The challenge of selling software that is essentially provided without cost is a complex, indeed. Savio Rodrigues and Matt Asay have both posted their thoughts and perspective on a subject that doesn't seem to have a consensus opinion associated with it. Unfortunately (or fortunately), I'm not comfortable attempting to 'clear the air' by expressing something of an authoritative viewpoint. What I am comfortable doing is, examining the varied complexities related to nurturing a business model constructed around open source software. So I'll take a stab at doing just that.
Savio mentioned how the majority of open source software users are "locked into the free price base," which is a well documented reality that isn't about to change overnight. Some might even argue that this free user base is a foundation structure for successful open source and I would have to agree, to a large extent. However, I have to wonder how this affects customer acquisition strategies for commercial open source vendors? Even if an overwhelming majority of customers begin their experience as free users (with a comparatively small number converting to paying customers), is the transition from free, unpaid user to that of paying customer taken for granted as a natural progression?
In my eyes the motivation for/circumstances behind pursuing a business relationship after engaging an open source community is anything but a natural occurrence. Is it a logical decision? Is there value in doing so? Yes on both accounts. All the same, to consider it a mere consequence of developing good software in the open and distributing it free of cost is a bit naive. For this reason, I have to question the logic behind Matt 's assertion that moving from,
"...isolated, departmental implementations (low dollars) to company-wide deployments (big dollars)...is the natural progression for any successful commercial open-source project."
Matt continues to say that,
"[Invading] the enterprise through free downloads and let[ting] the positive experience percolate throughout an enterprise until the CIO pushes a site-wide license."
This, too, might qualify as a stretch, if not a crass generalization. Yet this is the predominant view of exactly how open source works to win customers and generate revenue. Reality is a bit more complicated than the sort of 'silent penetration' mentioned above. By the time open source has penetrated the enterprise, associated visibility into the evaluation process becomes asymptotic to zero. Follow-ups become next to impossible, and since expiring trial periods are non-existent there are no reliable mechanisms for tracking it. Therefore, as Matt's post attests to, "leads" materialize almost solely based on impetus from the end use spectrum. Which, in a way, takes some of the wind out of the sales process.
These facts are worth considering within the context of debate surrounding the prospect of an open source company reaching the $100 million mark. Reaching that level implies that it's feasible to generate as much revenue with an impaired ability to discern and reach large chunks of an unpaid user base (the most reliable source of potential customers). This ignores the unfortunate reality that software (even open source) is still bought and sold. Yes, proprietary companies are inclined to ramp up sales teams to hawk their products and yes, it is a competitive advantage that open source companies aren't shackled to doing the same, but the $500k+ deals that don't happen by way of traditional engagement [driven by vendor outreach] through a sales team are the minority. Likewise, until the procurement process fully adapts to the realities of open source, relying on technology to sell itself will make the $100 million mark a barrier for most.