My own take on this subject:<p><a href="http://www.smashcompany.com/business/how-much-do-websites-cost" rel="nofollow">http://www.smashcompany.com/business/how-much-do-websites-co...</a><p>That was written in 2008, but there is nothing that I would change (although, some of the anecdotes are clearly shaped by the strange gold-rush mentality that infected a lot of first-time entrepreneurs, circa 2004-2008). Back then I was part of a team that helped inexperienced entrepreneurs shape their first software product (almost always a website). We were getting too many solicitations from entrepreneurs who had no idea how much things would cost. They would explain their idea to us and we would say "We can build that for $50,000" and they would be like, "What?!? Hey, the guy down the street said he could build it for $500. I'll go with him." So I wrote this essay to help set expectations for clients. Back then, I was living in Charlottesville, Virginia, in the USA, and some of the anecdotes are specific to that time and place.
Well, one conclusion we can draw from the article is that 90% of code is not considered good by author's standards (which I not only agree with but take one step further), yet we weren't presented a case in which crap code could be mitigated.<p>I mean, when your employers insist on employing the bad engineers because they talk big jargon, that shows the Real Cost of Stupid Management more than anything else.
Crap code really depends on the project. I used to write crap code for clients (unintentionally, of course!). They did not think it was crap code, but looking back ten years, I think it was. But the applications did everything they were supposed to, were stable, and they had a short lifespan so no one else was going to need to come in and edit the code. Sometimes "gets the job done" code is the best. :)