jgd@gatech.edu> (06/24/91)
jim@equi.com (Jim Allard) writes: Preface to comments: Several years ago one other engineer and myself designed and built the switching system for an AOS. The name will not be mentioned because quite frankly I can't prove everything to legal standards that I'm going to discuss. I will use this experience as the basis of my comments. Nor will I provide any further technical details, as these could be used to identify the AOS. I will also apologize in advance to those who have been cheated by the system I constructed. My 20-20 hindsight leads me to believe that we were selected to implement this system because we knew little about the practices of the long distance business. I know nothing of Jim Allard's company other than what's contained in his post. I commend him for his company's practices. The AOS I worked for is definately NOT his. However his defense of AOS systems falls on deaf ears. His is a minority of what I suspect to be one. > 1. Hasn't anyone noticed that AT&T LD rates have dropped dramatically > since divestiture? Does anyone really believe they (AT&T) would have > done that on their own? If you had a monopoly, would you? Wake up > and smell the coffee ... competition among other things have been > responsible for more realistic pricing in the LD market. There has > also been an increase in charges for local service (not necessarily > corresponding but understandable). And for which we can thank MCI and Sprint. Since most AOSes tack on many dollars in charges over and above the mileage and per-minute rate, they cannot by any stretch of the imagination be considered to have applied competative price pressure to AT&T. > 3. Every public IXC except AT&T provides 800 or 950 access to their > network as a convenience to their customers. AT&T steadfastly sticks > to 10XXX as the ONLY means for non-subscribed customer access (or > access by their customers from non-AT&T phones). If you owned 200 > switches at your hotel chain, would you give guests unrestricted fraud > potential? > Fact is, it cost us $1500 per switch to allow guests full AT&T > capability with fraud protection. We spent the money because some of > our clients' guests like AT&T ... go figure, but we believe in > service. Interestingly, only 5% of our AOS customers ask for other > carriers, a percentage of that for AT&T. We carefully provide > instructions for access to the caller's carrier of choice and various > other alternative billing methods. 95% of our callers CHOOSE to use > us. And yes, we clearly brand our service at the beginning and end of > our involvement (which we've been doing since well before it was law). > This based on two years of handling greater than 400,000 calls per > month. > 4. This business of an AOS connecting callers to the AT&T network is > a joke. While we can technologically do so, they refuse to take the > call, claiming possible fraud and high error concerns as some of their > reasoning. It would be a nice service to 'their' customers, and we > would have to pay origination charges for the entire length of the > call. Their position reminds me of the kid who takes his ball and > goes home if you don't want to play by his rules. Here is the meat of the problem. AT&T was forced to adopt these policies. While not a supporter of AT&T I do understand why they have become hardnosed. Let me outline some things we did at the AOS. And we were not alone. If a motel or other institution signed up with us, the first step was to install redirectors (devices that receive touchtone digits, store them, dial another number and regurgitate the digits) on all outgoing lines. These redirectors were programed to dial an 800 number into our switch. In most cases, we even intercepted local calls -- dialed via LD back to the local area -- in an effort to block ALL access to other carriers. Our switch received the numbers and tried to complete the call if possible by redirecting it out a channel on the T. If it failed or if operator assistance was required, the call was directed to an operator workstation. Our software interpreted the prefix digits if any (ex: 10288) and posted a message on the terminal as to what brand the operator was supposed to emulate. If 10288 was dialed, the operator was supposed to sound like an At&T operator. The women were given intensive training as to how to slur the phrase (ex: "T'N'T) so that they were not exactly being technically fraudulent. Whatever the desired carrier, the call was processed and billed by us with all the charges added on. The next issue was credit card validation. There are two levels of credit card validation. The first and most simple is algorithmic and the second is verifying that the account is good and the charges can be applied. The various credit granting companies and TPC all charge significant fees for credit card validation. Our bunch decided on a much cheaper way. All credit cards (except some now-obsolete AT&T business accounts) have checksums built into the account number. Each vendor (Visa, MC, AT&T, etc) use different algorithms but we knew them all. Most credit card vendors are identified by the first digit of the number. The AOS decided that since the profits were so high, they could forgoe most actual account validation and only perform the algorithmic check. They reasoned that they could eat the few bad accounts. There was one exception. AT&T/BOC calling cards. When our software received such a number, the switch grabbed an outgoing line, dialed an internal LD number using 10288, listened for the bong, fired off the PIN and listened for the response. If a "thank you" (digitized sound pattern matching) was heard, the credit card number was considered validated and the AT&T connection was terminated. Our switch then completed and billed the call through our trunks. If any other response was detected, the operator was given an approrpriate message. If the operator got such a message or if she wanted to not handle the call for any other reason, she could hit a function key which would put some simulated switching audio out to the caller while the call was dialed to 10288. Our software listened for the AT&T operator voice and upon the first silence thereafter, cut the caller over. The caller would likely never know he was using an AT&T operator. This was before the AT&T operators were taught to thank the customer. Even if he did, he would likely not know anything about AOS systems and would not be confused about the duplicate operators. In any event, we tried to bill the call too if possible. Thus the customer would get two bills for the same call, one from us and one from AT&T. Over 90% of the customers would pay the charge without question! What sheep we are. We were shown what I now believe to be forged documentation that indicated there was an agreement with AT&T to do this. I wondered at the time why we did not just tie directly to AT&T's system. Since our switch could churn out thousands of such "validation" calls per day, it had a huge impact on AT&T. Unfortunately AT&T made this task too easy with the nice bong and digitized voice messages that never changed. I know from discussions with people from other AOS companies that we were not the only ones doing this kind of "validation". I've observed that AT&T rarely does anything unless forced. The AOS industry forced them to revise their messages and their policies. > 5. My company DOES NOT apply any additional surcharges to calls we > handle, even though it's legal in many states and for interstate > traffic. The same stipulation is in every contract we write. Our > rates mirror AT&T's. We guarantee not to charge more than their > standard time-of-day discounted rates. In fact, our billing programs > round down (not up) to ensure compliance with the policy. Yes, we > also issue immediate credit for mis-dialed calls. I'm really glad to hear that. You are the exception. Most companies round up and some do worse. By "rounding up", I mean that if a call lasts four minutes and five seconds, it is legal to round that call to five minutes. At my AOS, I was told by the DP manager that their billing software (as opposed to my switching software) automatically added a minute to each call AFTER rounding. I think some attorney generals have had something to say about that practice. My AOS would also let the institution add any amount they wanted to as a "trade surcharge". Motels typically added $5. If a customer complained about any aspect of service, the policy was to automatically give credit. They reasoned that it was bad to attract attention plus since the profits were so high, it did not matter. > Is AT&T deliberately designing a card system which will create serious > customer dissatisfaction in an effort to pressure aggregators into > presubscribing to AT&T? I don't think the last legal shot has been > fired on this issue and I hope the public isn't that gullible. No, I think AT&T is trying to come up with a system that will not too terribly inconvenience their customers while preventing scumbag AOS and COCOT operators from abusing their billing and validation system. The tragedy is that we all lose from this experience. > I sincerely hope this does not come off sounding like a whining > step-child. We expect to be successful or fail on our own merits. If > we can provide the service at reasonable costs to the calling public, > on a level playing field, I say let the market determine the winners > and losers. Keep the politicians out of it, or we'll end up with > three LD phone companies about five years from now. That's not what I > call choice, and the playing field is far from level. You don't come across as whining but at the same time if the government shut down every agregator in the nation I'd be very happy. If someone else wants to build a physical plant and enter into the fray with AT&T, MCI and Sprint, I'd more than welcome it. The resellers are a whole 'nuther matter all together. Note to the net.nitpickers. NO I'm not a telephone professional. I knew enough about telephony to recognize a T from a local loop but I know little about long distance. Which is probably why I got the job. I got a crash course in the art while working on the system but that does not an expert make. If I use terminology incorrectly ... sue me :-) [Moderator's Note: Gee whiz ... and over on eff.talk not long ago you said ** I ** was an asshole ?? If you notice a snall difference between what you submitted and what I printed its because I removed your parenthetical gay-bashing remarks about AT&T employees. PAT]
"Mark R. Jenkins 619-458-2794" <marcus%cpva.span@sdsc.edu> (06/25/91)
In article <telecom11.487.1@eecs.nwu.edu>, emory!Dixie.Com!jgd@gatech. edu (John G. DeArmond) writes: > jim@equi.com (Jim Allard) writes: > Preface to comments: > Several years ago one other engineer and myself designed and built the > switching system for an AOS. The name will not be mentioned because > quite frankly I can't prove everything to legal standards that I'm > going to discuss. I will use this experience as the basis of my > comments. Nor will I provide any further technical details, as these > could be used to identify the AOS. Hmmm. I wonder if the rest of Mr. DeArmond's description (trapping access to LD carrier, faking operators) explains what happened to me on a trip across the US last September. I stopped in a hotel near Little Rock, Arkansas, and made a long distance call from my room. There was no way to get to AT&T by dialing (I know, I tried) so I called the hotel operator and asked for an AT&T operator. After some pauses, I got an operator which I "assumed" was AT&T and gave her my AT&T credit card number and completed the call. Upon checkout the next day I was given my room bill which included a charge for the long distance call I had made at some pretty exhorbitant rates. I explained to the person at the front desk that I had made the call through AT&T using my own credit card and that I wouldn't pay them (the motel) for the call as well. I assumed that they had just timed the call and had billed me a flat per-minute rate automatically assuming that I used their long distance carrier. They took the charge off the bill without too much arguing. However, I never received a bill from AT&T for that call. Mark Jenkins <Marcus@CPVA.SAIC.Com>| My views do not necessarily match yours. Science Applications International Corporation
Nick Sayer <mrapple@quack.sac.ca.us> (06/26/91)
marcus%cpva.span@sdsc.edu (Mark R. Jenkins 619-458-2794) writes: > After some pauses, I got an operator which I "assumed" was AT&T and > gave her my AT&T credit card number and completed the call. > However, I never received a bill from AT&T for that call. I wonder if a solution to the problem of fakers is to ask "Is this American Telegraph and Telephone?" If they answer 'yes' then either they are AT&T, or they're lying, n'est pa? Does this work? Nick Sayer mrapple@quack.sac.ca.us N6QQQ 209-952-5347 (Telebit)