Flashcom's Verizon customers sold at $100 a head > I tracked down Moon Global

reply to Anon

Re: I tracked down Moon Global

Not so. In the beginning, PacBell created major issues because they wanted to bill for the loop, and we would bill for the ISP services. This did not work out very well. In the end, the ISP was charged a lump sum for the customer loop and the service.

As to the bandwidth and colo, there are alot of providers who have recently empty cages, racks and partials. The SMS500 is like a 5u unit and the supporting equipment such as a few sun Netra servers (for DNS and Radius) and out of band management would take up like 12-15u of space. This could easily be had in half a rack or less in a shared colo room. The bandwidth costs can be negotiated until day turns into night. The costs will obviously be varialbe and as such your monthly income will vary. However, a 50K buffer leaves a lot of room for bandwidth charges.

The KEY (and would have made this much more clearly had I known you were going to post this on the front page ) is that without Advertising, without the staff that it takes to shepard an order through the provisioning and installation process, and without the need for expensive back end automated processes for billing, trouble ticketing, and accounting (all outsourced) DSL CAN be a profitable business.

reply to justin
I have to side with Justin here. 5k a month is a GROSS under-estimation of costs for co-location and bandwidth. 5k a month is like 50$ a month to the big dogs. Costs for that kind of pipe start along the lines of 2-3k for 2mbit and up. Nowhere the near the full speed a DS3 offers. Co-location is not chump change either. All in all the costs would really be much much higher then 5k.

I think this story is a nice fairy tail but in practice, its just that, a fairy tail. There are so many more aspects to the business then getting customers, bandwidth and location.

One company that seems to be doing things right is Flex.Com in Hawaii. If only more ISPs followed the same practices they do...

ohhhblahdee ohhhbladah...
--
..::drum and bass cannot be stopped::..
[text was edited by author 2001-04-05 10:28:33]

I beg to differ. You would really be surprised at how little traffic 1700 largely residential users pass on a sustained basis. Most of the traffic req could easily be satisfied with a 5-10 meg ethernet handoff. lets say your DS3 and bandwidth costs amounted to 10K a month. You still have a HUGE profit margin. The actually COST of the DSL line is in the human support factor. Remove the support factor or get it on a managable cost basis (such as outsourcing) and you can easily turn a profit. The problem is that BIG isp's employ dozens of customer care people, dozens and dozens of tech support and provisioing people, and a good sized application development group to work on the systems and an accounting department. Not to mention big C-Level salaries which easily reach into the hundreds of thousdands of dollars with kick ass bonus plans. You wanna know why the ISP's are going out of business?

In this scenario, ONE PERSON could actually run this ISP and outsource everything at a relatively FIXED cost thus ensuring that the margins are met and the bills get paid. Give me a copy of ICVerify and and Excel spreadsheet and I could personally bill 1700 customers in a weekend.

The real costs are behind the lines...

The *theory* is great. I still think the practice will fail. Even if you have prequal customers, you still need to advertise. How long do you think things can stay as is and still turn a profit? Costs change daily. You can negotiate costs all you want but they will go up. You have to advertise and seek growth. These things cost time and money and for a small while, MAYBE one person can manage it all, but in the long one you need people. And people are just the beginning. Until the model you speak of is done and is shown to be profitable I will continue to see the idea as nothing more than a fairy tale. Its easy to say companies have too many tech support personnel etc but the fact is, with the current state of end users today, companies need those personnel.
--
..::drum and bass cannot be stopped::..

Dear Friends,
Fascinating discussion.
The theories behind fixed-cost capitalisation, and growth based forward depreciation bases are being compared in a practical setting, which, if the costs were predicted accurately for the fixed cost model, would allow for a relatively high margin.
Who among you wonders how a "trading firm" survives, when there a often monthly losses, which would appear unsustainable?
Leaving aside covered fixeds, there is always revenue from the co-commissions on trades.
Thus, trader joe loses $20k, while costing himself commissions, whereby the firm supplying his desk may take have of those commissions going to the clearing firm. Thus, if Joe were to have lost his money on trading 400k shares, at around 1.5c per, total commissions before rebates= $60k, thus the firm, garnering 50% of the commissions, before fixeds has a profit before rebates. After rebates, costs, &c, although the individual is in deep shit, the firm is about even on its capital, on a cost basis. Given even marginal risk control, most trading firms do well for the limited partners. Analagous to the EU's fifty dollars per mensam.
Although no one, myself included, would recommend deacon's business model for the fainthearted, there are many theoretical reasons why it would succeed. The cost variable seems to be the fact in dispute here. My own experience is that the costs would probably be higher than deacon's estimates, thus protracting the profitability window. Therein lies the problem. SAP would not view the monthly costs as having been investment, but recurring. Nor can I foresee the ability to depreciate the original cost of buying the prequals, since, if deacon's analysis is correct, the value thereof would remain for accounting purposes, constant. Thus, the purely EBIDTA model which he employs as his basis for asserting profitability.
Should an unindemnified expense occur, the cost is immediately apparent against the current income, and thence, reduces operating capital.
The trading firm system works, because the firms, as does Deacon's model, rely upon outsourcing, such as AT financial. When AT's servers are down, the trader is out of luck. Given the miniscule amount of down time for such a service, losses are infinitesimal. Now, look realistically at DSL. Can anyone name one mail-server, or internet based mail service where one would willing put mission critical communications?
Thus, Deacon's model has one important hole, which lies in the bandwidth philosophy expressed. While true that I, as a home user, use very little bandwidth, I, as do many, rely upon my connexion for my scheduling, etc. The argument may be made that this is foolish, but I would wager that more of us do this than not, to say nothing of how angry I would be if my children could not access their homework, also 'net based. Thus, the "pure" residential user is apocryphal, and multiple redundancy or indemnification becomes necessary.
Then the final CLEC issue, which also affects SAP remains, can one base future profits upon the fixed costs, having reduced initial capital expenditure based upon buying assets at pennies on the dollar? My guess is, with decent luck, probably. But as easily as Deac says. Not likely. The run-out date for profitability would have to be clearly aligned with the cash burn rate.
In short, a very good idea for an entrepeneur seeking reasonable income, who is not betting the rent money for this year, or his bairns' meals. A dirt-cheap chance to drive a Porsche by 2002? Circumscript "maybe." A business with minute personnel costs, low initial capital requirements, and a source of sustainable income, where income and no growth is wanted, may indeed work.
Could one sleep while running it? Not likely.
All good wishes,
yazdzik

reply to Anon
Great concept Deak,.....but what are you going to do when those companies that you outsourced to start breaking their contracts.......because someone they were outsourcing to went bankrupt..........rotflmao

Guide.....guide.....guide.....where is my tech support......guide....guide......guide
--
The Future Is Purchased By The Present!

reply to Anon
ignore this dupe post
[text was edited by author 2001-04-05 13:23:41]

reply to Anon

said by deakonblues2k:

---

Not so. In the beginning, PacBell created major issues because they wanted to bill for the loop, and we would bill for the ISP services. This did not work out very well. In the end, the ISP was charged a lump sum for the customer loop and the service.

---

reply to Anon
Lets pray that you never have access to a copy of ICVerify,

and you definately need to get some training in Excel......lmao
--
The Future Is Purchased By The Present!

While I see a lot of posters, with the exception of deakonblues2k none of you appear to be in Southern California or running an ISP there. Now some of the things in his plan I can't comment on.

1. Cost of DSL loops from GTE: $32 and change for ISP with ~400 customers--bump that number up and you get better profits.

2. Cost of Frame DS-3 $2700 for a 22mbps CIR--note this is just the backhaul for the DSL loops.

3. As to Co-lo costs--I don't know the cost exodus, but rack space in my neck of the woods is about $1500 per month.

4. Internet bandwith costs--Typically are cheaper at colo spaces since they can use a throttled 100mbs lan connection--Broadwing does this at 1 Wilshire. 3 Mbps (dual ds-1) is 1800 from Sprint. 9 Mbs was something like 6500.

5. I would definately not outsource mail or webhosting. One of the thing keeping the smaller ISPs alive is that they can offer a variety of services.

6. Outsourcing level 1 tech support is a good idea. That type of support is easily replaceable.

7. As he says expansion is the real cost. If you could walk into 1700 customers with an acquisition cost of $50 it would be a sweet deal. I would then only advertise as needed, however word of mouth seems to be the best way. (I include this site in word of mouth. As long as you can keep a pool of customers at the 1700 mark you could make the model work.

reply to deltat2000
Sheeeeeeeeeeiiiiiiitttttttttt. All you looking for a way to make a run at it. Watching the market lately? Intel, Microsoft, SunMicro...all getting hit hard. NASDAQ from 5000 to 1700? Don't you think the trickle-down effect will hit everyone in the Internet biz? When the big guys start losing money, they look to OUR pockets. More bankruptcies to come before the dust settles. Hang on to your bucks and do your due diligence as Justin was suggesting.
--
With apologies to Friedrich Nietzche, "Whatever does not kill me is still a pain in the ass!"

reply to justin
First, DeaconBlues' comments about a profitable future, and the new model where ISPs are built or grow on acquisitions of failed ISPs, is basically a good and sound concept. You will see a whole new breed of companies doing ISP consolidation, and moving forward serving existing customers without the debt and red ink run up by failed marketing and bad business plans.

But I also read in the post some of the bad planning that gets ISPs into trouble. The devil is in the details, and in the end, the Flashcom Verizon customers did not pencil out for most potential buyers and investors sponsoring a buyout through the court, because the numbers would not justify the real costs to set them up. While they claim 2000, we counted 1648 as of December 1st, and about 200-300 in attrition each month. In the hearing, the counsel for Flashcom estimated 1000 circuits could be migrated by April 13th, and I heard from Verizon today the number of live circuits is now under 1000. I think that it is optimistic to expect all will re-sign, and at best you can expect 500-600 of them to become the new ISPs paying customers. No matter how wonderful Moon Global might be, there will be a percentage that will say "screw you!" and go with their own choice no matter what.

I am going to explain why 500-600 customers do not offer the profit that DeaconBlues suggests is there. I do want to also make sure people reading this that what I lay out here does not apply to Moon Global, which has existing Verizon infrastructure, circuits, and adding these customers are probably profitable in their scenario, we will assume that they met Verizon's and the courts requirements to complete the deal. What I don't see is it working as a stand-alone startup.

One of the obvious issues in this business that I think gets overlooked is reaching critical mass, where you have enough customers to meet your expenses and cross into profitability. Everyone wants a piece of the action, everyone wants to run their own show, and what we have found is, the ISPs that are hanging in there want outrageous sums to sell their operation legitimately. They are profitable, but the employees pull up each day in new cars and the owners pull up in rusted pickup trucks held together by bond-o. This is because nobody has reached their target numbers, and the pricing wars have lopped off too much of the profit in the DSL market. (We still have Verizon undercutting their resellers in this market)

The small isp's challenge is not that they can't be profitable or have terrible business plans, but there are hundreds of them, duplicating efforts, each with a fraction of the number of subscribers on their backhaul circuits that they expected or can handle. There are hundreds of ISPs in this situation. And the ILECS are getting rich on multiple backhaul circuits, all at 20% utilization. You have to wonder why Covad never figured this out and actually walked all these "underperformers" to Zyan and CAIS.

In Verizon land, you charge $50 of which the phone company gets $32.50 and $2 for USF. So a typical ISP might squeeze a $15 per month gross profit. Then you have bandwidth costs which everybody outside of the business tends to overestimate actual traffic levels, but that's a different story. Add the usual support, email, news, and other overhead costs. But everyone forgets the backhaul costs. A DS3 can cost between $1600 and $4200 from ILECS, but add the meet-point mileage and then colo-center cross connect, and you can be at $4000-$5000. (Covad, Rhythms, Northpoint, and New Edge all provide delivered backhauls at about $4000) This monthly recurring cost should be divided by the number of live, paying customers.

While everyone thinks "hey, you can jam 2000 customers on that circuit, piece of cake!" and this puts the per head costs for backhaul at less than $2 per head. But it is the actual number of live customers affects your bottom line and monthly cash flow! The grim reality is, there are many ISPs with 300 to 500 customers on their Verizon T3's and this means the backhaul cost can run $10-13 per head. At least bandwidth can be scaled on a usage or per meg basis from colo centers, so the real bottom line killer is that backhaul.

It is really easy to see with colo space, bandwidth, equipment leasing, backhauls, cross connects, plus technical support, credit card fees (3%+),IP address management, mail and news services could easily run 10-12K per month. If you net 2000 subscribers it's worth it, but net 600 subscribers and at $49.95 at least, you're upside down almost 3 grand a month by my calculations, and that is if everyone pays and the "one man show" isn't drawing a salary from the operation. Actually, he'd need a good one from elsewhere to kick in the shortfall.

But Flashcom got $60K of your money, and there is still the circuit supercedure debt issues to get through with Verizon. They might float it for awhile but ultimately they can shut you off; they have you by the short and curlies. The only remaining unknown is how many of the remaining Flashcom customers are at premium service levels, ie, those left at Gold or Platinum that would greatly enhance the revenue figures if there are a healthy percentage. But, given the numbers provided by the court, it wasn't promising enough for anyone to bet the farm.

This is why this "deal" only appealed and worked for a few local ISPs who #1, already had ATM circuits, equipment, and ample space left to accommodate them, #2, have a working provisioning process in place to migrate hundreds of customers en masse.

The ultimate model that will work as DeaconBlues suggests is a well-funded company that takes this deal along with dozens each week, rolling up and consolidating ISPs that are in trouble or hopelessly overbuilt. The ROI could be there, but not given the realities of this particular Flashcom deal, starting up from scratch. If something looks too good to be true ($100 a sub)it probably is.

Finally, potential bidders were undoubtedly spooked that Verizon Online keeps extending their $39.95 per month, free camera, and free modem and installation "Big Deal". It was supposed to end in March, but now runs through the end of May. Given that Verizon now charges ISPs $200 for the modem, and $60 for the install, and most independent ISPs can't meet the $40 price for reasons listed above, Verizon sales will continue to be flat. All independent ISPs report sales have dropped to a trickle. It is uncertain what control Moon Global will have to retain customers after they are migrated. There is nothing stopping them from not paying their bill or sticking around for 30 days then switching to Verizon Online.
--
Longest sentence in the English language: "I do."

Wow! I am so excited! This thread has brought about some of the best information and discussion I've seen lately! So Very Kewl!

Ok, just a few last bits. As to outsourced locations for customer services...

Critical Path which bought supernews last year is your mail and news provider. In reality you would be simply amazed at how many fortune 500 companies are using CP. And if not CP, then try Comtouch. Either way, this type of mission critical mail outsourcing is the way many many many are going.

As to Technical support. Last I checked Telocity was outsourcing virtually ALL technical support. Tier 1 2 and 3. Once again, a fixed cost per customer is the key.

As far as webhosting is concerned, Flashcom did not offer it and as such, we would not have provided it. There are however, tons of companies that have this as their business model.. Try Netopia.

For a success story where this model has worked, look at MM Internet in Long Beach CA. They have a fantastic rating as far as customer satisfaction, have remained small and service ONLY GTE customers in southern California. Additionally, they do very little advertising. Check it out. This model WILL work. Not for me though...I'm on to something else!

Ok.....discuss amongst yourselves....

Gee Dork,opps meant deak, can we meet and discuss this new scam of your.
Perhaps we can hook up with deltat2000 when he comes to visit you.
Then we can both show our appreciation for your great c/s skills.

reply to Anon
Interesting this one post generated some lengthy replies...
[text was edited by author 2001-04-05 17:21:44]

reply to Anon
Gotta go with the Deakon on this one. If you know your way around the ISP/telco biz you can cut some pretty sharp deals on cross connects, loops and transit fees. Although the monthly pretax $'s in this case wouldn't be much it could be done as he's noted. I might put in my own mail servers rather than go with CP but that's based on not knowing what the per user rate would be. Anything more than a few $'s per EU and you are burning up profit versus the one time for hardware/software for the mail servers. Tech Support - you would have to outsource it and pray the network you've built is reliable, but the folks at Flashcom did build a reliable network so I'm guessing this one would be no different. I had 14 months of service and 1 hour of downtime.

reply to creepndth

am i missing something?

reply to basshive

Re: I tracked down Moon Global

said by basshive:

---

I have to side with Justin here. 5k a month is a GROSS under-estimation of costs for co-location and bandwidth. 5k a month is like 50$ a month to the big dogs. Costs for that kind of pipe start along the lines of 2-3k for 2mbit and up. Nowhere the near the full speed a DS3 offers. Co-location is not chump change either. All in all the costs would really be much much higher then 5k.

---

reply to NoConnect
Hey NoConnect............great idea....would love to get together with you an dorkonglue...opps...deakonglue..opps.

Whatever his name is....Guide..........guide..........guide.....
--
The Future Is Purchased By The Present!