowledge situation; which generaly doesn't work out well.
I won't fully argue against either buy vs build - but a short term save of money for 75% of a system... not always a the trade off you want to make.