Kim Morrison wrote:
>
[snip]
>
> The fact that it was a challenge to sell your customer on a
> product for which you paid $40 is not surprising.
> Positioning e-smith at $40 against Microsoft at $1,500+ has
> got to raise questions in the typical small business person's
> mind about the value of e-smith against Microsoft.
Interesting you should say that. I had a similar conversation with Joe last night -- sometimes pricing yourself or your product too low is as bad or worse than pricing too high. Rightly or wrongly many people (not all) base their value judgement in part on the price.
For example, I once needed to hire a Mac expert to come in, examine a publishing process that used a hacked together VB application and an HPIII LaserJet, and recommend a Mac-based replacement that would output directly to a RIP and ultimately enable direct-to-plate output. The Mac expert I found thought the solution would require about 8 hours to work out, and sent me a proposal for $600 ($75/hour). The owners of the company rejected it, because 1) there was no way anyone could figure it out in 1 day and 2) it couldn't have been much of an expert if he only wanted $600. As I had not disclosed the name of the consultant, I asked him to submit another proposal, one which required him to come our site, spend two nights in a hotel, and charged $150/hour for three days. I presented the proposal for 24 hours effort at $150/hour, plus $700 travel and expenses, and it was approved in less than an hour. The solution did in fact take less than 8 hours to identify and implement, and company ended up paying $4300 for something they could have had for $600.
One must also consider the value of knowledge and the sale of knowledge as a product. When I was starting out in consulting my first job was for a doctor. This was years ago, when Apple ][ computers were the state of the art, and the PC was newborn. The doctor wanted once of the first PCs to "play with", and also wanted more memory and hard drives for his Apples. I did the work and charged him $10/hour, which was good money (I thought) at the time. He looked at my bill, took out his checkbook, and said, "Take this as a lesson. I am not paying you for what you did. The actual effort was not that great. What I am paying you for is for doing what you know how to do that I do not know how to do for myself and do not want to know how to do. When you come into my office to see me, I may spend only five minutes with you, but I charge you $35 for an office visit. You don't pay me the $35 for the effort I expended in that five minutes. You pay me for the years of schooling, my experience, the time I spend keeping my skills and knowledge current. You pay me to do for you what you cannot or do not know how to do for yourself, not for my actual time and effort with you." And with that he wrote me a check for $30/hour.
What e-smith has done is very much like this doctor. Some of the people on these forums probably know how to setup firewalls, mail servers, Apache, ftp and ssh, pptp and ipsec VPNs, and so forth. For them, e-smith is not just cool software, but also a novelty or a time-saver. If their backs were to the wall, these people could probably take Red Hat or Mandrake and create their own servers. I'm betting, though, that the majority of e-smith users and customers, myself included, either don't know how to get Linux to do most of the things that an e-smith server provides or feel that it would be far to much effort to have to figure it all out from scratch. And that is what we, or at least I and the company I work for, pay e-smith/Mitel for. Just like going to the doctor.
The new IPSec VPN in ServiceLink is a prime example. I've seen this beast in action, and it is simply amazing. It is simply simple. I've also seen our own network services group (approximately 50 people) struggle for over two years with VPNs using Win NT/2K, and while their solution is perhaps more robust in logging and error reporting and provides a few more configuration options than ServiceLink, it also takes a MCSE 2-3 days to setup a new VPN and requires constant tinkering to keep it running (each MCSE manages about 8-10 VPNs.) ServiceLink VPNs take a minute or two to setup -- and that's if you take your time and triple check everything! -- and require virtually no recurring maintenance. When you buy (or sell) ServiceLink, you're not only buying the software that provides the mechanisms and the time on the servers that provide the service -- you are also buying the knowledge that went into creating the system that makes it so simple. Ditto that for the basic server -- the firewall, the email, the i-bays, etc.
> Being open source, we naturally wouldn't raise our prices to a
> competitive price for the software alone. We've been working
> toward crafting a solution where we add value "around" the
> software and are therefore able to raise the price to
> competitive levels. I would never want us to go back to a
> $40 media-only package since, to the average SMB customer, it
> makes our software seem like a toy.
Yes, the $40 model. Let's be generous and peg costs at 25%. If you have 25 employees who impose an average overhead of $35,000 (salaries, taxes, benefits, fixed asset allocations, etc.) per year, you would need to generate a minimum of $875,000 per year just to have a prayer of breaking even. At $40 per copy ($30 after costs) you would have to sell 29,167 copies each and every year.
Let's take the $395 model and be even more generous and skip costs. Taking the same 25 employees and their $$875,000 of overhead, you'd still need to move 2,215 contracts per year.
Anyone familiar with business in the US/Canada area will recognize that I've understated the case when it comes to the expenses incurred in running a 25 person company. However, even with the favorable bias, the numbers don't work.
> We shouldn't go back to a publishing business that doesn't
> pay my staff's paychecks. Heck we put the software up on a
> free download site and we offer a nice PDF manual, isn't that
> more than you can expect from any other software company?
No, you are on the right track. There is refinement to be done, especially concerning the blades model -- though I suspect this work is already underway and I await the announcement
> I fully realize that part of your desire for a package is
> that you'd like to support the business and I really
> appreciate the sentiment. Actually, your on-going moral
> support and your contributions on these forums have always
> been fair trade. I did notice that you poked at our security
> model in your email. What would be a really helpful
> contribution would be a detailed listing of which security
> practices you question or that you'd like to see
> implemented. That's be a terrific way to support the business.
Indeed, that contribution would be useful. The comment could be interpreted several ways, some of them having nothing to do with firewall or server security.
Scott
38 Replies
4262 Views