May. 28, 1999 -- Think of the absurd. A freshly completed construction project that theoretically links up several buildings, renovating and connecting a wide array of facilities and amenities, only to discover the construction teams handling the various jobs really weren't working together. The result: a costly duplication of separate sets of ducts for heat, water, electricity, in addition to unnecessary expenses tied to setting up discrete elevator banks, fire alarms and sprinkler systems. And the financial fallout doesn't end with construction cost overruns. These duplicate systems, which don't add a penny of additional revenue, tack thousands of dollars onto the building's operating costs and reduce its efficiency.
Not likely in modern construction? Perhaps not, but it's pretty commonplace in the world of operations support systems (OSSs). This hasn't stopped service providers from entering into an expensive love affair with OSSs, which are designed to support functions to create, provision, troubleshoot and bill for network services. For all the cost and associated hassles, service providers are infatuated with the strategic value promised by integrating these systems.
Last year alone, operators spent approximately $16.5 billion on OSS technology, according to a recent study by the TeleManagement Forum (Morristown, N.J.). And that doesn't count the billions spent annually on integrating the systems. That's a huge collective dent on the industry's balance sheet. It also represents a leap of faith by corporate executives and investors alike, because the benefit of an OSS is not as readily apparent on the bottom line as that of, say, a snazzy new application or ad campaign.
Why does it cost so much to buy and build an OSS? And why so uneven a return on investment (ROI)? Many providers and other experts believe the problem lies in the lack of communication. Many components of OSSs-billing, customer care, order entry, provisioning and service creation-often don't work well together. In many cases, they don't collect data in the same format or use common protocols to communicate. All too often this means that these systems may perform well individually but can't readily exchange or integrate important information that will improve their efficiency. And times may only get tougher. As services converge, either through corporate mergers or the creation of next-generation networks that bundle services on a common infrastructure, the inefficiency of these component systems is compounding at an alarming rate, experts say.
Too many moving parts
The problem of OSS integration isn't new. As long as providers have offered services provisioned over more than one network, some experts have called for tying together back-end systems. Right now it's just getting harder to keep pace. The number of disparate OSS components maintained by each service provider is multiplying, thanks to the rash of mergers and acquisitions (M&As) and services being offered. As those redundant, unconnected components multiply, service providers find themselves returning less profit and spending more on administrative overhead. Ironically, the growth from M&As becomes completely counterproductive to the supposed benefit from "economies of scale" gains that many executives cite as justifying these deals.
So a bad problem just gets worse. "Whenever you buy an OSS, you can take the license fee and add 50 to 150 percent right off the bat for integration costs," says Mike Schwartz, executive director of OSS strategic planning at Telcordia Technologies Inc. (formerly Bellcore, Morristown, N.J.) and one of the founding fathers of OSS technology. "We've had [service providers] tell us they spent three times more on integration than they spent on the OSS itself. A few of those companies dumped their OSS provider and came to us for another solution. But most of the time, they just keep pouring money down the rathole until it works."
Service providers confirm that their wallets are feeling the pain of OSS integration. "We're spending between $50 million and $60 million on an inventory management system, and about three-quarters of that is just auditing what we've got so that it can be tied together," says Eric Nelson, vice president of network technology solutions for ACSI Network Technologies Inc. (Annapolis Junction, Md.), a subsidiary of the fast-growing competitive local exchange carrier (CLEC) e.spire Communications Inc. (Annapolis Junction, Md.). "That's about half a million dollars for each node, and those are all expense dollars. There are very few CLECs ready to tee up that kind of cash."
And the cost is growing, according to the TeleManagement Forum's report. In 1998, providers spent more than $2.2 billion on OSS-related systems integration, more than they spent on element management, billing or service assurance. The problems-and related costs-in integrating OSSs will vary by provider. But it's safe to assume that for incumbents, the greatest costs come in integrating legacy OSSs with those for next-generation services. Their goal is to keep costs down as they broaden services. Competitive local service providers traditionally face major integration costs in setting up provisioning and ordering systems, which reflects their goal of getting new customers up and running quickly.
Drawbacks aside, OSS integration has had its proponents for years. They justify their support by pointing to the theory that operators can save money and reduce the need for additional staff by combining redundant systems and building "flow-through" capabilities that allow order entry, provisioning, operations and billing systems to interface directly with one another. The theory's logic remains sound, experts note. It's the execution that often falters.
Service providers spend too much on OSS integration, experts say, for two reasons: They don't know what they've got, and they don't know how to tie it together. "The root of the problem is not the OSS infrastructure," says Steve Gordon, director of marketing at the OSS consulting practice of ADC Telecommunications Inc. (Minneapolis). Rather, he claims, the problems lie in "bad data," which comes in the form of information that is inaccurate or data that is so randomly collected and dispersed that no central repository can tell operators what's happening in their systems. "I've seen [providers] spend money to build out their networks, only to find they already had dark fiber in those locations. Any time you do those kind of redundant processes, it costs you money."
The most difficult part of integrating operational and billing systems is finding a way to collect the data, says Charles Arsenault, senior adviser for systems administration at network service provider Teleglobe International Inc. (Montreal). "It's difficult to get billing information from different sources, correlate it and augment it with data from other systems," he says.
It may be difficult, but funneling all the relevant data into one place is vital for not only integrating systems but also helping them function afterward, says Nelson. "The greatest expense [in OSS integration] is getting the inventory information together," he says. "All of the design and flow-through tools out there today are useless until you get your inventory loaded."
The inventory problem is most acute among competitive local service providers, which frequently purchase their administrative and billing systems on an ad hoc basis without considering the subsequent need for integration, says Schwartz. These newer carriers, such as e.spire, are dealing with the problem through a combination of next-generation OSS integration packages-including the Telecom Business Solution from MetaSolv Software Inc. (Plano, Texas) and Net/Expert from Objective Systems Integrators Inc. (OSI, Folsom, Calif.)-and in-house data warehousing systems that make it easier to port administrative data between applications, says Nelson.
incumbent concerns
More established providers have greater concerns about linking legacy OSSs to next-generation systems that will support IP-based services. "The incumbents have a different set of problems because they have OSSs in place that can't support next-generation network services," says Schwartz. "For them, the question is how to provision traditional services and next-generation IP-oriented services without having to build two different systems." Companies such as Telcordia are providing tools for extending legacy OSS data to new service models, Schwartz says.
Under federal regulations, incumbent local service providers also must provide electronic access to their OSSs before being allowed to offer long-distance services. This electronic connecting of OSSs-called electronic bonding-is still in its infancy, but some progress has been made. Just last month, SBC Communications Inc. and Allegiance Telecom Inc. (Dallas) announced electronic bonds between their OSSs.
Whether the problem is inventory or new service rollout, virtually all service providers find it difficult to cobble together the disparate back-end systems they currently support with "customer-facing" systems such as billing and customer care. "The basic problem is that a lot of service providers don't understand what each of those systems can do," says Thomas Stergios, president and CEO of Presidio Consulting Inc. (Oakton, Va.), which specializes in billing systems and OSS integration. "Some [providers] look at them as discrete systems-billing, general ledger, order entry-and others look at them as discrete operational groups. They're both wrong. They need to see the OSS functions as a business process and then get integrated systems that can support the various functions in that process."
It's not unusual for providers to purchase off-the-shelf systems from different vendors that provide redundant functions, Stergios says. "They just buy these big boxes at $6 million to $8 million without modeling them to see how they might work together."
Standards such as common object request broker architecture (CORBA) and telecommunications management network (TMN) have been proposed as guidelines for building the interfaces that will let OSS component systems communicate. But service providers say not enough OSS and network element vendors support them. "We are at the mercy of software and network element vendors," Bob Ferry, chief information officer at wholesale service provider Pathnet Inc. (Washington, D.C.), said at the recent Global TMN Summit '99 conference. "We are trying to fit the square TMN peg into the round hole of network elements that don't support it."
integrate now, benefit later
So if OSS integration is so expensive and fraught with pitfalls, why do it at all? Some service providers still bet that the long-term cost savings of a common OSS infrastructure will outweigh the up-front costs and headaches of doing the integration.
SBC and its Pacific Telesis Group (PacTel) unit gave this belief some credence late last year when it reported that the merging of their organizations and the subsequent integration of their OSSs has yielded real benefits for their combined customer base. Repair times were reduced by an average of 60 percent in the 12 months following the merger, and service installation times were down by an average of 80 percent. SBC and PacTel also said they were able to gain additional savings by combining their resources-people, skills and operational systems-although they did not get more specific.
Other providers also report positive results from smaller integration projects. Williams Communications Group Inc. (Tulsa, Okla.) in 1998 deployed an automated service activation capability, the Service Activation Manager, that has helped the wide-area network (WAN) operator provision services more quickly, although the company also neglected to give specifics. "Moving forward, service activation may be bridged to upstream systems, creating larger, fully automated processes, " says Mike Modugno, project manager in network information technology at Williams. Upstream systems actually provision circuits.
Despite the high costs, experts say, OSS integration will pay off both for those building anew and those providers trying to nudge existing setups into the 21st century. "For incumbents, which are limited by their size and price caps on some services, the question is how to get earnings growth," says Schwartz. "For most of them, the answer to that question is to cut costs, and then cut costs some more. For them, the concept of OSS integration is becoming strategic because it can lead to lower costs down the road."
Competitive local service providers have a different set of issues because their financial success depends largely on how many new customers they can sign up-and how quickly they can do it. As a result, their OSS investment revolves primarily around new, faster provisioning systems. As they are called upon to show a profit, however, the broader picture of OSS integration will come into focus, says Nelson. "Part of the reason the OSS is in the shape it's in is because it's seen as overhead," he says. "As people start to focus more on cost management, the issues [of OSS integration] are going to get more attention in the boardroom."
Tim Wilson is editor at large for InternetWeek. He can be reached at tbwilson@cmp.com.
By: Tim Wilson
Copyright 1999 CMP Media Inc.
|
