- Very few companies want to take a chance on software products coming from smaller ISVs. There's been a great deal of consolidation, and companies are scared to use anything that doesn't already have serious backing and market penetration. Reasons cited often include not being able to hire developers or other IT people to manage the software, or concerns about the longevity of the ISV. In contrast, back in the 80s you had products like FoxPro being created by tiny Fox Software in Perrysburg, Ohio that went on to become some of the biggest developer tools for PCs. This attitude of not wanting to use something because of some other tangential factor was simply not in play. If you could solve someone's problem well, you had a customer.
- Somebody touched on advertising in one of the comments, and this cannot be understated. There is simply no reliable way to find your audience today that doesn't involve "spray and pray" with one's (already-limited) advertising budget. Advertising used to be a great way for smaller ISVs to become known and announce to the world that they were the "real deal". You would start out in the back of a popular magazine that targeted your potential customers with a B&W 1/4 page ad, and then move up as sales increased. It's just very, very hard to get your product in front of the right people that are making the purchasing decisions, and even when you do, the first issue above still comes into play.
- There were options like shareware for ISVs that were just getting started and didn't have any money at all for advertising or promotion. Many great applications started out as shareware and went on to become commercial successes. The key difference with open source here is the expectation that you were supposed to pay for the product if you used it, as well as, obviously, a lack of source code.
Of course, all of this is a giant feedback loop that accelerates over time, and here we are with a few software companies completely dominating various spaces. It could be that this is where all industries eventually end up, but it certainly is not ideal for small business creation. The part I find the most disheartening is that I'm not even sure if this is something that is specific to the software industry, in which case we're dealing with macro issues and software developers are really lacking any tools to buck the trends or change course.