I was just looking at the prices at a popular dedicated hosting provider. I noticed that a simple Linux server with them cost ~$300 a month, with 50GB/month of traffic. How do people who purchase dedicated server service turn around and resell it for profit? Suppose you sell a plan that offers 5GB/month at $15/mo. You'd "fill up" your bandwidth with 10 accounts, so now you're $150 in the hole.

Is there a formula to success? Do most people only use a fraction of the bandwidth you provide? Would any dedicated host-ers out there like to share what their mileage has been?

Thanks,
Andy

First... You want to get more bandwith for that price. $300 for 50GB is not too great a deal. Many hosts here can even provide you a better deal at this level of monthly charge.

Second... Find a host that has cheap bandwith upgrade. You don't want to pay $10/GB for additional data transfer. Many hosts may offer you cheap dedicated server deal but then the overage bandwith charge is way too expensive. Find a host that can offer you around $3/GB of overage data transfer and possibly less. But no more than $4/GB.

Third... You will be surprised that on the average, people only use 200-500MB (0.2-0.5 GB) of data transfer per month. You may offer your clients 5GB of data transfer per month, but on average, they will only consumer much less data transfer per month.

Hope this helps....

cheers,
:beer:

But also consider this:
If some of your clients were to use up 5 GBs, your screwed over :)

If some do you won't be 'screwed'. Most won't, and that is what overselling is about.

"The typical oversubscription rate is 4 times"
http://www.isp-planet.com/business/pricing3b.html

That's why there are many web hosts selling bandwidth (especially virtual hosting) below their cost.

Many web hosts would go bankrupt if people actually used what they pay for...just like cell phone plans.

Originally posted by energy
Many web hosts would go bankrupt if people actually used what they pay for...just like cell phone plans. And that's why it's priced the way it is. For the end user it can make sense to go with a package that has limits beyond what you'll ever use, because you're then locked into a monthly cost that's manageable for you. Just like, as you say, cell phone plans; people generally buy a package that allows more minutes than they think they'll ever need, and if they find out that's not the case they probably upgrade to a plan that does have enough rather than paying the high price of extra minutes.

You'd have a hard time selling hosting if you sold by, say, a by-the-meg rate from the first meg -- just like cell phone companies couldn't sell as successfully a cost-per-minute plan -- no matter (within reason) how low that rate was. People generally like to have a set cost they're comfortable with and not have to worry about it.

Originally posted by JayC
You'd have a hard time selling hosting if you sold by, say, a by-the-meg rate from the first meg -- just like cell phone companies couldn't sell as successfully a cost-per-minute plan -- no matter (within reason) how low that rate was. People generally like to have a set cost they're comfortable with and not have to worry about it.

I disagree. I think there is in fact quite a large market for such hosting -- assuming the price is reasonable -- with people who have small websites. I think the reason such hosting plans do not exist is a lack of interest on the part of web hosts; if it takes you five minutes to create a new account, you probably don't want to do that for someone who will only be paying ten cents each month.

Cperciva, I think you are right.
I am a reseller and one of the two reasons why I changed my host was for non-existing small sized plans.

Originally posted by cperciva
I think the reason such hosting plans do not exist is a lack of interest on the part of web hosts; if it takes you five minutes to create a new account, you probably don't want to do that for someone who will only be paying ten cents each month. Yeah, that's a good point. But I'd agree with the second part, too, which goes back to that "within reason" statement. There's a point at which it isn't going to be worthwhile for a host. So the cost would have to be more than that, or the measured increment would have to be higher. And once you do that you're heading toward the point where people won't want to risk the potential charges if they do suddenly get a bunch of hits for some reason.

There is a certain cost above just the space on the HD and simple data transfer. That $1.50/mo client is going to take as much, if not *more* of your time for customer service, and call me crazy, but I don't work for $0.05/hr.

IMHO, anything under $15/mo isn't all that worth it, but that's me. If they *never* send support emails, then *maybe* $10.

Pay yourself for heaven's sake. You work hard!

Originally posted by Chicken
IMHO, anything under $15/mo isn't all that worth it, but that's me. If they *never* send support emails, then *maybe* $10.

If they never send support emails, and account creating and billing is handled automatically, why should there be any minimum price (beyond of course the cost of resources consumed)? Maybe I'm missing something, but it seems to me that if a customer doesn't take up any of your time it's a good customer to have regardless of how much, or how little, they are paying.

My point is that the minimum costs are largely derived from the fact that hosting is still a largely manual process, requiring human intervention to create accounts, handle billing (to a lesser or greater extent), etc. Take the humans out of the process and everything would change.