Giving It Away
From Zanecorpwiki
I recently read a quote from a fellow named Gerry Cullen who stated, If you can't give it away, you can't sell it. I have no idea who Gerry Cullen is, except that he's a sales guy without a Wikipedia page (but then, I'm an entrepreneur without a Wikipedia page so I guess it's a wash), but this struck me as one of those obvious statements that is in fact completely wrong.
The problems of selling free have been frequently discussed, and I'm not saying anything new here, just summarizing. First is the oft cited problem of perception. The free couch is a minor meme uses to illustrate the idea that something which is free will often be perceived as worthless. If one could get money for the couch, then one would ask for money. Therefore, the perspective couch acquirer concludes the free couch is worthless and passes by.
The bigger problem with free is the implication for future cost. You spend an hour dragging the free couch to your living room only to find that it smells like piss when placed in an enclosed space and will stab you painfully in the ass 10% of the time. So you decide to cut your losses, drag it to the curb and stick a free couch sign on it, but after it rains and molds up, you know no one's going to be fool enough to take the thing and you're neighbors give you dirty looks. So you spend three hours talking your buddy into lending you his truck, hauling the thing to the dump and paying $50 to get rid of it. Now you're Saturday's shot, you're neighbor's hate you, and you blew the $50 dollars you had to buy a couch from Craigslist.
That's why no one gets more than one free couch in their life: they just cost too much.
Now free can work. I'm a huge fan of Linux, Firefox, Apache, and lots of other free software. A free software guy reading this might be cringing now because I've been talking about free as in beer, but the software I've listed is free as in speech, but that distinction doesn't matter. It's both, and so was the couch. Just like a user is free to rip apart Linux, write their own drivers and re-implement the scheduling sub-routines, the owner of the piss couch was perfectly free to re-upholster and re-spring it. In either case, if there's an option to pay money for a non-free solution (e.g., commercial drivers or a $50 couch from Craigslist) that solves our problems, that's what we're going to do. The question is whether the non-free solution is cheaper than the free solution.
Just like we learn to be wary of the long term costs of the free couch, we learn to evaluate the long term costs of software. It took years and years to convince corporate buyers that free can actually be cheaper because there's an inherent risk with free systems. Let's say I have to bits of software--spam filters, word processors, browsers, it doesn't matter. One will cost me $50 bucks and the other is free.
With the $50 piece of software, I have a good idea of my costs. $50 dollars plus some unknown future costs, $X which will come in the form of time and cost of upgrades and bug fixes. However, I know that the company selling me the software has incentives to keep me happy so that I'll tell my friends how much I like the software and they can keep selling it. This limits my expected future cost to something reasonable.
On the other hand, with free software, I pay nothing but in general I have no expectation of limitations of my future cost, which we'll call $Z. I may have to spend hours scouring the Internet, hire a developer to fix the problem, crawl their code and fix it myself. I may not even be able to fix the problem at all. If the software occupies some important part in my business, this failure with no clear solution could be catastrophic. So, even though the initial cost of the free software is less, the expected future cost Z of the free software is much, much greater than the expected future cost X of the non-free software. Therefore, I expect the total cost of the free software to be much, much higher than the total cost of the non-free software.
That's why successful free software is zealously supported by either a commercial entity, a large user base, or both. There must be some mechanism in place so that the user can feel secure that whatever problems they encounter can and will be addressed with little cost on their part. Even once these mechanisms are in place, as has been the case with Linux and Apache for over a decade, it still took hundreds of case studies and many, many more years for even the best free software to overcome the perceptual problems and be adopted into business.
With the success of some of these large projects, free became fashionable for some time. It still is, in fact, but that's beginning to wear off. Users are realizing that free in-and-of-itself is not necessarily a good thing. There needs to be some reason to expect that the free software won't start smelling of piss after a week. This means either a large community, or vetting by trusted third party, as happens with distributions.
Free, however, is no longer a selling point. There are still many good reasons to do free (in software and with other creative activities), but free, is better understood as an obstacle to overcome than as a feature in-and-of-itself because any half informed potential user is going to understand that nothing is free and that the initial monetary cost is merely a small part of it.
Which brings us back around to the free as in beer vs free as in speech distinction. Not as a socio-moral distinction, but from a purely economic standpoint. Free as in beer doesn't make software cheaper, but free as in speech does because individuals that care about software as their modern form of expression are willing to invest their time in it and the user benefits from that investment. Participants in the movement often bemoan that users who fail to recognize the nobility of free software don't understand what it's all about. The truth is, however, that the nobility must have an economic expression to be of any consequence.
We can lessen costs through innovation and process, but any creative act necessarily entails cost. We can shift the burden of these costs around, and by the proper arrangement of that burden we can maximize the effect of our efforts. For such optimizations to be successful, both producers and consumers we must at all times recognize that the cost must be paid.
