Features

magine the day when competitive and incumbent carriers can exchange ordering and provisioning information seamlessly. No re-keying data. No orders rejected because of incorrect or missing information. No time spent tailoring each side's electronic systems to enable the two to communicate.

As service providers attempt to end the interconnection nightmare through solutions ranging from e-bonding to service bureaus, one has to wonder: Can't we just all get along?

The concept is one of middleware. Each carrier would write to an open application programming interface (API), ensuring that its ordering and provisioning systems map the right information to the middle-ware, which would be the gateway to another carrier's system.

Sound too good to be true? Well, it is. But though we're not there yet, carriers are moving in that direction - pitfalls and pluses alike.

One of the key issues between competitive local exchange carriers (CLECs) and incumbent LECs (ILECs) is sharing information electronically. Known as e-bonding, the process is fraught with problems. For starters, every provider uses its own internal ordering and provisioning system. Each has a unique system and process, some more streamlined than others.

"Today there are as many different inter-faces required as there are trading partners," says Peter Campbell, chief information officer of Nextlink Communications.

If a CLEC places orders with four incumbents, it must tailor its system to map the correct data to each of those systems - all of which are different. Though incumbents publish their interface specifications, integration is still a cumbersome task for competitive carriers, which typically bear the burden of interoperability. Perhaps it is an unfair burden for CLECs, but there are far fewer incumbent legacy systems - especially with the continuing merger trend - than competitive carriers, which number in the thousands.

An RBOC might do business in 15 states, and it must interface with hundreds of CLECs for e-bonding, says John Barnicle, chief operating officer of Focal Communications. Though he is rarely sympathetic to RBOCs when it comes to automating work flow systems, he under-stands their concerns. Considering the number of application processes and databases that must be updated, "it must be daunting," he says.

E-bonding issues Mapping information between carriers' systems is no easy task. There are the obvious logistical issues when directing data to the correct fields in another carrier's system. Those are compounded by different naming conventions and service definitions. A simple but critical field such as the customer address can be intensely problematic.

For example, assume a service must be provisioned to a customer residing at 123 West Main Street.Where one system recognizes that address, another might require 123 West Main St. or 123 W. Main Street or 123 W. Main St. Add an apartment number to the mix, and it gets even worse: Apt. 4, #4, Unit 4, etc. Date and time stamping of messages also can pose problems (Figure 1).

Similar confusion can be found in how carriers define services. The required data elements vary between states, so carriers offering service across borders must keep up with them.

In addition, the number of related systems must be updated; 911, 411, line information database, calling name, PIC/Care and local number portability are only a few of the databases that must be part of the e-bonding procedure. Doing any one of those things in a coordinated fashion might be feasible, Barnicle says,"but to do eight or 10 of these at once can be a gamble. The key is to automate as much as possible into work flow engines."

Where once this information was used to provision voice service only, now it encompasses high-speed data, video, DSL and other technologies. Within those, carriers use different types of groupings to define a service.

Copper wires are one example, says Nancy Kaplan, vice president of Renaissance Strategy."Now look at the number of characteristics they have to keep track of on copper. There are applications for DSL service that weren't part of the old Bell system." Carriers must track the kind of copper, the number of wires and other things on the line such as bridged taps.

This is one of the downfalls of the Telecom Act of 1996. Competition breeds innovation, but in the sharing of data, it has caused problems.

"There is not a common understanding of what a product or a data service is," says Dave Deutschman, chief technology officer of Quintessent Communications, which develops back-office software."Many years ago, Bellcore created a common language that was used for products and features and site locations," he said. "Essentially, it was a universal service ordering code. Today, everyone does product codes and feature codes their own way."

These issues largely have been resolved in the access service request (ASR) side of the house. The Ordering and Billing Forum has defined standard forms and data elements. "But the local service side is up in the air," Deutschman says. "They may vary 20% to 30%, and that makes integration very difficult."

Compounding the problem is that not all RBOCs adhere to the standards. "The OBF standard is a nice concept, but it's not followed," says Jerry Sklar, senior vice president of OSS for Allegiance Telecom."Not one of the RBOCs is 100% [compatible with the OBF standards]."

That has forced CLECs to build to RBOCs' published specifications, though few have fully automated systems. Incentive to electronically bond is strong, however.

"It will cut down on errors, lead to some benefit in the [provisioning] time line and give us a tracking mechanism to hammer the RBOCs, then they send [an order] back," Barnicle says. "But it's important to remember that e-bonding is not the panacea."

Barnicle's chief complaint is that when Focal submits automated orders for service, the RBOC system will reject them outright if any errors exist, without specifying all the fields that must be corrected. If an order has three mistakes, it is returned three separate times, he says. "[RBOCs] have the capability to tell us `these three fields need to be filled out this way.' They claim to be wanting to be helpful, but it's disingenuous when we see things along these lines," he says.

The burden of cost Today, e-bonding is often ad hoc. In most instances, smaller carriers request service from larger incumbents, and they tailor their systems to present the correct data to the incumbent's system.

Like many smaller carriers, Electric Light-wave Inc. relies on faxing to place orders. It also can directly access, but not electronically bond to, the systems of Pacific Bell, U S West, GTE - now merged with Bell Atlantic, forming Verizon - and Southwestern Bell, says Becki Gilbertson, director of IT for ELI. "But it's not all automated. We have to enter [the order] in our system, then enter it in their system. When they provide the service, we have to enter that into our system again. It's still kludgey."

And sometimes it doesn't make sense to devote resources to e-bonding."Carriers uniformly complain about the cost," says Dale Quick, Quintessent chief operating officer. "It is a very expensive proposition."

It also can be acrimonious. CLECs must send a substantial number of orders to an incumbent to justify developing business rules and mapping interfaces."We do it with two or three of our highest-volume business partners," Gilbertson says.

One reason is because the usual incumbent attitude is "Fix your system to match ours," she says. "It's still a team effort. Every transaction has something coming back. It's not a one-way business transaction. It has to be a two-party game."

Still, the onus is on the CLEC to map to the incumbent. It is a time-consuming and cash-draining process - especially for small providers with limited resources. "From a productivity standpoint, I don't want my customer service reps to worry about if Bell Atlantic or BellSouth use the same interface for the same transaction," Nextlink's Campbell says.

With a fully automated system, competitive carriers can dedicate their resources to capturing and keeping customers rather than tracking orders. "We have 10% of our work force in an order entry and provisioning group. That could be pared back," Focal's Barnicle says. "We have another 15 to 20 people in a cost assurance or access management group. We could cut that or not add as many people to those as we scale up." Those groups represent 10% to 15% of Focal's total work force, he added.

That doesn't include money lost from slow provisioning. That is not a big concern for Focal, which primarily orders T-1s with a two-or three-week turnaround. But unbundled loop orders typically are processed in three to five days, Barnicle says. Shaving a day off a three-to five-day interval is more significant.

Investing in e-bonding can be a necessary distraction for carriers, but it also can lead to stopgap measures. CLECs "are gathering the customers they can now, and they are investing in processes that are less than totally efficient to do that, knowing that down the road they will have to make an investment in a long-term solution," Quick says.

Companies such as DSET and Quintessent help streamline the carrier-to-carrier communications (Figure 2)."We are mapping the correct data elements properly with the right information," Quintessent's Deutschman says. The software must apply the business rules, then map data to the right location and in the correct format for the recipient. The system must translate local service requests and ASRs going to the incumbent and the firm order commitments ELI receives from them.

In a fully automated scenario, Gilbertson would like the ability to track when ELI requests a circuit for use, when it's activated and when ELI starts to pay for it. By reading into each other's systems, carriers can validate orders automatically. If a disparity exists, Gilbertson also would like the system to resolve disputes."All disputes are manual," she says.

Help from the outside Some solutions exist to help automate the process. The two most common options are gateways and service bureaus, says Donna Nanney, product manager for regulated ordering at MetaSolv Software.

A gateway is software that resides at the customer premises and requires the carrier to tailor its system to do business with each of its trading partners. BellSouth has created the Telecommunications Access Gateway and a complementary front-end for CLECs called RoboTAG (see side-bar on page 52).

MetaSolv and its partners also fit in this category, though the system can feed into a service bureau as well. MetaSolv can get the gateway up and running in about 90 days, but it still requires third-party integration from consultants and integrators such as Ernst & Young or Arthur Andersen, Nanney says.

"It's not true plug-and-play," she says."But it's as close as you can get."

The integrator's work is ongoing. CLECs track changes to state tariffs, which can affect the automation, and the integrator must update the work flow system. Allegiance, which has developed its own system (see sidebar on page 56), suggested that DSET become an agent for the provider, tracking the moves and changes required for the electronic gateway.

"I have to engage them anyway; they have to change the mapping," Allegiance's Sklar says. "If they jump in with both feet, I'd be happy to pay."

In the service bureau or clearinghouse approach, the carrier has a single connection to the clearinghouse, which determines what the carrier needs to do business with its trading partners.

Seizing an opportunity, companies such as Telcordia Technologies and Extant have created service bureaus. These companies act as middlemen between carriers' ordering and provisioning systems. They regularly review the RBOCs' published APIs and stay abreast of the state requirements for service orders. They work with CLECs to ensure that the data flows to and from the carriers' systems in the proper format. That includes protocol translation and the mapping of data.

"The emergence of some third-party providers and service bureaus is helpful to everyone until more standard-based types of things can emerge," says Focal's Barnicle.

Incumbent carriers particularly seem interested in the service bureau approach. One reason is, "If you outsource it, the incumbent doesn't have to get involved at all. It's a laissez-faire attitude," says Paul Hughes, director of billing and payment application strategy research for The Yankee Group.

From the ILEC side, working with a few service bureaus beats trying to interwork with hundreds of CLECs. Lynn Notarianni, director of wholesale systems for Qwest, is a fan of service bureaus. "The service bureau is probably the most sophisticated interface concept today," she says. "The service bureau gets you 95% of the way there."

From Qwest's perspective, functionality and support are the same regardless of connection mechanism, Notarianni says. But automated systems are more efficient because users are less prone to make errors. "It automatically generates a request... and the machine is programmed to know precisely what it has to send every time to be successful," she says.

Service bureaus were born because competitive carriers need them."No vendors have come to us without a CLEC on the other side," Notarianni says."We believe it is the most efficient way for ILECs and CLECs to work together. You don't keep repeating development with every CLEC. They build to our interface and establish a way to keep up with releases."

In days long past, ordering and provisioning systems were easy. Even after the breakup of AT&T, the RBOCs continued to use the same systems. Now carriers complain the landscape is hard to maneuver, Kaplan says. "Most say,`I wish Ma Bell were back.'"

Surely competitive carriers would dis-agree. Many have developed their own systems, which they consider improvements over the RBOCs', Kaplan adds.

CLECs can push incumbents to revamp their systems, but they won't bite unless the suggested changes improve the existing process. "With ILECs, it's not active resistance, it's passive resistance," Kaplan says. "When you come up with a better system, they'll go along with it. But they will not take dramatic or active steps to improve the system."

And ILECs have reason to maintain their stance, Kaplan argues. Some CLECs have growing pains and are working with old or nonexistent OSSs. That's why the service bureau, despite its cost, is a good interim solution, she says."[CLECs'] houses are not in good shape. In the short term, they can go outside, but in the long term, they are going to have to get their back offices straightened out."

Though service bureaus worked well in the cable and wireless industries, they may not translate to the wireline market, Hughes says. A concern with service bureaus is that CLECs will lose their ability to differentiate based on services.

Notarianni agrees. "To have an environment that is 100% ubiquitous goes against the grain of competition."

The pace of change What will drive RBOCs to automate systems more quickly and cooperatively is the carrot of offering long-distance service. To do that, they must meet the 14-point checklist included in Section 271 of the Telecom Act. "E-bonding impacts several things on the checklist. Then they are incented to work on it," Focal's Barnicle says.

Nextlink's Campbell agrees. "The great motivator for the LECs to fix [e-bonding] is greed," he says. "When they are seriously trying to use my service and facilities, the shoe goes on the other foot. When SBC or Bell Atlantic goes outside their footprint, then they'll understand what the issues are."

RBOCs - generally slow movers - have a valid reason for dragging their feet in this scenario: Their legacy systems are main-frames and never were meant to be open or to interface to outside systems. "It's a significant undertaking," Barnicle says.

Incumbents also resist because they fear they will lose business. "They don't feel compelled to help us through these things any more than the law makes them," Barnicle adds. "They are spinning their wheels, and they seem content to do so."

With two motivated partners, the automation process can be fairly smooth.Within three months of Nextlink's merger with Concentric Network, the two companies had built an internal communications system to exchange order information, Campbell says. "Why does it take me years to do that with the LECs?"

The prospect of building an open API might be too simple to work. Incumbents face technical and logistical issues of mechanization; CLECs are too independent and disorganized in the operation support system realm to have a collective voice. In the short term, the e-bonding process likely will continue on an ad hoc basis.

But not all are disappointed with the progress. "I don't think there is ever going to be a magic place in the middle that knows when to turn around and send an order to PacBell, U S West or GTE," says ELI's Gilbert-son."Some people out there think it's going far too slowly, but I don't have that position."

BellSouth developed the Telecommunications Access Gateway, the company's published application programming interface, to let carrier customers obtain pre-order information and submit firm order commitments.

But TAG requires carrier customers to develop their own front-end systems, and not all CLECs are equipped to do that, says Marcia Moss, operations support systems manager for interconnection services.

As a result, BellSouth developed Robo-TAG."It does all the front-end work for the customer," Moss says."It includes a graphical user interface screen. Business and usage rules and data requests are built into the software, so the customer doesn't have to do all the design work." It shortens implementation time from four to six months to 45 to 60 days and cuts costs from at least $250,000 to $100,000 or less, she says.

That's a partial solution, but RoboTAG doesn't feed back into the CLECs' systems."They still need integration on their end," Moss says.

BellSouth couldn't develop a system that interfaced with all CLECs' systems - there are too many, she adds."I've meet with 100 CLECs in the last two years, and I have yet to see one that is identical or where there is no translation necessary to get from one system to another."

Because carriers input data directly to BellSouth, the incumbent can process orders more quickly, Moss says. But Robo-TAG doesn't obviate manual processing of orders. Some complex services do not yet have standards, nor have business rules been developed to process them electron-ically."Not all telecommunications services can be electronically ordered today," she says.

In addition, CLECs bear some responsibility for keeping up to date. Their electronic systems can be "months behind" the incumbents, making e-bonding nearly impossible, Moss says.

As a budding competitive local exchange carrier, Allegiance Telecom decided to play like a player. The company's strategic vision included a broad footprint and e-bonding. In fact, the latter was required so Allegiance could scale to a number of markets, says Jerry Sklar, senior vice president of operations support systems at the company.

In 1998, Allegiance partnered with MetaSolv Software for its ordering and provisioning system and with gateway vendor DSET. But that's only part of the solution. The carrier also works directly with incumbents. To ensure solid communications, Allegiance forms a team of product managers from the CLEC, the incumbent and the software vendors. Up to six people are assigned to work with each of the incumbents. They foster relationships and cooperation between Allegiance and its partners. Operational groups meet weekly, monthly or quarterly to resolve issues.

"The product managers facilitate the relationship between all those parties," Sklar says."It's a full-time job." They are charged with understanding ILEC issues and explaining Allegiance's issues to ease the e-bonding process.

Allegiance's most advanced system in place is the one that links to Bell Atlantic North, Sklar says. Work started in June 1998, and by the following January, Allegiance rolled out its first electronic local service request gateway to Bell Atlantic.

"It was the first in the industry, and it set a template for what to do with the other ILECs," Sklar says. Working relationships with other ILECs followed and similar teams have been established.

"It took us six months to work out a system with Bell Atlantic, but it took only 90 days with Southwestern Bell."Allegiance just completed the product with BellSouth and has automated systems in place with Pacific Bell and Ameritech. Next up are Bell Atlantic South and U S West, expected in the fourth quarter."We'll be the first CLEC to complete e-bonding with all the RBOCs," Sklar says.

Allegiance has won recognition for its efforts. "We won the Supercomm [SuperQuest] award for the best e-bonding gateway out of 160 entrants," Sklar says."The runner-up was BT."

The company has no secrets, he adds. The key to its success is its relationships with trading partners and its dedication to making those work.

- With e-bonding, firm order confirmations are sent within four-to-six hours, if not immediately; manual confirmations can take 24 hours to process

- The e-bonding system with Bell Atlantic saves an average of 20 business days to process customer orders through the system

- The volume of customers that can be processed has increased quite a bit with e-bonding superscript *

superscript * Undisclosed percentage Source: Allegiance