What struck me most about the article on Enterprising Networking Planet about open source support and when you go commercial was the stark distinction made between paying and non-paying user. The article's investigation of the two ends of the market for open source support, namely those who pay for commercial support and those who don't, brings the question of where the two overlap into question. In essence, this proverbial middle ground will drive the next phase of growth for commercial open source support providers. My reasoning is that the needs of both the unpaid and paying user are mostly alike outside the fact that the former chooses to forgo pursuing a commercial relationship with a support vendor. The assumption that those using open source in mission critical/production environments choose to mostly opt for support doesn't hold much water especially when you take a peek at the unique download numbers for most commercially supported products in relation to the number of paying customers.
Yet this mostly anonymous crowd of non-paying users represents the exact source of growth opportunity for commercial support providers. However, engaging them shouldn't be misconstrued as "converting" them to sources of revenue. Instead, it's probably more prudent to actually support them as self-supporting users. This might sound contradictory but it's worth considering that while some non-customers will eventually convert most aren't necessarily in need of being convinced to enter a business relationship. One of the main reasons: a great deal of self-supporting users are already aware of the risks and difficulties associated with forgoing commercial support but have decided to bare the burden anyway. With this in mind it is of little consequence if a support provider offers X, Y, Z in the same way that paying customers aren't inclined to let the prospect of saving money by relinquishing commercial support tip the scales in the other direction.
Nonetheless, as the aforementioned article pointed out, reliability underpins a large majority of the decisions to pursue support...especially in the face of troubleshooting and problem resolution. Obviously, non-paying users face the same issues only they have committed to the path of going it alone, one that isn't easy but still proves rewarding if traversed with care. Interestingly enough, this expression of independence does not imply that there is a absolutely no market for support within this category, only that the market is for a different type of support. It helps to admit that any variety of support contract creates an according dependency on the capabilities of an external vendor, which isn't wholly negative but isn't always preferable within the context of choice afforded by open source.
One of the positives presented by opting for an open source solution is being able to acquire quality software in the process of building associated internal competencies. Oddly enough, the intersection between employing open source and building self-sufficient competency is where support providers should examine how they can evolve into a more relevant entity. Perhaps instead of attempting to become the sole source for upgrades, vulnerability assessment and/or migration, support providers should seek to offer a wider array of flexible support (and licensing) options for those who choose to go it alone? Pricing could resemble something similar like car insurance with payments of a low, flat rate (monthly, quarterly, yearly) for emergency/disaster support. Or maybe even pay-as-you go support (troubleshooting, short-term consultation, etc.) that is billed by the hour. This would afford non-paying parties with the option to pursue just-in-case interactive support that better aligns with a mostly self-sufficient approach.
Obviously these forms of support would be offered at substantially cheaper prices than is customary but they would also appeal to more non-paying users thereby increasing derived revenue from an open source user base. For commercial open source vendors that provide commercial and community versions this might take the form of extending this brand of support for the latter. Either way, it's my perspective that the picture of successful open source in the next 5 years or so will include yet-to-be conceived forms of support options that better meet the needs of the unpaid user.