There are a couple factors at play here. First off, the average programmer is, honestly, not very good. The can just barely make anything useful, good luck having it also be well designed and coded, secure, robust, fast, etc. Second, programming is like wizardry to most non-programmers, it's hard enough for a professional software developer to judge another, it's nigh impossible for J. Average Businessperson to do so, even if they have a dire need for software development. In the average case you might have some contract go out which ends up being won by some enterprisey line of business app development house who poops out some horrible cobbled together tool based on Excel, MS Access, and a winform app somewhere in there to glue things together, and then bills the customer for 6 or 7 figures (I'm not even joking). Compared to that, your average startup is the dream team.
And what's fantastic about funded startups from the perspective of a software muggle is that not only might they have a product already in existence that you can check out and find reviews from other customers but they've also been given the stamp of approval by investors, investors who are likely far more knowledgeable than the average software consultancy customer. Just look at the Obamacare website debacle as a case in point of how incredibly difficult it is to a: find a quality development shop who can actually deliver what you want, and b: do so within reasonable budget and time constraints. And the difference between finding a good development shop and a poor one isn't merely a better quality product, it's a factor of maybe 2x or more in development time (which is, of course, tremendously valuable) and a factor of 10-100x in cost.