How many of you have staff?

[Zachary McClung]

Yeah, I told him that those figures seem kind of weird, but he insisted that is how it should be. Well, thanks for going into detail.

My pleasure! How long ago was it that he told you that? There might have been some truth to it four to five years ago. When I started in the industry there was $20.00 - 30.00/month shared hosting plans with 500 MB of space & 1 GB of bandwidth. I miss those days.

When you look at it in that sense a customer is paying $180.00/year and they would be making $180,000.00. The server may have costed $6,000.00 per year. You would still be at a stretch with 12 employees but I'm sure with the write high end clients your friend could be correct. That sure is not happening these days though.

 

[handsonhosting]

1500 clients per tech would be an insane environment and likely lead to a high turnover of staff (depending on issues that come up etc).

We don't put more than 100 clients on a server and only 12 customers on an ecommerce server. This is so that if an issue does come up on a server, we can isolate it without causing widespread headache for employees. Ever had 300 people asking why their site is down? Ever wish there were only 10 asking the question?

We have 3 system admins to cover 300 servers, but it all depends on what is going on, and what tasks are automated.

Our support staff run double shifts as sales staff. 12 employees to handle 1000 accounts would be nice, but hard to do on a shoestring budget.

 

[Zachary McClung]

1500 clients per tech would be an insane environment and likely lead to a high turnover of staff (depending on issues that come up etc).

We don't put more than 100 clients on a server and only 12 customers on an ecommerce server. This is so that if an issue does come up on a server, we can isolate it without causing widespread headache for employees. Ever had 300 people asking why their site is down? Ever wish there were only 10 asking the question?

We have 3 system admins to cover 300 servers, but it all depends on what is going on, and what tasks are automated.

Our support staff run double shifts as sales staff. 12 employees to handle 1000 accounts would be nice, but hard to do on a shoestring budget.

handsonhosting:

You think 1,500 accounts per tech would cause to hi turnover? I'm not sure how many of your servers are shared & ecommerce; however, even if you had 12 clients per server you would have 1,200 accounts per server. You also charge $37.00/month/ecommerce account which allows you to utilize your strategy.

If you placed 100 accounts that were at $2.99/month your strategy does not work well.

 

[Artashes]

When I started in the industry there was $20.00 - 30.00/month shared hosting plans with 500 MB of space & 1 GB of bandwidth. I miss those days.

Sure, but then hardware was equally expensive.

 

[handsonhosting]

however, even if you had 12 clients per server you would have 1,200 accounts per server.

I think something is missing there, and I do want to know what you're saying. Can you fill me in?

I think I ended up going on two different tangents in the one post :uhh:

The techs per customer being 1500 is just high (at least to me). It means that customers are being treated as a number and not as an individual. Difference between small business and big business. I couldn't fathom having a single rep with that many accounts. 500 I can see, but 1500 is high I think - to me anyway.

The 12 people on a server versus a few hundred or thousand just had to do with sanity. Nobody likes a burnout, and in the support world, we all go through it every now and then and need a complete break away from the computer. The turnover was more in reference to the employee rather than the customer. Running a staff at 65-75% capacity usually equates to happier environment. It SOUNDS like money wasted, but sanity of personel is a key priority for us anyway.

 

[Zachary McClung]

Sure, but then hardware was equally expensive.

I wouldn't say the drop in hardware was in the same portion that the drop in shared hosting fees were. I seen an Ev1 Server had from years ago and they were running $99/month. You could hold 500 - 1,000 on those old P3 - P4 when the site was only 10 MB. Even the big 8 GB dual-core at that time was $399/month. You could make more money back then, then what you can now.

 

[Zachary McClung]

I think something is missing there, and I do want to know what you're saying. Can you fill me in?

I think I ended up going on two different tangents in the one post :uhh:

The techs per customer being 1500 is just high (at least to me). It means that customers are being treated as a number and not as an individual. Difference between small business and big business. I couldn't fathom having a single rep with that many accounts. 500 I can see, but 1500 is high I think - to me anyway.

The 12 people on a server versus a few hundred or thousand just had to do with sanity. Nobody likes a burnout, and in the support world, we all go through it every now and then and need a complete break away from the computer. The turnover was more in reference to the employee rather than the customer. Running a staff at 65-75% capacity usually equates to happier environment. It SOUNDS like money wasted, but sanity of personel is a key priority for us anyway.

Yes, you did have two rants there. I agree with you that 1,500 accounts per tech is high. It is happening though. You cannot sell $0.99/month hosting and pay a tech if you don't have 10k accounts per tech. It just doesn't financially work.

I like your strategy of 100 per server for shared and 12 for ecommerce. Your ecommerce pricing structure allows you to accomplish this. Your shared service with margin allows you to do it as well. You also do not provide unlimited everything (I checked your site). Clearly you have websites that understand quality.

Not everyone has that. Many of the datacenters that I do consulting work for has 140 servers per tech. They are mostly made up of dedicated server clients so the support load is not that high; however, shared hosting companies now a days with overselling and competition has to increase customers/tech. I was there once on a smaller scale I know how to lose my mind.

 

[handsonhosting]

Doing homework on me eh?

Limiting things means less profit, but overall happier customers, and in the world of the hosting business, customer service and satisfaction are the main things that separates anyone these days.

Unlimited is a myth as you know and those low priced packages with every feature, bell and whistle just can't be supported if a site becomes popular.

Years ago (early and mid 90's) when we were all dealing with static sites, sure you could cram a few thousand accounts on a server, but these days with blogs, shopping carts, forums etc, a low end server oversold to the hilt screams "run away" faster than the Knights that say Ni I miss those days

 

[Zachary McClung]

Doing homework on me eh?

Limiting things means less profit, but overall happier customers, and in the world of the hosting business, customer service and satisfaction are the main things that separates anyone these days.

Unlimited is a myth as you know and those low priced packages with every feature, bell and whistle just can't be supported if a site becomes popular.

Years ago (early and mid 90's) when we were all dealing with static sites, sure you could cram a few thousand accounts on a server, but these days with blogs, shopping carts, forums etc, a low end server oversold to the hilt screams "run away" faster than the Knights that say Ni I miss those days

Doing the research is what leads to solid discussions. A low end server what is that? When I was in the hosting industry I always had customers on dual-core, 2 GB or better. When the dual 5130's, 4 GB came out we switched our boxes over to it. We were one of the first. It costed a pretty penny but your write customer service is the way to go. We ran 200 - 250 accounts per server. They always were below .2 with plenty of room for bursts. I never could say lets jam 1,000's of service on a P4. I'm just saying it was possible. Honestly, I'm pretty impressed to see some still doing what your doing Conor. There are very few hosts like us still around.

 

[romes]

My friend told me this I'd say a year n half ago...Not sure..might've been a lil longer but yeah, 100 accounts per server is not a lot. If you buy a good server ranging from $300-$600 per month you can have over 400-500 accounts per server and still have optimal performance. If you have a good server admin and good server, you could increase that number to 600-800 clients per server.

Servers now-a-days are cheap and still maintain their power which is good for us hosts...This means more clients on a server and higher returns.


Also, for those who can afford it, it is good to have a back-up server that would have the exact files your current server has but the only diffirence is that when your main one goes down, all you have to do is have your clients point their domains to the other server and their sites will be just fine. In my 5 years in the industry I've seen a few hosts do this and it works. But, a server failing is not likely as servers are maintained to the highest standards (depending on your DC).

 

[Zachary McClung]

My friend told me this I'd say a year n half ago...Not sure..might've been a lil longer but yeah, 100 accounts per server is not a lot. If you buy a good server ranging from $300-$600 per month you can have over 400-500 accounts per server and still have optimal performance. If you have a good server admin and good server, you could increase that number to 600-800 clients per server.

He was off by a few hundred I suppose. Even for a year and a half ago. The concern is not how many clients you can fit per server. When it comes to reselling staff is going to be your biggest cost. You can find a decent Dual-Core, 2 GB that will hold 100 sites on it, charge $5.00 per month, and make $350.00/server. However, your customers probably will not receive support at all because you will not be able to afford the tech. The server is really a small factor in shared hosting.

Servers now-a-days are cheap and still maintain their power which is good for us hosts...This means more clients on a server and higher returns.

I do not agree with this one bit. Cheap and more clients per server in the current market doesn't work. Your dynamic sites are using 10x to 30x more power than what they use to and server power has only increase by 4x - 6x.

Also, for those who can afford it, it is good to have a back-up server that would have the exact files your current server has but the only diffirence is that when your main one goes down, all you have to do is have your clients point their domains to the other server and their sites will be just fine. In my 5 years in the industry I've seen a few hosts do this and it works. But, a server failing is not likely as servers are maintained to the highest standards (depending on your DC).

Having clients point their domains to a second server sounds like a kidde way of accomplishing a failover situation. Most find backups on a second hard drive or a remote backup up just fine. If they were to have to repoint their domain you would be looking between 12 - 24 hours worth of down time. They do make a load balance solution that solves this problem; however, you cannot do it for $5.00/month that is for sure.

 

[romes]

Well, having your clients point their domain to the other server would be temparary till the problem on the main server was found and fixed. Also, this is just a added feature that most clients would like. This means peace of mind when it comes to your site.

Yes, this can not be done for $5.00 a month. This solution would be for clients who pay a little more. Anyways, just throwing it out there lol

 

[handsonhosting]

A low end server what is that? When I was in the hosting industry I always had customers on dual-core, 2 GB or better. When the dual 5130's, 4 GB came out we switched our boxes over to it.

Sadly the low-end approach is something that we see every day (as our clients come to us from those hosts). Hosts running P4 2.4ghz with 512MB ram and trying to operate a few hundred sites on the server. Or the Dual P3 700mhz, or the Gasp - RAQ4... yes, they are still in operation Often it's the mentality of "if it ain't broke, don't fix it", and while I agree to that mentality (for the most part) there comes a time that change MUST happen. Running a software script from 2002 that is still using PHP4 and can't operate on PHP5 even after you've made thousands of dollars in your ecommerce store just doesn't cut it. Not only is it risky, but it's against PCI Compliance rules and is putting data in a compromising position.

I do not agree with the statement by Romes where you could put 400-600 customers on a server.. or maybe it's that I don't agree that it SHOULD be done on a site. Maybe it's different target markets. Most of our customers operate online stores and utilize between 700 and 10GB of data. While we do have many older customers that are actually using less than 1MB of space, the days of those customers are a thing of the past.

All our new setup machines are Quad Core Xeons, and even the older ones (over the past 2 years) were all Dual Core systems. Users that have been with us longer than 2 years have likely MOVED servers within our network as we rotated hardware out and retired older machines. Customers like this as it puts them from a Single Core system to a Quad Core - and while they may not see it instantly, they recognize the numbers when it says they are getting 4 times as much power for the same price, and WE did the move with less than 15 minutes interruption.

But, a server failing is not likely as servers are maintained to the highest standards

A server WILL fail at some point, and more often than not, it's failing within 2 or 3 years. Either memory, hard drive, cpu, motherboard, powersupply etc.. something is going to die, it just happens. The question is how fast can you diagnose it and fix it. The backup option is a powerful tool, and even more powerful if you can include it within the cost of operation. We spend tens of thousands of dollars on our backup solution and give it away to customers for free. We could be making additional profit with that money, but a happy customer will bring you twice the profit (and referrals) in the long run.

Hmm.. somewhere I went WAY WAY WAYYYYYY off topic on this thread - I'm sorry guys! Maybe we need to split this one up?

 

[romes]

Well, loading compacity for a server varies between what kind of clients you have. 90% of my clients do not even use 200mb of space and 1GB of bandwidth per month. So, I could have 100-400 clients per server and not have a problem. Also, I am not the kind of host to try and stuff clients onto a single server just to make a profit...Once the server being used reaches a certain point I cut off all signups for that server and get a new one.

I was just throwing those figures out there. If you have the right server with the right specs, having a lot of clients on a single server will not ruin performance...all I am saying.

 

[handsonhosting]

Agreed.

As long as you're diligent on it, and are actually paying attention (unlike what most hosts do) then putting additional clients on there is not a problem.

Having an account limit of say 100 or 200 is relative for sure. 100 people with moderate sites using 30-50GB bandwidth is vastly different than 200 accounts using 5GB bandwidth

We're on the same page Romes

 

[romes]

We're on the same page Romes

lol Yeah

 

[hostton]

I have 2 staff and now looking for another staff to handle marketing.

 

[SenseiSteve]

Growing to where you need more support staff is a great place to be. I know in the copier industry, it's customary to have one technician to 300 machines. I'm actually trying very hard to grow out my department so I can hire another sales rep - not there yet

 

[siforek]

I've tried hiring staff in the past, but it's never worked out. I've always concluded if you want something done right, do it yourself. Of course I'm not dealing with thousands of tickets a month..

 

[hostton]

I've tried hiring staff in the past, but it's never worked out. I've always concluded if you want something done right, do it yourself. Of course I'm not dealing with thousands of tickets a month..

What problem you are having with them?