Revitalizing the Pay-Per-Call System

by Brian McConnell, author of Beyond Contact: A Guide to SETI and Communicating with Alien Civilizations

For well over a decade, 900 numbers have been synonymous with nefarious services, dial-a-porn in particular. This is unfortunate, because the original concept behind pay-per-call was quite forward-looking and anticipated today's information economy by many years.

The original concept behind the 900 number, and its provincial cousin the 976 number, was to turn the telephone network into a billing and collection platform for third-party information providers. Information providers would set a price and would collect a generous share of the revenue of the tolls. The carriers would create the marketplace; information providers would compete to offer services.

A huge marketplace for ideas did indeed spring up as a result of 900 numbers. The only problem was that the ideas being exchanged were not dinner-table conversation. What could have been a foundation for an information services economy instead turned into a vast red-light district. This was a classic example of the law of unintended consequences. This did not have to be.

Before I go into a discussion of what went wrong and why, let's first look at a simple example of how badly telephone companies are screwing customers, and how revitalized pay-per-call services could, paradoxically, fix this problem.

Directory assistance is a mundane but heavily used service. We all use it from time to time. We also overpay for this service by huge amounts. Most cellular companies now charge $1 to $1.50 per call, for a service that is heavily automated. The carriers do not provide this service. They outsource directory assistance calls to third-party providers. They also limit competition to this sole provider on their network. They keep most of the money, while sharing a pittance with their vendors. The real cost of providing this service is much lower than the retail price. Consumers spend roughly $1 billion per year on directory assistance calls, several hundred million of which is excess retail markup.

What would happen if you instead could call third-party providers (for example, 900-DIRECTORY) directly? Each provider would be able to set a price, and because you could call many providers, they would be forced to compete for business. For a commodity service such as directory assistance, price would be a key factor. A company that offered similar service but at a price of 50 cents per call would quickly undermine the carrier's overpriced directory assistance service. Just as with any other economy, competition would reward innovation and bring prices down.

Demand for telecommunications services is highly elastic. If the cost of directory assistance were reduced to a quarter per call, people would use it much more heavily than they do today. Once the price for a service declines below a certain threshold, consumers begin to perceive it as free and will use it as much as they want. For example, long-distance calls are now so cheap that people rarely pay attention to the amount of time they spend on the phone.

It might seem counterintuitive that pay-per-call services will make some information services cheaper, but that is exactly what will happen if it is done correctly. They will also stimulate the development and growth of a wide range of information services.

What Went Wrong?

Why did the experiment with pay-per-call services go so horribly wrong? In a word, pricing. 900 numbers were introduced more than a decade ago, when long-distance prices were five to ten times higher than they are today (20 cents per minute was "cheap" then). To make money, information providers had to charge steep fees, usually $1 per minute or more. The carrier would usually charge 50 cents per minute just to take the call, so to make any money, you had to charge much more. There aren't many things worth paying $1 per minute for. That limited the possible uses for 900 numbers from day one. Legend has it that one of the executives at AT&T who was involved in the creation of the 900 service realized what people would be willing to pay for, and went off to start one of the first dial-a-porn ventures, and became sickeningly rich in the process.

High pricing wasn't the only problem. 900 numbers quickly became associated with porn, the one killer app for pay-per-call telephone service. Prudish legislators and regulators imposed myriad rules and regulations, which had the effect of rewarding perversity, as well as spoiling the system for nonpornographic service providers. Among other things, they imposed restrictions on what operators could discuss (the so-called Helms Amendment) and, most importantly, gave customers a free pass to refuse to pay their phone bill. So, for example, if you spent several hundred dollars on calls to "chicks with blankety-blanks," all you had to do was tell Pacific Bell, "Hey, it was my teenage son, not me!" and you were off the hook (no pun intended). Charge-backs became a huge problem, so companies had to increase their prices even more to offset uncollectible bills.

This turned into a vicious circle, and within a few years the only companies offering 900 services were dial-a-porn services (which could charge several dollars per minute and still do brisk business) and software technical support. The situation got so bad that AT&T exited the 900 number business several years ago.

What Might Have Been?

Now, let's imagine an alternate history where 900 numbers operated exactly the same way but with a different business model. In this alternate world, the carriers simply take 20 percent of the toll charged to the caller above the standard carriage fee. In 2005 dollars, an example of this formula would work something like this:

Carriage fee: 2 cents per minute (the standard bulk rate for toll-free service)
Retail rate to caller: 20 cents per call
Average call length: 2 minutes

Gross revenue per call: 20 cents
Cost to carry call to provider: 4 cents
Revenue per call less transport: 16 cents
Commission to carrier: 3.2 cents per call
Net revenue to service provider: 12.8 cents per call

It's a small difference, but it completely changes the economics of the business, because the cost structure is much simpler, with low variable costs and no fixed costs. Basically, this is just a simple commission model.

A simple change, but it solves a big problem for information service providers. Most information services are features, not stand-alone products. Once again, directory assistance offers a great example.

Metro One, a Portland, Oregon-based provider of enhanced directory assistance (DA) services, is one of the main shops to which carriers outsource DA calls. The past several years have not been kind to Metro One, as carriers have sought to squeeze vendors for every penny (while raising rates for their subscribers). About two years ago, Metro One recognized this trend and created its own directory assistance service called Infone ( It costs 89 cents per call, about 30 percent to 40 percent less than what the carriers charge. It also offers more features.

There's just one small problem. The only way Metro One can collect fees is by credit card. Most consumers are wary of subscribing to services like this because they assume that they are signing up for yet another $10-per-month drain on their credit card. Metro One doesn't charge a monthly fee, but because so many companies have a habit of autocharging customers a few dollars here and a few dollars there, customers are rightly suspicious. The bottom line is that building a subscriber base for a service like this is difficult. It is also expensive to bill customers for small amounts. Credit card companies extract transaction fees that cut into margins. Billing systems, accounting, and customer service personnel also cost a lot of money.

If Metro One had instead been able to offer service to users via, say, 900-1INFONE at a cost of 89 cents, it would have accomplished the same end with a lot less hassle. Customers would not have to subscribe to the service in advance. They could just call it when they needed it. The cellular provider would bill the customer, and would pay Metro One 80 percent of the charge on the back end (or whatever commission it had negotiated). Simple, cheaper to operate, and it makes impulse usage much easier. Most importantly, it would have encouraged other companies to compete with Metro One, which would have rewarded innovation (for example, using speech recognition to drive costs down) and would have brought prices down further. Take operators out of the loop, and there is no reason directory assistance should cost more than a quarter per call.

There are a lot of services that people will pay a token fee for. New handheld devices that blend conventional telephony with IP-based services expand the realm of uses dramatically. Let's take streaming media as an example.

Suppose you want to listen to a game on your favorite radio station while you are out of town. You call a pay-per-call number for that station at a rate of 4 cents per minute (plus normal airtime, though you probably have a flat-rate plan that renders the airtime effectively free). That's cheap enough that you don't really think about the price. A phone system answers the number with a live audio stream, or if the phone is voice-over-IP-based, streams hit bit-rate audio or redirects it to an MP3 stream. You listen to the game for a half hour, at a total cost of $1.20, less than a cup of coffee. Prices for call transport services are going down, so in two to three years' time, a service like this could be priced even lower.

Using the formula above, the phone company would charge the station a couple of cents per minute for call transport, with the station keeping the remaining 2 cents per minute. Doesn't sound like much, but the station would be answering many calls at once. If an average of 20 people were listening to programs this way during a 10-hour business day, that works out to 360,000 minutes per month, or $7,200 per month. Chump change adds up when a lot of people are on the line.

Free-Market Competition

While you can debate the merits of using a telephone or VoIP device to listen to a radio station, the point is that people are willing to pay for telecommunications services, and that they are willing to pay different amounts for different services. One service might be worth a dollar per call, another only a penny per minute. There is no such thing as free telecommunications. Even free internet phone service requires a nonfree internet connection.

There would be competition not only among information providers, but also among long-distance carriers. Competition for toll-free telephone service is fierce, with market prices now below 3 cents per minute on many networks. These same companies would be natural players in a revitalized pay-per-call industry. They already have all of the call transport and billing infrastructure needed to provide this service. (In fact, many long-distance companies once provided 900 service.) Competition among carriers would drive costs for 900 service down to levels similar to those of toll-free service, which would make pay-per-call service attractive for a wide range of applications.

Wireless carriers also would benefit. Basic voice service has become a commodity, with most customers paying bulk rates for bundled or flat-rate plans. Premium-rate services are a great way to increase revenue, because wireless operators would receive a share of the collected revenue. (When you call a toll-free number, both the originating and terminating telephone companies share the collected tolls; similar rules would apply to 900 numbers.)

The public telephone network is more than a global intercom system. It is also an elaborate currency exchange platform. Most people don't think of it this way, but every phone call is logged, categorized, and priced (even if it's a free call, telcos itemize everything). Pay-per-call services allow service providers to assign an arbitrary price to a call, and by doing so, turn the telephone network into a marketplace for information services.

Revitalizing pay-per-call service will have a dramatic effect on the development and growth of wireless information services, most of which will be accessed by telephone or phonelike devices. Wireless information providers all suffer from the difficulty of acquiring subscribers and collecting revenue. If every new information service were just a nickel phone call away, those companies would have a much easier time attracting customers and investors.

People scoff at the idea of micropayments, but it is important to note that most of the telecommunications industry is based on them. Most telephone calls cost only a few cents. Yet, all of those 5-cent calls add up to real money--hundreds of billions of dollars, in fact.

Avoiding Past Mistakes

Reintroducing premium-rate numbers does not mean a return to porn and psychic lines (which ironically were unable to foresee their own insolvency). A few simple rules can be employed to protect the system and consumers:

  • Identity verification: require callers to verify their identity using a PIN code upon the first call to a provider.
  • Low pricing: price transport services to reward low-cost, mass-market applications (such as audio content on demand).
  • Micropayments, not macropayments: carriers set price caps, so the system would be used only for low-cost transactions (for example, a 75-cent call for driving directions, a 10-cent call to listen to a recording of an archived radio program, and not a $200 call to "Candy Cane").
  • Charge-backs: the caller is responsible for all charges made from his or her phone unless it is reported lost or stolen. No more free pass to say, "My buddy must have made those calls, not me." Carriers' response should be, "Well, don't lend your phone to your buddy anymore."

By focusing on low-cost, mass-market applications such as wireless content delivery, carriers can avoid many of the problems from the first generation of pay-per-call services.

Premium-Rate SMS

As it turns out, the wireless industry already has part of this problem figured out. Premium-rate SMS service has emerged as a popular payment system in Europe and is now making headway in North America. PSMS is based on a concept similar to that of pay-per-call service, except that SMS is the communications medium.

For example, let's say you want to get your local weather conditions and forecast from The Weather Channel. Just send an SMS to FCAST (42278) with the city name or zip code as the text message. The service will reply with a 36-hour forecast. The service costs 75 cents per message. PSMS services range from about 25 cents to several dollars per message. As it turns out, it's a great micropayment medium.

The cellular industry already has one toe in the premium-rate services pool. With all of the billing infrastructure in place for PSMS, it is a straightforward next step to offer similarly priced pay-per-call services.

Brian McConnell is an inventor, author, and serial telecom entrepreneur. He has founded three telecom startups since moving to California. The most recent, Open Communication Systems, designs cutting-edge telecom applications based on open standards telephony technology.

Beyond Contact

Related Reading

Beyond Contact
A Guide to SETI and Communicating with Alien Civilizations
By Brian McConnell

Return to the O'Reilly Network.