When my colleague Charles Babcock spoke to Red Hat CEO Jim Whitehurst, one sentence that fairly leapt out at me was "[Red Hat's] biggest competitor is simply people who stop buying subscriptions." So where do they go from there?

First, some context. The above statement appeared in a paragraph about the value of software subscriptions as Red Hat's major revenue stream. Obviously there are any number of reasons why people would stop buying Red Hat subscriptions, with them finding another product line being one of the biggest  Another would be that they've reached a comfortable plateau of functionality for the time being and don't see a pressing need to throw more money at stuff that simply goes unused.

Digression: There's also a third possibility. What about the folks who get their sea legs with an open source product (Red Hat or otherwise), and then realize they'd rather rely on the people they have in-house to continue the job? There is more than a little irony at work here. One of the reflexive assumptions about open source is that when you put it to work for you, it's empowering -- you can take steps with open source software that weren't possible at all (or only after great difficulty) with proprietary stuff.

But does that mean the average open source vendor's doomed to run into this wall, where they end up with their customers causing their own version of empty-nest syndrome? Probably not. I don't see this scenario being more than a very small percentage of how subscription-supported open source ends up -- and even if it does, I'd gamble that the people who soldier on by themselves may well contribute any work they accomplish back to the community.

Digression over. My main thought here is what kinds of work an open source vendor with a subscription-based model has to pull off to keep both new and existing subscriptions coming in and not going out, and figure out how to make subscriptions genuinely useful to customers who could go any number of other ways -- roll their own solutions, go proprietary, or ally themselves with yet another open source vendor.

The best way to do that is to be proactive -- to go after the people you already have and ask them: "What else are you trying to do?" (Not, "What do you need from us?") As colleague Mike Fratto put it in a message aimed at vendors in a slightly different industry: don't pretend that just because no one asks for something it's not something they want. Sometimes the greatest and most pressing needs go unarticulated precisely because they're so huge.

Follow me on Twitter: http://twitter.com/syegulalp