At least in the web world (which is the only one I have a whole lot of experience working in), there seems to be an interesting progression that developers go through if/when they leave the corporate world and try things on their own. Recognizing that fixed price development is very often a nasty trap is where many developers and organizations start to actually wonder about how their business model could change.

1. Fixed bid projects. Attempting to offer custom development as a package deal seems like a great way to undercut the competition; in reality it's anything but, and (among other things, including those mentioned in the article) reduces the perceived value of your work, which hurts in the long run.

2. Hourly consulting. Better, and works well for quite a while, allowing you more flexibility in your engagements and with your rate. This might even be a good long-term strategy for individuals; however, it requires very good planning and workload management, which are more difficult than many people think. It also does limit your earning potential, but this is much more of a factor if you expand into a consultancy, in which case your expenses will drop your margin.

3. Product/subscription service development. Often the subject of HN discussion (and rightly so), the prospect of recurring revenue is obviously a powerful one. However, even more than when selling development services, this requires a solid understanding of your target audience(s) and the ability to market to them effectively, in addition to understanding end-user customer support.

Where do we go from there? I personally have no clue; I've been slowly climbing this ladder myself.