Date: Sat, 6 Jun 1998 16:44:53 -0700 (PDT) From: Michael Dillon <michael@memra.com> To: inet-access@earth.com Cc: linuxisp@friendly.jeffnet.org, freebsd-isp@FreeBSD.ORG, iap@vma.cc.nd.edu Subject: Re: US West and RADSL (fwd) Message-ID: <Pine.BSI.3.93.980606162546.13151F-100000@sidhe.memra.com> In-Reply-To: <Pine.LNX.3.91.980606122248.20041D-100000@hypermall.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, 6 Jun 1998, Jawaid Bazyar wrote: > The cheapest DSLAM setup that can host more than a single customer and > scale to anything reasonable costs in excess of $10,000. Yes, you can get > a onesy-twosy Pairgain modem type thing for a grand or two, but do you > really want to pay $2K per port long-term? I didn't think so. Wait a sec. What if the customer is paying $120/month for that circuit on a 3 year contract? The $2k port cost is paid for in just under 18 months and the next 18 months is profitable. Now what if you are buying 100 of these Pairgain type thingies? I think most ISPs could negotiate some sort of discount based on that quantity. > Alright, the Denver metro area as an example has approximately 30 central > offices. Instantly, in order to reach the whole potential customer base, > you're looking at $300,000. Just in equipment. Sounds like a job for your local neighborhood bank. They love to finance equipment purchases by established profitable companies with a steady growth curve and a stable customer base. > Now you have to tie all > that together, in which case you're probably still looking at (minimum) > 30 T1 ports into an ATM cloud, at $400 per month each, for a total of > $12,000 a month. Some of this ADSL gear, like Westell's FlexCAP2 RADSL gear, can do pretty decent speeds like 2.24Mbps downstream and 1.088Mbps upstream. In some cities you may be able to tie together the COs with your own RADSL circuits. http://www.westell.com/products.html > Not to mention co-location/rent fees, anywhere from a > couple hundred to a couple thousand a month. Anybody looking to supply a whole city with ADSL is basically getting into the CLEC business and needs to be able to raise CLEC style funding even if they are going to specialize in IP only and use some cheaper technology as a result. > And yes, you *have* to hit the entire area for this to make economic > sense, because we're seeing loop qualification rates of 15% to 25%. That > means that less than a quarter of the phone lines coming into the office > are even capable of having DSL run over them. Depends on a lot of things including the DSL gear that you use. HDSL works on less lines than ADSL and CAP modulation works on more lines than DMT modulation. It also depends on the copper plant. Some cities are better than others. > So, take your existing customer base, divide by two to weed out those who > won't pay more than $20/mo for anything, divide by five to get the number > you can reach at all with DSL, and divide by two again for those who are > happy with their existing internet service. > > Are you *really* going to invest $300,000 cash and $12,000 to $50,000 a > month so you can serve DSL to maybe 1/20th of your customer base? Exactly > which "most ISPs" can afford that? Ohh, right. The *big* ones. You are making a lot of assumptions about the business model here that I don't think are justified. Yes it is true that you can't fling ADSL at the customer base and expect enough of it to stick. You need to plan carefully and for smaller ISPs it will not be possible to do anything other than be a DSL-enabled ISP of one of the LECs. But some people will be able to make a business case for supplying DSL service from just one CO. -- Michael Dillon - Internet & ISP Consulting Memra Communications Inc. - E-mail: michael@memra.com http://www.memra.com - *check out the new name & new website* To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-isp" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSI.3.93.980606162546.13151F-100000>