Building a better mousetrap

Just about every kid at some point plays a board game called Mousetrap, where the goal is to cobble together a whole bunch of seemingly unrelated elements into a system that in the end traps the mouse, winning the game.

IP billing as it exists today looks a lot like that game, as service providers and solutions developers struggle to determine how to best string together disparate, largely incompatible billing, rating and mediation platforms to produce the coherent, detailed invoices that customers will understand and pay, thus generating revenue. Which, of course, is how you win the game if you're a service provider.

They're also trying to figure out how they got to this point on the game board, and what has to be done and by whom to make the process work better.

Recognizing opportunity

Billing at its most basic seems like a pretty simple deal: Keep track of the services your customers use, list them accurately, send the invoices out regularly and then wait for the cash to start rolling in.

But even this simple endeavor has managed to elude the grasp of many service providers, which don't have the ability to rate and mediate adequately to distinguish amongst the various services they provide. The result often is that they're forced to bill on a flat-rate basis, which robs service providers of the opportunity to get fair market value for certain services that eat up more capacity or take higher priority in the transport protocol (Figure 1). And sometimes they're not able to bill at all.

As if this isn't enough of a bugaboo, billing challenges encompass much more than this simple picture suggests, at least in the IP arena.

For starters, billing — when done well — can be used as a tool to reward customers for their patronage and to give them reasons to purchase more — and more expensive — services, says Bob Bell, product strategist for billing solutions provider Geneva Technologies.

“A carrier has to incentivize its customers to use other services that are higher in value to increase [average revenue per user] — that's what the carrier's game is,” he explains. “Billing is a way to do that, by providing product discounts based on usage. In other words, a customer can earn so many MP3 downloads. That really needs to be embedded in your billing and rating capability.”

Billing also can be used as a strategic weapon, allowing service providers to differentiate themselves via their pricing strategies, says Phil Mutooni, senior manager of business support engineering for voice-over-IP (VoIP) provider iBasis.

“That's one of the most overlooked aspects of billing because everyone seems to think this is one of those ‘after-the-fact’ issues,” he explains. “Most people look at telecommunications as being about services, but in fact, it's a billing system wrapped around services. If you can't bill, you might as well forget about it.”

This may have been less of an issue in the circuit-switched arena, where vendors produced only a handful of new services each year, says Michael Couture, director of strategic market management for billing solutions provider Amdocs. Consequently, it was a relatively easy matter for service providers to keep their billing systems in step. In the faster-paced IP world, there's been an explosion of new services, which presents great opportunity for service providers.

“It creates a lot of potentially billable services that will help to protect the voice services and the customer base, and reduce churn,” he says.

Couture is quick to acknowledge that this vast opportunity comes with a tradeoff.

“Where you once had a handful of switch vendors and new services each year, [service providers] are now looking at hundreds — if not thousands — of new services that they want to be able to charge for and not just include in a flat rate. That means a lot of impact to the billing system,” he says.

The challenge

While the sheer volume of new services seems to be vexing enough for service providers and solutions providers, the technical nature of IP also presents a notable hurdle in creating a workable billing system, iBasis' Mutooni says.

“The problem is that IP is a service-agnostic protocol, so supporting a common format for Web-log servers and for call-detail records for voice over IP, the public network, media conferencing — the list is endless — presents a fairly big technical challenge,” he says.

Another problem is the attitude service providers take toward billing, Couture says. “Certainly one of the challenges of building an application is that accounting and billing is usually an afterthought,” he says. “That puts the onus on the billing and mediation vendors to make something of what little information we're given for the events that occur.”

Joe Lynam, CEO of solutions provider eBillit, agrees, adding that since the advent of the Internet, service providers have been focused almost exclusively on the development of applications — what he calls the “sizzle” — on the front end, which leads to trouble.

“Billing has always been, ‘We'll get to it,’” he says. “There's a lot of complexity associated with billing, from the platform to the right way to bill a customer. Deciding on whether you're going to use a flat-rate or metered model, and then deciding on the appropriate delivery and presentment mechanism, is critical because you're going to have gaps if you don't do it properly.”

Andrew Bonner, director of market development for ADC's strategic solutions group, further suggests that service providers need to consider the back office a more integral part of the service offering. He says that back-office planning must be in “lock-step” with product development and marketing activities.

“They can't leave the billing system until the network is in place and everything is turned on,” he explains. “It has to be planned at the same time.”

Even when billing is treated as more than an afterthought, how service providers cobble together their systems often gets in the way of successful billing. All too often they use the same strategy that a child would use when playing the Mousetrap game — build it as you go — resulting in unwieldy systems with too many incompatible parts, says Mark Ozur, chief technical officer for VoIP provider edge2net.

“The new IP providers tend to buy a lot of off-the-shelf solutions and hope the vendors have sorted all of this stuff through,” he says.

The quest for interoperability

The real problem, counters Dominick Tolli, senior vice president of operations for VoIP provider Net2Phone, is that too many solutions providers take a niche approach to developing billing platforms, which too often forces service providers to take matters into their own hands.

“We have vendors say to us, ‘This thing is hard; just let us do the rating piece.’ And we say to ourselves, ‘Well, thanks. That's not the hard piece.’ So we ended up building, believe it or not, about 95% of our platform internally,” he says. “It's not the best, and it's certainly not something we'd want to sell, but it works.”

Peter Janca, global solution executive of customer care and billing for IBM, agrees. “Everybody has a different idea on how to do this,” he says, “so there's no consistency. There are some solutions providers that take a holistic view of the billing problem, and there are others who take a focused niche view. Part of the basic decision an IP supplier has to make is, ‘What kind of animal am I?’”

The decision is often made for the solutions providers by the unique nuances of the IP space, Mutooni says.

“Solutions providers are trying to do software development at the speed of thought,” he says. “As soon as someone in marketing comes up with an idea about great packaging or some innovative way to do pricing, that somehow has to be implemented. And obviously there is a software development cycle, so that presents a challenge. [That's why] a lot of folks are saying they're just going to focus on a certain niche.”

Amdocs' Couture believes this is especially true in start-up companies, where a lack of research and development resources and inadequate staff tend to limit focus. “Best of breed then becomes the default for vendors that don't have the scope of the solution,” he says.

A best-of-breed approach may be better for certain service providers than an all-encompassing “telco-in-a-box” solution, says ADC's Bonner.

“For a Tier 1 company that can afford the specialization of components, the ‘telco-in-a-box’ approach makes no sense whatsoever,” he says. “It usually does not represent a best-of-breed solution end-to-end, which certainly leads to future trouble if you find some function outside your preferred vendor that you wish to deal with.”

Another reason to work with multiple vendors, adds Mutooni, is that there is safety in numbers.

“Since we're in an unpredictable environment these days where you never know when someone is going to go out of business or be acquired, working with multiple sources offers a more reliable solution than working with one source,” he says.

Perhaps, but edge2net's Ozur, while a proponent of the best-of-breed approach to building a billing platform, acknowledges there's no way of knowing whether various components will be interoperable.

“The vendors haven't sat down in any integrated way and tested all of those solutions. Hence, there's lots of places for breakage,” he says.

Net2Phone's Tolli is even more adamant in his disdain for the best-of-breed approach, something he says is quite difficult to pull off because of variations in not only hardware interfaces but also the software that runs the systems.

“So you might get three boxes together, one each from Sun, IBM and HP, and you think, ‘Boy, this is going to be some hot system.’ Then when you go to put them all together, you find out that it's going to be a nightmare. Only if it's standards-based can you swap in and out and buy the best of the best.

“If it's not standards-based and just cobbled together, it just introduces too many variables that are going to give you trouble down the road. Plus, it doesn't give you a single point of contact to get help when you need help.”

Fixing the breaks

While the debate will continue for the foreseeable future as to whether an a la carte “best-of-breed” solution is ultimately better than an all-encompassing “telco-in-a-box” solution, so too goes the debate on the role that standards will play in the future of IP billing.

The consensus seems to be that standards initiatives such as the one being undertaken by the Internet Protocol Detail Record Organization (IPDR.org: see sidebar on page 38) make sense in theory, but more doubt exists concerning their practical implementation.

“IPDR is a great example of an initiative that has come a long way in a very short period of time, and that's very encouraging,” says Amdocs' Couture. “Those types of initiatives benefit everyone in the industry because they create the potential for building an application that's readily billable.”

Any move toward standardization helps, he says. “It would certainly facilitate and increase the speed in which new services come out and operators are able to bill for them — no doubt about it.”

IBM's Janca isn't as sure. “The market is moving so rapidly that by the time you get a standards-based approach, it would be obsolete,” he says. As evidence as to how quickly things can change, Janca points to the competitive local exchange carrier (CLEC) market. “If you had asked me a year ago who my best customers were, I would have said, ‘CLECs. There are thousands of them, they're going like crazy and they have infinite amounts of money.’ I wouldn't say that today.”

Although he believes an IPDR standard will one day become reality, Kamran Sistanizadeh, co-founder and chief technical officer for Yipes, cautions that a standards-based solution will have its own set of limitations.

“Whether these IPDRs can be exchanged in a transparent fashion between different billing systems [between carriers], I think that remains to be seen,” he says.

While he acknowledges a need for a common data-CDR format, iBasis' Mutooni agrees with Sistanizadeh that service providers and solutions providers should carefully manage their expectations concerning emerging standards. He adds that initiatives such as IPDR may be more effective in promoting much-needed dialogue than creating a panacea-like answer to the IP billing dilemma.

“The nirvana for software development is when someone thinks about something and it's totally implementable overnight using some drag-and-drop interface or plug-and-play module, but that's not how life works,” he says. “The answer is converging on a model that has an open development environment and architecture that supports easy integration.”

edge2net's Ozur further suggests that the answer should have its genesis with equipment vendors such as Cisco Systems and Lucent Technologies, which could support the cause by providing the proper elements in their solutions to aid the billing process.

“I'm not saying they should become billing companies, but I am suggesting that they have never seriously considered billing as a fundamental notion of their data processing,” he explains. “The hooks into the equipment needed to generate the right information to bill properly just aren't there.”

In the end, market conditions may force service and solutions providers to speed up the process of solving the IP-billing bugaboo.

According to Peter Brown, director of IP OSS systems for Global Crossing, the billing system selection process is now being affected by venture capitalists who want to make sure the service providers they are funding are capturing revenue and controlling costs.

“Before, I think people would buy billing systems without doing a great deal of due diligence,” he says. “Now, with revenue assurance getting a boost as part of the market downturn, service providers have their feet to the coals more than you could ever imagine. If you're in a buying decision, you're under a lot more pressure than ever before.”

This, in turn, is putting a great deal of pressure on the solutions developers, which need to act before service providers get to the point where they feel cornered, says Net2Phone's Tolli.

“The question becomes, ‘Can solutions providers come up with adequate platforms before service providers build their own systems because they no longer were able to wait?’” he says. “If that happens, they might have difficulty making the sale because people can't stand still. We have to find a way to do it, and we're going to trudge through it.”

Spec a little, build a little

by Glenn Bischoff

At first it seems like the ultimate conundrum: how to build an IP-billing standard one step at a time but do it at Internet speed.

But then you speak to Steve Totten, chief editor of technical specifications for industry coalition IPDR.org, and it all seems to make sense.

Two years ago at Billing World in Dallas, AT&T, Narus and TeleStrategies convened to discuss the potential for an IP detail record organization. “The meeting was held at 7:30 a.m. on the day after all the hospitality suites, and 250 people showed up, which was about half the show's attendance,” Totten said.

The Internet Protocol Detail Record Organization was thus born. Among those joining AT&T, Narus and TeleStrategies as charter members were Lucent Technologies, Hewlett-Packard, Daleen Technologies, Convergys, Clarent, Telus, Agilent Technologies, Accenture and Xacct Technologies. Today the roster boasts more than 100 service providers and solutions developers.

Two months after the get-together in Dallas, the fledgling group held its first formal meeting in Palo Alto, Calif., and by December 1999 had version 1.0 of a technical spec. Just 14 months later, IPDR.org unveiled version 2.5 of its proposed network data management-usage (NMD-U) specification.

This version provided benchmarking for the language — XML (extensible markup language) — and transport protocol — SOAP (simple object access protocol) — both of which are at the heart of the specification.

Selecting XML for the language was an important decision, Totten says.

“Very quickly we got into the idea of practical implementation,” he says. “So we've been trying to leverage a lot of existing work and didn't go into this with the idea of inventing all new notations languages and coding schemes. We tried to base it on things in the Internet world that were already pretty well accepted.”

Though they may be well accepted in the Internet world, no one knew how well XML and SOAP would work in an IP-billing application. There was some concern that XML might not be scalable because it is a relatively verbose language and that SOAP might not be reliable enough to handle anticipated volumes because it is a new protocol.

Consequently, the benchmarking activity was an important stage in the development of the specification and the organization, Totten says.

“The philosophy I put into place, which everybody agreed with, was, ‘Spec a little, build a little,’ rather than spec the ultimate conception and then publish something that would turn out to be irrelevant three years later,” he says.

The concerns were confirmed by the benchmarking. For version 2.6, which was scheduled for release this month, the HML language will be altered to contain a condensed encoding of the IPDRs, while the firewall transfer protocol will be beefed up to make SOAP a fully reliable transfer mechanism, Totten says.

Looking to the future, Totten has two end goals: practical working implementations by service providers of business-relevant forms of IPDR and greater involvement of service and solutions providers in the effort.

“Even though they're dedicated to making something happen, they have day jobs. That's one of our challenges,” he says.