Recovering lost revenue: Where it leaks and how to get it back

Now more than ever, the industry needs to preserve resources and generate revenue. The problem is that venture capitalists and investors are turning into turtles, crawling back into their shells. Money from outside sources is difficult, if not impossible, to acquire during times like these.

What does this mean to service operators? They will have to look within their own networks to uncover new sources of revenue (and maintain profitability). While market share and, subsequently, subscriber growth take a back seat to more conservative practices, cost avoidance and revenue recovery become companies' priorities.

Indeed, the industry as a whole is reporting a loss of anywhere from 2% to10%. The loss cannot be attributed to one source or problem, but rather to a series of issues. Collectively, these issues are the basis of the loss. In fact, many companies already are identifying as much as a 5% revenue recovery just through process improvements--without even addressing the larger problems.

However, this is not easy. During the heydays of the '90s, service operators invested in new markets and the equipment necessary to expand in those markets. Market share became the investor's mantra.

Revenue leakage was largely ignored and portrayed as minimal when compared with the investment returns realized through market growth. This left companies ill prepared to manage the revenue stream and the data that feeds into billing systems creating that revenue. The tools needed to fight revenue loss are, for the most part, nonexistent at small operators' sites and minimally in place at the facilities of larger operators.

However, this is rapidly changing as service providers begin shopping for the tools they need to recover revenue. The choices are many, and understanding the investment return is often extremely challenging.

Where does all the revenue go?

During the last two years, a lot of effort has been placed on finding lost revenue. Most companies start with process improvement programs in the area of billing and customer care. Once they have solved these problems, they begin examining the network. There are a number of industry reports and statistics that classify the most significant areas of revenue loss.

To understand the tools that are needed to recover lost revenue, one must first understand where loss occurs and how it occurs. There are a number of areas in which revenue leakage adds up, but four are reported by the industry as the most significant: call detail records, intercarrier billing, service level agreements and fraud. This article focuses on these four major areas, which also promise the highest rate of return on a company's capital investment.

Call detail records

Switches produce detailed records on every call that is placed. This requires investment in software modules for every switch in the network as well as downstream equipment to pull the billing data from the switches in the network, process the data and feed it to a billing application.

Switches must be configured before they produce billing records. Upgrading a switch's software may result in mistakes that will arbitrarily result in the absence of billing records for specific types of calls. For example, one large service provider received an invoice from another carrier for long-distance traffic that was terminated but could not reconcile the bill with its own records. It took almost a week for the carrier to ascertain that an error had been made during a switch upgrade, and the violating switch had not been producing billing data for an entire month.

Such was a costly mistake that resulted in the development of new procedures to ensure that all switches are thoroughly tested after any upgrades. Still, the task is a manual one.

Inaccurate call duration is another factor. Switches were equipped with the ability to delay the creation of a billing record for a configurable period of time. The period, known as guard time, was implemented when Signaling System 7 (SS7) was first deployed in networks. The result is that many calls are undercharged by two seconds or more.

The fix is easy. Change the guard times to their minimums, and the call durations are more accurate. But guess what the default value is? Vendors set that software at two to three seconds, which means every time a switch is upgraded, someone must check to make sure that guard times are reset.

Switches are not the only source of loss. Polling devices can lose their connections with switches, resulting in lost billing records. Records themselves also may be inaccurate or missing critical data. So the data is sent to an error log instead of being processed. Error logs must be reconciled manually--another costly procedure with very little return. Many other systems in the downstream process have become overloaded and are causes of failure. There is not yet a reliable solution to replace these systems, but the solution is coming.

Then, there is the data itself. Switches provide call detail records in a specific format. Unfortunately, there are hundreds of possible formats. In North America, the Automatic Message Accounting (AMA) format is the standard, but there are hundreds of different formats of AMA. A system must collect all these formats from various switches and reformat them into a single format for the final billing process. The process adds to the cost and complexity of a billing system and introduces yet another source for revenue loss.

Intercarrier billing

After the divestiture of the Bell System, service providers struggled with how to route traffic outside of their networks. They implemented intercarrier agreements with simple billing schemes. As services became more complex, so did these agreements.

Simple access charges now are augmented with new charges based on minutes of usage, services accessed and the types of calls terminated. Managing these agreements has become a nightmare.

To further complicate matters, companies have found new and innovative ways to route their traffic through other service providers' networks at a better rate. These other companies then send the very same traffic to the terminating company disguised as local traffic. The result is that the originating carrier is able to route its traffic through less expensive agreements, while the terminating carrier loses out on potential revenue.

This is commonly referred to as carrier bypass, although many dub this as carrier fraud. In reality, there is nothing fraudulent about the practice. Think of it as least-cost routing.

Figure 1: Normal routing bypass

 

Figure 2: Alternate routing bypass

Carriers can manage this to their advantage. The secret is having the tools and reports available to identify the type of traffic coming into the network from other carriers. The reports alone do not fix the problem, however. They provide the data needed to negotiate higher rates from the carriers sending the traffic.

The issue will continue to be the manipulation of this data by intermediate carriers. Most companies rely on switch-based records to determine how to bill for the traffic. Some companies have changed over to SS7, but its use is not yet widespread. As calls are routed through tandems, the tandems will remove specific parameters from the signaling data prior to routing the traffic to another network. The intent is to conceal the origination of the calls, identifying the traffic as bypass traffic from another carrier. Some companies have seen the calling party number changed to a generic number or eliminated from the signaling stream altogether.

If the concealment can be determined and proved, higher rates can be charged for the traffic based on the contention that its origin is suspect and the data necessary to prove otherwise was not made available. Keep in mind that this is simply the tactic of the moment. As companies arm themselves with the tools to detect the occurrence of bypass, carriers will find other creative alternatives.

One often overlooked section of intercarrier agreements is known as Percentage of Interstate Usage/ Percentage of Local Usage (PIU/PLU). All interexchange carriers must identify the percentages of usage, usually based on their own perspective (and naturally, in their favor). Proving the carriers wrong is impossible without complete visibility to all traffic coming into the network.

Service level agreements

Intercarrier agreements will usually include service level agreements. The SLAs guarantee that a carrier's traffic will be given highest priority (or at least the negotiated priority) by the terminating service provider. Failure to provide these service levels results in penalties that are sometimes substantial.

One service provider (a hub provider of sorts) nearly went out of business because it could not manage its SLAs, and when the network began experiencing trouble, penalties quickly multiplied.

Managing the service levels is one challenge; monitoring the service levels poses a completely different challenge. Companies must arm themselves with data (in real time), identifying at any point in time measures such as answer-seizure ratios. Imagine being able to see automatic performance reports every hour, rather than after exhaustive manual studies.

A robust and reliable quality of service (QOS) tool will provide the data and can be used for reporting a number of other metrics as well. However, these tools are only as reliable as the data provided to them.

The good news is that a wide variety of tools are available to manage SLAs, and this is probably one of the easier areas in which to prevent revenue loss. An added plus is that these tools also can be used to closely monitor and increase network performance.

Fraud

Fraud is perhaps the area of the four that receives the most attention. Fraud vendors have done a tremendous job educating the industry on fraud issues and providing the tools needed to combat subscriber fraud.

Believe it or not, some service providers still do not have fraud detection tools in place and have no plans of adding them. To these companies, fraud is not seen as a prevalent issue.

However, fraud continues to plague the industry, and fraudsters have become more and more sophisticated. A simple search on the Internet will reveal countless Web sites dedicated to the topic of hacking into telecommunications networks and committing various types of fraud. In fact, many fraudsters have turned their sights on the signaling network itself and are actively learning how to compromise network elements causing service interruptions, not to mention accessing free services.

Fraud tools have another use. Law Enforcement Agencies have received new powers under the Homeland Defense Act, enabling them to monitor the calling activities of any suspect (under court order, of course). They have already begun working with the security departments of many telephone companies to receive comprehensive reports on suspect activities.

Producing these reports is chiefly a manual process, requiring personnel to pore over switch-based billing records. Time can become an issue, and these departments need to be equipped with the systems necessary to automate these reports. Not only will an automated process increase the margins on providing these reports (law enforcement agencies have been authorized to pay for this information), it also will decrease the lead time required to get the critical data into the hands of law enforcement.

Don't underestimate fraud. It may not be as significant an issue as, perhaps, intercarrier billing, but it is still a prevalent problem and will continue to increase in complexity and sophistication.

How to become the CFO's hero

We have identified the problems. Now, how do we solve these problems? What does a company need to fight revenue loss in these areas collectively and still maintain company ROI objectives?

The answer lies in the core of the network. The signaling network provides the richest source of data for any service provider. Capturing this data is simple--much less complex than creating, capturing and processing switch-based call data. Once this data has been captured and stored in a database, it may used by any department. Its largest value is in the area of revenue assurance. Let's look at each area individually.

Call detail records

Addressing CDRs is, perhaps, the first step. While subscriber billing will require switch-based billing data for some time to come, there are other types of traffic outside of the switch that are better suited. SS7-based records contain much more information than switch-based records; SS7 is the perfect source for the applications discussed previously.

It makes sense that the first phase in tackling revenue loss is to capture the data that already exists! Yes, the data exists in the network today. All that is needed is a means to capture the data and store it for use by applications.

A number of vendors currently provide monitoring systems used by engineering and operations departments to collect network data and provide statistics and surveillance tools to help manage the network. However, these systems rely on hardware "probes" that sniff signaling data off data links and process the data into CDRs.

Probes are sufficient for monitoring and surveillance applications. However, they represent new opportunities for revenue loss due to the nature of the hardware itself and the way it connects to signaling links. The probes consist of shelves, equipped with interface cards, which are then connected to a signaling link (usually via a patch panel) typically in a simplex configuration. The hardware was not developed within the requirements of a billing system (Telcordia specifies 99.9999% reliability in billing systems) because it was not designed for these types of applications.

Still, many companies have deployed these systems and have begun feeding their revenue tools with the data derived from these probes. These companies are realizing that the solution is very expensive, error prone, and not the best means for revenue applications. They are great for monitoring systems--but not designed for billing.

New platforms are becoming available. They still connect to signaling links like a probe, but they provide far greater reliability than a conventional monitoring system and support the performance required for revenue applications. One signal transfer point (STP) vendor has even added software to its STP link interfaces to capture the signaling data at the link, copy it, and send it to a database for use by other applications.

These next-generation systems can be placed side by side with existing monitoring systems and used to feed revenue-critical data to billing departments much more efficiently. In the process, certain types of CDRs may be offloaded from the switches in the network and migrated to SS7.

The concept is not new. It was documented in a patent in the early '90s, clearly detailing the use of SS7 as a source for billing data. Several service providers have migrated to SS7 for a billing feed with incredible paybacks. Once a company moves to SS7-based CDRs, it has captured the data that solves the other problems in the network.

Intercarrier billing

SS7 is by far the best source for producing intercarrier billing records. To begin with, SS7 is indisputable. Switch-based records continue to be questionable, especially because they are created after a call has supposedly been placed. In fact, at least one company has been charged with placing false calls to increase billing records to other carriers.

SS7 on the other hand is created by the switches to connect calls in the network. If there is no signaling, there is no call. Therefore, SS7 records are indisputably far more accurate that switch-based records. SS7 also provides much needed visibility to detect bypass scenarios.

For example, one service provider used SS7 to create CDRs, which were, in turn, stored in a common database. The carrier then used a report generator (customized for creating intercarrier billing reports) to process the data. The carrier was then able to determine which carriers were sending traffic into its network without providing originating calling numbers or even originating carrier identification. Once the carrier was able to capture (and quantify) the traffic, the carrier then went back to the other carriers and charged them higher tariff rates, on the basis the carrier did not get the information from the traffic necessary to determine its origin.

The carrier financed the implementation of the collection system through recovered revenues, in months instead of years. The carrier also solved a bigger problem. The polling systems responsible for collecting AMA records from the switches were quickly being discontinued, requiring a substantial investment to replace them with newer devices. By migrating to SS7-based CDRs, the carrier has been able to eliminate this equipment and maintain the newer polling devices for subscriber billing records. The carrier will soon be looking at ways to use SS7 for subscriber billing as well.

Not only did this migration eliminate the polling equipment in the network, it quickly simplified the downstream processes required for switch-based billing data. Collectors, parsers, correlators and formatters are no longer needed. Furthermore, the carrier is now dealing with one common data format (comma-delimited files are much simpler to integrated into applications than, say, an AMA format). Now that the company has infrastructure in place to collect SS7 data, it can begin adding applications to take full advantage of the relatively small investment.

Service level agreements

A service level agreement can cost a company millions of dollars in penalties if tools are not available to monitor network performance in real time. Real-time monitoring is only possible through SS7-based CDRs (switches create CDRs when a call has been released).

A reliable QOS application will provide a number of metrics that can be used to manage SLAs as well as ensure network performance issues do not go unnoticed, resulting in lost customers. However, QOS should not be seen as simply a defensive measure. It also can be used to verify that calls going to other networks are handled according to SLAs.

These tools also can be used to determine trunk utilization, allowing a company to carefully groom its networks, resulting in cost avoidance. One RBOC has realized millions of dollars in cost avoidance by examining the trunk utilization on every trunk in its network, providing real-time data during any period of time. The company now engineers trunks based on actual traffic statistics and makes growth decisions based on actuals rather than models.

Several other carriers are looking at similar programs as they begin reducing operational costs and closely manage their network growth. Only SS7 can provide the data needed by QOS and trunk utilization reports to effectively and efficiently manage SLAs and trunk growth.

Fraud

The biggest investment in fraud systems is typically the additional investment in switch software required to produce CDRs or the deployment of conventional monitoring systems to collect and create SS7-based CDRs. However, we have already covered a phased approach to implementing a revenue assurance program based on SS7 data. The phased approach not only provides the business case, it ensures a quick return on investment for collection equipment.

Fraud applications can be scaled for networks of any size. They offer an added layer of revenue protection. While a fraud business case may not be substantial enough to warrant investment by itself, fraud should be a part of a larger revenue assurance investment strategy for all service providers.

When approached in this manner, fraud detection becomes affordable and strategic. Such applications also allow service providers to efficiently support local law enforcement agencies as well as the new Homeland Defense Department. If all service providers embrace the concept, it will prevent eventual government mandates (like CALEA).

Think of fraud detection as one tool in the toolbox of revenue assurance tools.

Conclusion

Revenue assurance requires the cooperation and teamwork of several departments. If companies choose to focus on just one or two areas of revenue loss, they will never realize the investment returns that justify purchase decisions in today's tight economic state.

If companies implement a phased approach to revenue assurance, they will realize substantial returns and realize significant increases in profit margins. At the core of revenue assurance is the very data that make it all possible. The data already exists throughout the world today in the form of signaling. SS7 has gained wide acceptance as the principle source of call detail records, which can be used by numerous applications.

The author has classified SS7 data as the three Ws of your network: Who uses the network, for what do they use the network and when do they use the network. Understanding the three Ws is critical to any service provider, and SS7 provides a single source for this intelligence.

Because SS7 already exists in the network, the data does not have to be created. It needs to be captured through collection systems and stored. The process is significantly simpler than what is required to create, collect and process switch-based CDRs. The result is a much quicker return on investment for a revenue recovery program.

Once the data has been collected, it can be used by numerous applications. All departments may use the applications. Therefore, SS7 should be a shared resource (not just an engineering necessity). Marketing departments, sales departments, finance departments and customer service departments will find that knowing the Three Ws is an important tool. Many will wonder why they didn't learn about it sooner. More importantly, the data allow service providers to recover lost revenue and implement revenue assurance programs that will ensure the company's margins are optimized and that profitability is maintained.

Travis Russell is Product Marketing Manager for Tekelec.

Visit Tekelec online.