Two little stories:
A friend of mine has been trying since 10 years to transform his consulting company into a product company - without success. The projects he has implemented were too expensive in terms of development and maintenance to have enough time apart. And none of the projects was general enough to produce a code-base that could serve as a product for several customers.
Another friend implemented a little access-based niche-domain tool during an internship for a big enterprise that he was able to resell to several competitors and even several times within the same enterprise. He is working around 5 hours per week and makes a confort living.
Honestly, I think the first story gives the more likely scenario.
I'm biased to the security software market, but note that it's a market where companies have constantly tried to play the "productize-the-service" card. What I've observed is, this almost never works.
In fact, taking a consulting approach to product development (working closely with a first customer) also slowed us down drastically, and our product actually has very little to do with our service (what we're launching has nothing to do with reverse engineering software).
There are also some high-profile failures behind our strategy: building a straight-up consultancy to "fund" a full time product group.
But then again, who knows? Building a business is hard. It is not like applying to college, where you clear some "acceptance" hurdle and then you do your homework and everything works out. With the best people, the most credible strategy, and even the right market timing, circumstances can still screw you over.
Maybe the one just got lucky and the other was unlucky? It's just anecdotal evidence. The HotOrNot founders got rich by scribbling something on a coaster (or something like that) - doesn't prove we should spend more time in bars to create a successful startup. Or maybe it does :-/
