Earlier quoted context omitted.
The current leaders are way overpriced IMO ($9 to $49 per month per agent–are you serious?) I'm assuming these services are targeting companies with people working full time on support, where that $9 to $49 is a drop in a bucket compared to their salary.
That's the problem. Even small startups use them. Imagine a 4-person startup with 2 co-founders, another engineer and another designer. No full-time support staff, so everyone jumps in from time to time. Unless you make a generic agent account (which is lame for assignment/follow-up), you're talking $100/mo for support. I don't discount the value extracted but to me it reeks of enterprise, per-seat licensing a bit. I…
I've been meaning to do a blog post on such, but an average single support request costs us around $8. So if you've got a business where the average number of support requests per user are dramatically less than 1.0 per user per month, then it can work out, but I suspect there's a proportional support load to those SaaS systems that you're looking at where it's closer to, say, 0.5 per seat. If you lower the price beyond $9/month, then all of the sudden your profit evaporates.
It's non-obvious until you run the numbers, but it goes something like this:
Support Person Salary $60,000.00
Fully Loaded Employee Cost $90,000.00
Per Hour $46.88
Support Requests / Hour 6
Cost / Request $7.81
The interesting thing for where Bushido fits into that is that in some ways it'd be pushing things towards utility pricing (the same way a lot of the cloud sort of stuff has done) -- i.e. the cost of supporting a software stack would be proportional to the cost you generate further up the chain, rather than based on some metric of an "average customer".