These are great points. IF you can get away without a phone number, you train your customers early on to use other means. Yes, you may lose out on customers, but this is built into your profit model.
Taking phone calls costs money. Typically, people who would prefer to call are going to be not as technical, and will indeed require 'support' that goes beyond your product, because if your product isn't working, you have to fix it or at least diagnose it on that phone call, even if it's because their networking is down. You essentially become their "IT" department.
So, not having a phone number will cause you to lose sales, but probably with those sales are the 'expensive' customers that you probably would not have made money on after factoring the extra support costs.
That being said, if you must have a phone number, increase the cost of your product to cover the extra cost of that.
Very true!
In my first job at a small hardware manufacturer, all we had was phone support (this was late 80's - early 90's, email wasn't widespread). The people who needed the most support often became very good customers because they knew they could rely on us to help them when they needed it. Almost all customers needed pre-sales support and a fairly high percentage needed tech support after purchase depending on what they bought. One of our product manuals was always being upgraded because the product was so complex that we began to anticipate the customer questions and answer them explicitly in the manual. The average sale was probably around $300 but we had a high number of repeat customers many of whom we knew by name.
As you said, the cost of the product must cover the anticipated support. If you are covering the cost, then make this an opportunity to differentiate your company even more by offering excellent support beyond what your competitors do.
So, I think this is one of those lessons learned from the school or hard knocks, for me. Sometimes it is just as important to be clear with yourself and the rest of your company about what customers you aren't trying to serve as being clear about what customers you are trying to serve.
You cannot satisfy everyone, but early on, when you have very few customers, you may tell yourself, "This customer would be worth doing something very different for." It is tempting to do anything you can do make a sale.
Sometimes, this is just a sign that you aren't targeting the right people yet, and it's time to evolve and grow. But, if you aren't careful it just leads to you trying to be all things to all people. I had this problem with my previous company...the questions would come, "Can you do X, Y, and Z?" And my thought process would be, "Yes on X and Y, and I could totally make Z happen, if I just spend two weeks developing it, making this a losing sale, but I'll amortize across future sales of Z." And I'd respond, "Yes." I'd make Z possible, and then I would never sell to another customer that needed Z. To be fair to this thought process, Virtualmin wouldn't exist had I been successful about saying "No, that's not something we do." But, counting up the successes vs. the things that merely distracted me from my real goals, that's batting about .025. And, I had wanted to make something like Virtualmin for years (and I'd even taken a stab at it three or four years previous and written 10kLoC, or so, and had some basic functionality and a half dozen users)...this customer just gave me an excuse to spend enough time/money on it to make it a reality.
Because I (almost) never have to spend an hour on the phone with anyone, either to make a sale or to support the product, I can spend that hour helping a dozen people solve their problems. Thus, sales are cheaper for us...so we can charge less. Support is cheaper for us, so we can do more of it, without having to charge extra as our competitors do.
If I ever said, "Yes", to the question, "Can I call you?" all of our math about how we make a profit on software that is as large as ours, while being as cheap as it is (and being 90% available in an Open Source form), would completely go out the window, and our prices would have to become too high to be competitive, or we would just start losing money on every sale, as I did in my previous business. I'm not saying answering a phone now and then would put us out of business...but I am saying that I don't want to be a telephone IT support business ever again (you're spot on in your assessment that you become every customers IT department, because you're the only company that answers the phone and has smart people on the end of the line--I can't even count the number of times I've fixed peoples routers because they couldn't afford a Cisco support contract). Before offering to answer phones, think about whether your business can support pricing that high (if you're unfamiliar with Cisco support pricing...do a little digging...be prepared to gasp), or if it will price you right out of the market.
Taking phone calls costs money. Typically, people who would prefer to call are going to be not as technical, and will indeed require 'support' that goes beyond your product, because if your product isn't working, you have to fix it or at least diagnose it on that phone call, even if it's because their networking is down. You essentially become their "IT" department.
So, not having a phone number will cause you to lose sales, but probably with those sales are the 'expensive' customers that you probably would not have made money on after factoring the extra support costs.
That being said, if you must have a phone number, increase the cost of your product to cover the extra cost of that.