Cutting Through the OSS Hype: Notes from TeleManagement World

At this year’s TeleManagement Forum in Nice, France, a number of trends became evident.

Competition has become much tougher, particularly in the inventory management market, where new vendors continue to crowd in. Vendor messages continue to change and become more product focused, but many of the touted benefits still sound the same. Especially in areas related to IP services and service assurance, stories continue to be told that bring very little value to service providers. Integration remains a costly challenge that the TeleManagement Forum, among other groups, continues to investigate, beginning work on the foundation for a shared information or data model with hopes of bringing integration costs and efforts down to a minimum.

Product Focus

During just the past three years, marketing has made a significant shift from componentized suites to configurable tools, and now everyone is talking about tailored products. According to Ed Pinnes, executive director, Integrated OSS Systems Engineering, Telcordia Technologies, Telcordia is even taking a significant step away from its traditional approach to the business. The company has increased its investment in product development and is looking to minimize custom work while pursuing open, strategic vendor alliances for the first time.

In a similar vein, NetCracker, the poster child for J2EE-based, componentized suites, has productized each of its components, touting pre-integration, ease of implementation, and ease of use, according to Julie Wingerter, the company’s vice president of strategy. These are just two examples, however, of a show floor that was vocal about product and reticent regarding professional services or customization.

The shift to a product focus is not surprising. Carriers want solutions that pay back immediately and have grown wary of integration or custom development projects. They have been burned too many times in implementation, and now they expect vendors to take the cost and risk of implementation, configuration and integration on themselves. This only goes to reinforce the message that Steve Laufman of Qwest voiced back in February at OSS World 2002, where he insisted that to compete for carrier business OSS vendors need to invest and listen more, and charge a lot less.

Integration Still a Challenge

Integration remains one of the toughest and most daunting challenges, and vendors offer a million and one new answers for solving ongoing integration problems. In the past several years, carriers have almost always attacked integration from a technology perspective, with a goal of making systems talk to each other. The industry has invested millions in enterprise application integration (EAI) technology—only to have vendors turn around today and point out all of its flaws, after spending years selling its long-term benefits.

At the heart of this problem is a lack of commonality in data models. In simple terms, different systems think about different elements of a carrier’s business in different ways. Naming conventions and relationships are handled differently, which is what makes integrating two or more disparate systems a nightmare.

The extreme cost of integration is crippling service providers. More importantly, this cost can cripple any vendor’s return on investment story, because application-focused numbers often go out the window once the actual environment and integration costs come into play. To help close the gap in data modeling, TMF has launched its Shared Information/Data (SID) model initiative. This initiative focuses on documenting the relationships and roles that business entities and processes play, driving toward a model and methodology dictated not by technology, but by business. The core idea behind the model is to come up with nomenclature and models for relationships that can deliver enough commonality but remain generic enough to allow differentiation. Every process or interaction involves specific information exchanged among organizations that play specific roles. These roles and common interactions need common definitions. Says Martin Creaner, TMF’s chief technical officer, the goal is to “reintegrate the network into the business process and make it a profit center instead of a cost center.”

From a very practical point of view, a shared data model could be useful in implementation and integration. OSS applications can only do their jobs when loaded with the right data, rules, policies and processes that dictate how they interact with or manage the carrier’s business. Carriers spend millions for systems integrators to bring in armies of inexperienced consultants who do little more than map data from one system to another. They bill thousands of hours creating the data element and process documentation necessary to the people who actually implement and integrate systems. Often, this arduous process is poorly managed and can’t keep up with the speed of the business. By the time the documentation is done, requirements change. Turning to common data and process models could help shorten this process immensely, because the process and data ingredients for the most part would be modeled, leaving implementers to focus just on the pieces that are unique to the carrier in question.

Those experienced in OSS integration and implementation see the SID as a good idea but, as with many past efforts, believe certain practicality issues will arise. For example, one OSS vendor executive who has seen it all says a shared data model is tough because “you’d have to tear existing systems apart to change the data model. You might have to throw away an entire OSS. The rule of thumb is that if you have to change more than 20 percent of the code, you should just start over.” He estimates that changing an application’s data model would impact at least 35 percent of an application’s code. Further, he notes that past efforts to create a common data model have failed, and the vendors or carriers that chose to build on them quickly learned they made a big mistake.

On the positive side of integration, says this executive, technologies like Java are already driving down the cost. Where the integration-to-application cost ratio was once 3:1 or worse, in most cases it is now much closer to 1:1. Telcordia’s Pinnes, likewise, cites groups such as the OSS-J and vendor partnerships that actually deliver real products as helpful forces in driving down the cost and difficulty of integration. Open, common technologies in telecom, which perhaps started with CORBA and moved into technologies like XML and Java, have helped reduce technology costs and difficulties, but they have not been able to solve the larger business and data modeling problems that persist and increase cost and complexity.

Inventory Is Hot

The inventory market seems like the center of the OSS universe these days. New vendors continue to arrive from all areas of the business and the globe. Some business actually appears to be happening, particularly with mobile operators. The list of inventory vendors includes Cramer, Granite, NetCracker, Incatel, Arkipelago, Smallworld, Axiom, MetaSolv, Visionael, Eftia, Open Telecom and Ai Metrix.

Axiom and Incatel, two relatively new entrants on the product scene, represent another trend in OSS: vendors that grew up in PTTs and are now productizing the carrier platform. These can be added to a list that includes Open Telecom (Telstra), Sodalia (Telecom Italia) and others—like Telcordia. Axiom (TeleDanmark) and Incatel (Telenor) are both Scandinavian companies looking to expand into Europe, the Middle East, China and other developing areas.

Arkipelago, a Swedish vendor focused on managing network build-outs with its network planning and inventory management application, was swamped at TMW with interest from 3G mobile operators from around the globe. Granite and NetCracker also discussed opportunities in the mobile arena. According to Michael Holm, vice president of sales and marketing for Arkipelago, operators are building out their 3G networks, and in the meantime need to clean up the shop. They’re revisiting their networks, looking to optimize them and find the most effective ways to roll out new 3G resources in line with existing physical infrastructure. As a result, they seem to be interested in nailing down their inventory resources and improving their provisioning capabilities as they adopt ISP-type services and networks. From the air interface to the backbone, 3G operators have inventory needs, and they are out shopping for systems.

However, inventory is a crowded market, and everyone will be vying for the same contracts around the globe. China may be adding 5 million mobile subscribers per month, according to industry estimates, but that doesn’t necessarily translate into a wealth of contracts. Companies like Arkipelago, which thrives off a strong channel relationship with Ericsson, could have the inside track as equipment vendors prime new deals around the globe. The sales process today is like “truffle hunting,” says Mark Mortensen, chief marketing officer for Granite Systems. “You need to get a pig and a shovel, head out into the woods, find the customer, dig ’em up, and drag ’em back out into the open.”

Granite and staunch competitor Cramer Systems are generally considered the leading vendors in inventory management, and like much of the inventory market itself, each faces key questions regarding product development and the future. In the near term, inventory may be the sector to watch for signs of renewed OSS success.

Activation’s Uncertain Future

The activation market remains a conundrum. It seems that most vendors either want to avoid it altogether, or have failed in the past several years to make any real advancements. Outside of Syndesis, the apparent leader in the activation market, activation vendors are not terribly impressive. Two of the primary vendors, Orchestream and Astracon, are telling very similar stories today to the ones they did upon their inception. Their focus remains on IP-VPNs for the most part, and the story is not any more compelling today than it was a year ago. Their arguments stress a reduction in the time it takes to plan and configure VPNs, but the VPN business today may not hold enough dollars for carriers to take notice. Large carriers aren’t in a big rush to replace or cannibalize their ATM and frame relay investments with VPNs and other, more operations focused priorities, are taking precedent.

While the market is not clear on Astracon’s future, questions have especially grown around Orchestream, which recently saw its CEO and North American president leave the company. It did, however, have a customer on hand from Telewest, a mid-tier carrier based in the United Kingdom, which vouched for its ability to automate activation of MPLS VPNs on its backbone, something Telewest decided could not be handled manually.

The Network Data Layer

More general than activation, the network data layer is a mess of systems. Element managers, domain managers, activation platforms, configuration management systems, inventory reconciliation engines, auto-discovery tools and mediation platforms all serve different data collection needs at the network layer, but this burgeoning area shows very little coordination.

Undoubtedly there are many needs at the network data layer, but activating

IP-VPNs does not appear to be among the priorities. New companies, such as Intelliden and the reinvented CoManage, are attacking data, transaction and change management needs. Yet many questions still surround the issue of where network-to-OSS intelligence should reside. With equipment vendors moving up the operations stack, and OSS vendors getting closer to the network, the network data layer is ripe with potential opportunities—if they can only be identified and isolated.

Kimber Lewis, president of Cramer Systems, summed up the carriers’ general attitude toward these complex problems in saying, “The element management area is a mess. From an inventory perspective, I just need some information passed up to me about what’s happening on the network. I don’t really care where it comes from, because I can make sense of it once it gets to me.” Carriers are perplexed about how and where it is best to collect, filter and analyze network data as it relates to operations, and many are testing solutions that help consolidate and simplify network change management and inventory reconciliation.

Service Assurance

Is Still Confused

Service assurance continues to be a big buzzword around the industry, but it’s not clear that anyone is buying it. No need to reiterate here all of the reasons a carrier might be opposed to implementing service assurance solutions. At a basic level, carriers aren’t convinced they need to offer user-specific SLAs and other guarantees and measurements that will expose them to new billing liabilities. OSS vendors insist they will have to offer complex service assurance, lest their competitors steal customers away. However, there’s no huge push to move performance reporting and service level agreements beyond general network measurements, regardless of available vendor products.

Service assurance vendors should get the hint: carriers do not want to hear about automatically giving money back to their customers when services falter.

SLAs and performance reports are most important in intercarrier and large enterprise services. The economies of scale are such that a detailed SLA offering isn’t likely to carry enough of a differentiating economic punch to make a carrier or enterprise change providers. They’ve already learned how painful it can be to shift suppliers, only to see their new one go by the wayside. An SLA won’t heal the open wounds or make carriers and enterprises eager to jump back into the fire. Automated SLAs can help enforce control in intercarrier and carrier-to-large user relationships, which can contribute to improved regulatory compliance and more accurate billing. These kinds of behind the scenes carrier issues aren’t often discussed, but understanding them is critical in today’s sales environment.

Where service assurance is largely succeeding is in the back office. The issue is not so much about giving customers guarantees as it is about improving the ability to deliver services and driving costs out of the equation. Trendium, for example, has had success in measuring key performance indicators and providing visibility into the interactions among key service delivery systems. By delivering critical process metrics and analyses, carriers can prioritize their systems investments, plug revenue leaks, and reduce the time and cost to deliver service. It sounds like marketing, but it actually works. Looking at the business from a horizontal perspective reveals a range of costly inefficiencies that can be remedied. The problem has been a lack of awareness of the holes in the business.

It’s All About the Horizontal

The shift toward horizontal business awareness represents perhaps the greatest trend evident in today’s OSS market. Carriers need to understand their processes in every aspect and element of the business. They need to understand the relationships among customers, services and the network, and this is where spending is going. Carriers are cleaning house and trying to drive costs out of the business. They know they need to be more customer-focused and process-aware. Vertical automation in functional silos is very old news. Horizontal integration at a deep and complex level has proven nearly impossible.

Thus, carriers want relatively simple means to identify, prioritize and fix business problems that are mirrored in their systems and data. That’s why companies like Trendium, Sodalia, CoManage, Intelliden, Cramer, Visionael and others are being lauded for their ability either to capture or manage critical processes and relationships, or derive and deliver the critical information necessary to define and measure business performance without the need for an EAI bus or other massive integration project.

The key is a focus on driving cost out of the business while simultaneously improving one’s ability to serve customers. If a vendor can’t tell a compelling story that covers each of these angles while demonstrating a return on investment in some kind of real dollars, it’s going to be a very long and lonely 2002.



Edward J. Finegold is general partner with Stylus Telecommunications LLC, an independent consulting and tactical analysis firm. He can be reached at ejfinegold@styluscom.com.



Joshua E. Barbach, Stylus’ director of research, contributed to this report.