
|
View Full Version : Cheap Colocation - Potential Startup...thoughts?
WTFHosting 04-29-2001, 03:03 AM Alright, we're thinking about starting a small colocation company. Lease space in a neutral fiber facility with a dedicated line. We also know how competitive it is out there.... So before we invest any time, or money, we'd like to know your thoughts on our offerings. <edit> Our niche would be in the very low-cost unmanaged colocation market, where users are experienced with administering their server and wouldn't require any services from us. </edit> I know the people here have a tendency to tear new companies to shreds...so please tear softly. :)
Again...based on your feedback we will decide if we're even going to attempt this...and/or what we need to work on.
http://www.inthrive.com
<edit> Just some info included on the site:
FREE Setup & Installation
FREE Rack space (up to 2U)
FREE Manual Reboots
FREE IP Addresses (up to 32)
FREE 24/7 On-call support
99.9% Network Uptime Guarantee
3 Month Minimum Contract
No content restrictions (must be legal)
Pricing by Kbps is on the website.
</edit>
Note: The site isn't finished and is just up for information.
I don't think this is technically advertising...but I stuck it in this forum just to be safe.
MattG 04-29-2001, 09:58 AM Not sure there will be any objections to the rates - which are quite low . Would like to hear how you plan to support the large number of signups at those incredible rates...
Best wishes
JeremyL 04-29-2001, 01:09 PM I really don't see what their is to support. I am sure that these are completely unmanaged and that they just plug in what you send them for the server and turn it on. What that takes 30mins at most if the setup is ready?
If they do offer any tech/admin support other than reboots I am sure they will charge somwhere near $100-150/hr and can easily have an employed admin do this and can hire more as needed since they will be getting paid by the hour.
So the really the only thing else they have to worry about is connectivity up to the rack, but if they are leasing the space from a large NOC I am sure that the NOC is handling that part as part of the contract (I may be wrong).
Then again the the initial investment to do this would be large, so unless you got some extra dough to loose till you get the customers then you might not want to do it.
I did notice on thing on your page, you said "up to 1U" on free Rack Space. What does that mean?
cbaker17 04-29-2001, 01:29 PM I may be wrong but havent you placed this exact same message in like 2 other forums??? If im wrong im sorry if im right please keep your posts to one forum thanks.
WTFHosting 04-29-2001, 02:36 PM JeremyL,
Thank you for your comments. The idea was to offer unmanaged colocation at extremely low prices. Since we will be in a neutral fiber facility, we will be talking with several backbone providers to negotiate the lowest price. And you're right, the investment would be quite large, which is what we are planning for. Regarding "up to 1U free rack space" is we will not charge for the rack space if the customer's server is 1U (one rack unit). Since rack space is dreadfully expensive, we'd have to charge for any clients sending larger servers.
cbaker17,
I have only posted this in the advertising forum, even though it's not an advertisment post. (?) <edit>I think you're referring to jazon007 who has posted his question wither or not building a colocation facility in his "spare" warehouse space would be profitable</edit>
MattG,
Our target niche would be the low-cost colocation market, where users are experienced in managing thier own server. As JeremyL pointed out, there would be a large investment to start, and it wouldn't be a bad idea to limit growth until we can recoup some of our start costs. I'm not sure how many sign ups we might get--most clients are weary of new companies, especially those offering very low rates.
WTFHosting 04-29-2001, 08:39 PM Maybe I should make this more of an interest check. If we were to say, offer these services in a few months, who would be interested? What points would concern you? Do you feel the pricing if fair, if not, what would you change? We're doing this research to see what the community would see as the perfect unmanaged co-location host. Any and all comments are appreciated and will be taken into consideration.
dektong 04-29-2001, 09:42 PM WTFHosting,
I am sure I saw the price was $1.5/GB before... right?
Without knowing additional information, I can't comment on the price. What is your total connectivity (i.e. the total amount of bandwith your switches/routers are connected to? What is the total amount of bandwith the NOC has (DS3, OC3, etc). I assume these are capped bandwitch, not burstable, rigth? What bandwith measurement method do you use?
OK, these are my preliminary questions...
<edit>
Just did a traceroute, you are using CARI.net as your provider? I am not sure about it, but I think they are pretty good (in terms of reliability, connectivity, and support), aren't they?
</edit>
regards,
-dave
WTFHosting 04-29-2001, 10:14 PM dektong,
We're still experimenting on the pricing. It's really completely dependant on the cost of leasing space in the neutral fiber facility and the cost of the dedicated line (what we can negotiate). Currently our pricing is for capped bandwidth connections, although we're open for suggestions. Which bandwidth measurement method would be considered more desirable for potential clients? With our current structure, we would just cap the ethernet ports at the speed purchased and wouldn't need to monitor how much traffic actually passed through the port.
Right now we're considering Colo.com and Equinux for our neutral fiber facility. I'm uncertain of their total network bandwidth, but around 7 major backbone providers do have a presence in their NOCs. We would most likely start out with a dedicated line around 10mbit, and slowly increase bandwidth on an as-needed basis. This allows us to keep startup costs low and expand as our customer base grows.
As far as Cari.net, the site Inthrive.com is hosted on is by no means where we would be selling co-location space out of. As I've said, this is just an inquiry into interest and looking for suggestions so we are by no means set on moving forward with anything. However, we have had pretty good luck with them.
cimshimy 04-29-2001, 11:22 PM If you do this, you should go with colo.com in NY, because I live very close to their New York facility and I could come and work for you over the summer.
Andrew
dektong 04-29-2001, 11:33 PM Doesn't it make more sense to go with whatever NOC closest to him, not to you? :D
cheers,
:beer:
WTFHosting 04-29-2001, 11:56 PM Sorry...NY is a little far to fly to install new servers... :D
Would have to be the Dallas, TX facility...
cimshimy 04-30-2001, 12:07 AM That's why I was kidding. :D
Andrew
dektong 04-30-2001, 08:23 AM Originally posted by WTFHosting
Currently our pricing is for capped bandwidth connections, although we're open for suggestions. Which bandwidth measurement method would be considered more desirable for potential clients?
Personally, I like my bandwith not to be capped at all since I want people to be able to access my site at the highest speed possible. This is important for those who get hundred of thousands of hits per day in which it is very likely at any given second quite a few people access the server. Now, capped bandwith in this scenario will probably end up being too slow for those people trying to access the server. So if possible, offer the ability to burst to 10mbps, for example. With respect to this, more people will prefer the average bandwith measurement and most people will avoid 95th percentile based measurement. In fact, I know many hosts here offer average bandwith measurement although they are being billed by their upstream provider based on 95th percentile usage. This way, they will get more clients....
Depending on your target market, some people will think 120GB of bandwith for colocation is just too much for them. Some people will utilize lower bandwith consumption (20-50GB), so they will probably find other place to be better for their needs. If only you can offer this, then you will (of course) get more customer. I guess you can only afford offering free rack charge because you want your customer to buy at least 120 GB of bandwith to cover the rack charge? It's fine with me... again it all depends on your marketing strategy.
As I've said, this is just an inquiry into interest and looking for suggestions so we are by no means set on moving forward with anything. However, we have had pretty good luck with them.
I never heard bad things about CARI.net either.... :)
Anyway, good luck in your new venture! All the best for you....
cheers,
:beer:
WTFHosting 04-30-2001, 11:36 PM From the looks of things I'm going to need a handful of clients lined up before attempting anything. Would anyone be interested in colocation with us in a 2 to 3 month timeframe?
sergio 05-01-2001, 01:28 PM Originally posted by WTFHosting
From the looks of things I'm going to need a handful of clients lined up before attempting anything. Would anyone be interested in colocation with us in a 2 to 3 month timeframe?
But what about not capped sollutions?
WTFHosting 05-01-2001, 03:17 PM sergio,
We originally wern't going to offer burstable bandwidth, however if there's significant interest we can modify our pricing structure.
<edit> Actually we could probably offer both. Offer fully burstable bandwidth and charge by the monthly average for clients wanting burstable bandwidth. And offer capped bandwidth for clients wanting to make sure they don't go over budget...thoughts? </edit>
SaluHost 05-01-2001, 07:43 PM I would think most people are going to want burstable. You have the cheapest plan at 81GB transfer, but only 256Kbps. There's no way you would ever get 81 GB transfer at 256Kbps, you'd have to be transfering full speed constantly. Plus 256Kbps is what, like 32k? That's pretty slow IMO, so you should definately make it burstable. Just my opinion though.
sergio 05-02-2001, 03:59 PM Originally posted by WTFHosting
sergio,
We originally wern't going to offer burstable bandwidth, however if there's significant interest we can modify our pricing structure.
<edit> Actually we could probably offer both. Offer fully burstable bandwidth and charge by the monthly average for clients wanting burstable bandwidth. And offer capped bandwidth for clients wanting to make sure they don't go over budget...thoughts? </edit>
You can try to sell prepayd bandwidth usage in blocks. Like 2$/Gb in blocks of 10Gb/month.
dektong 05-02-2001, 04:09 PM Originally posted by SaluHost
Plus 256Kbps is what, like 32k? That's pretty slow
I should agree with you. Especially, for those offering web hosting, capped bandwith can be really bad call. Consider you are hosting 100 accounts, each accounts will probably have some people accessing their sites, so at any given time your server could be accessed by a lot of people. Let's say, 20 for example... and these 20 people are sharing one 256kbps line? Not good... In fact, I knew of a hosting company who has 800 clients or so and their connection was capped at about 256-386 kbps.... Very slow!
cheers,
:beer:
dektong 05-02-2001, 04:15 PM Originally posted by sergio
You can try to sell prepayd bandwidth usage in blocks. Like 2$/Gb in blocks of 10Gb/month.
Then, I am pretty sure he can't offer free colocation :D
WTFHosting, yes... offer both burstable and capped bandwith. That will be the best option everybody can get.
cheers,
:beer:
WTFHosting 05-02-2001, 06:46 PM The one thing I've run into is clients seem to like to be billed by actual transfer (GB/month), however we're billed by our upstream providers in Mb/sec...
Should we simply take the average monthly transfer in Kbps/sec and bill for the bandwidth tier the client falls into? It does seem to affect our pricing structure...since we show price per GB of transfer, yet the actual transfer will be lower...it's hard to show pricing by $/GB if we're charging by average Kbps...see where it gets tricky? Thoughts?
sergio 05-03-2001, 10:05 AM Originally posted by WTFHosting
The one thing I've run into is clients seem to like to be billed by actual transfer (GB/month), however we're billed by our upstream providers in Mb/sec...
Should we simply take the average monthly transfer in Kbps/sec and bill for the bandwidth tier the client falls into? It does seem to affect our pricing structure...since we show price per GB of transfer, yet the actual transfer will be lower...it's hard to show pricing by $/GB if we're charging by average Kbps...see where it gets tricky? Thoughts?
May be you should make discount for Gb downloaded in not peak hours?
rodcon 05-18-2001, 01:37 AM Great Idea. I have been thinking the same thing, but along different lines. Im more interested in offering a low cost unmanaged Dedicated server deal myself.
I figure colocation would be a good source of income during startup since there probably would quite a bit of space going to waste at first.
I am also located in Dallas. Keep us posted, and good luck.
Rodcon
rodcon@rodcon.cncdsl.com
|