Customers expect their billing for any service to be perfect. So does the service provider. In the telecom industry, there are only three indicators of quality that the customer receives: price, actual service and the bill.
The bill is the only physical representation of the service. The customer expects it to be accurate, timely and complete. But perfect billing is not as easy as buying the perfect system. In some cases service providers have spent hundreds of millions of dollars only to discover that their billing system still does not deliver what was intended. The service provider often forgets that the system is only a tool and must be surrounded by tight business processes and effective management tools to complete the picture. A system may do exactly what it is told, but unless there are valid requirements, process controls, problem resolution tools and metrics to measure the health of the system, it can still fail. The key objective is to prevent and identify errors as early as possible. Rework and recovery are the enemies of the billing process and increase costs significantly.
Why Do We Need to Understand the Process?
Billing is not the mystery it seems. It is a logical process that can be mapped, dissected and understood at the business user level. The more people who understand the process, the more it can be analyzed and the less likely that valuable time will be spent on unnecessary "witch hunts." Most organizations react to a dip in revenue by blaming the billing system for either dropping records or mis-rating events. By implementing the proper process controls, however, an organization can constantly monitor and evaluate the quality of billing, and look to other possible problem areas when reduced sales orders, increased customer cancellations, etc., occur.
It is critical to understand the process because if there is a billing problem, the root cause must be identified as quickly as possible, the impact assessed and corrective action taken to assure that the problem does not recur. Ideally, all this should happen before the customer receives the bill.
In the Beginning ...
To understand the process, start with the creation of a billable event and follow with a thorough understanding of all the event collection parameters, editing criteria for billable vs. nonbillable events, and data added to records at the initial stage. At this point, address the following questions:
How often are billable events collected from the source?
How do I know that all events have been collected?
Are there sequence numbers on the events so that I can determine if there is a gap in records received by the source?
When in the process is it decided if a record is billable (complete) or non-billable?
What is a billable or non-billable record?
Do I maintain a trend, by each source, to assess whether there is an abnormal increase or decrease in the number of records or in billable vs. non-billable distribution?
Is there an archive file that stores all event records for additional analysis and retrieval in the event that a recovery process is necessary?
What, if any, information is added to the event record at this point, and what drives the values of what is added?
Understanding everything that happens in this portion of the billing process is critical to understanding all subsequent steps in the billable event life cycle.
Who Should Be Charged What?
The next logical step in most billing systems is the rating process. Billable events from the collection process are passed to rating to obtain the initial rates for the event. This step is driven by the most sacred of all billing system tool, the rate tables. A well-designed system uses tables that can be controlled and maintained by the users, instead of hard coding embedded in the programs. This facilitates a rapid response to market circumstances and permits easy changes to product/service/event pricing. The goal is to change the programs as infrequently as possible and to allow the tables to drive the rates. This flexibility also brings responsibility. There must be a knowledgeable set of users who are intimately aware of the rate tables and the potential interrelated impacts associated with any changes to the tables. There must also be tight controls on who can change these tables, and a process to ensure that it is done accurately. Ideally, there should also be a test bed of sample customers associated with the rate tables to permit performing "What if?" analysis. This would allow the originators of a possible rate change (e.g., product managers) to assess the actual impact the change will have in production and to avoid surprises down the road. Questions to ask regarding the quality of the rating process include the following:
Who should control the rate tables?
What are the drivers that change these tables, and what sign-offs are required from marketing and finance?
How can you determine the accuracy of changes prior to production? Are rates validated each time events are rated to assure that problems are detected at the earliest possible point?
What tools are in place to support this process? For example, can rate groups be updated vs. updating each item on a line by line basis? Are report capabilities in place to review only the changes and ensure that nothing else has changed?
Accurately rating an event requires matching the event to a customer so a particular product/plan can be determined. Once this is done, the system will access the appropriate rate for those conditions. This process implies that either the customer file be extracted on a regular basis or that the rating process must read the customer file in real time. The owners of the rating process must be able to determine that the customer information is accurate, timely and complete. Ideally, the data should be as current as possible (i.e., less than 24 hours old) to assure that the customer is billed accurately. If a copy (extract) of the customer file is used, controls must be in place to assure that all customer information has been included. To understand the quality of this process, consider the following questions:
How often is a new copy obtained?
What tools exist to ensure that the whole file was extracted? For example, are there control reports provided on-line that review and validate the number of customers, file size, etc.?
How is the data validated? For instance, are there thousands of customer records with no customer number or product/plan type?
Are records for cancelled customers included, and if not, how is usage past the cancellation date handled?
What is appended to the original event record, based on what is obtained from the customer file, product/plan, billing cycle, customer number and so on?
How are records sorted and distributed from this process: into individual cycle billing files or just output by date?
Since the billing process relies on receiving valid data from other systems (customer data, accounts receivable data, and so on), these inputs must be accurate.
Is the Billing Correct, and Are We the First to Know?
Once rating is complete, service providers bill end-customers on a regularly scheduled basis. Some do this monthly; others with substantial volume bill on a cycle to even out cash flow. A critical piece of the actual billing process is the ability to determine if the billing is correct and meets quality standards before it is sent to the customer. Billing verification requires carefully thought out processes and support tools. Primary among these is a comprehensive "invoice verification checklist" that can be used as a guide to determine the accuracy of all items on the invoice. With complex services now combined on a single bill, a checklist of this nature can include from 200 to 400 items. These range from determining whether the balance forward is correct to verifying the content and accuracy of the marketing/dunning messages included on the invoice. Tools commonly required to support this process include automated sampling selection, call repricing tools, tax table visibility, product manuals for discount or pricing plan structures, and visibility to rating tables. In smaller operations these may start out as manual processes and evolve to automated processes, but the key factor is to ensure that this validation process occurs. The worst case is to find out about a problem from your customer.
As an output of the billing process, an "invoice register" should be validated and generated by the billing process. This should summarize all customers who have been billed, including the number of events, minutes, dollar amount due, and so on. A summary at the end of the register should provide totals and information that can be used for trending and analysis, including items such as average calls per customer, dollars/minutes of use, total customer with zero usage, and total customers by product type. Ideally, this should be the same file of information posted to receivables once the billing is approved.
But billing is not always perfect. When a problem is detected, it is important to have an ad hoc tool that can rapidly analyze and report on the impact of the problem-such as how many customers, events and dollars have been affected by the error. It is then a management decision whether to rerun the cycle or to apply future credits and handle the customer impact via customer service. This is a difficult decision, but at least it is an informed decision. The information and the impact can also be immediately communicated to customer service and finance so that everyone is prepared and there are no surprises.
In addition to all of the billing accuracy concerns, constant vigilance is required to ensure that billing runs on time and includes all traffic associated with the cycle. Accuracy is a key factor, but timeliness and completeness are also valid customer expectations. No one wants to be billed two months late, or to find only a portion of their traffic listed on their current bill. An event which is billed late and outside of the current billing period is suspect and may not easily collected. This obviously affects corporate finances.
What Are All of These "Leftovers"?
Managing suspense/error files requires constant attention. Unless errors are corrected and recycled quickly, the objective of timely billing will not be achieved. Millions of dollars are lost or written off due to records that are never billed or billed so late that they are uncollectible. In the case of providers who bill events via the LECs, very stringent rules on the age of events govern what can be billed on a LEC invoice. Calls or events that are outside of these time frames are returned as unbillable, and the related revenue is lost.
Most commonly, errors occur at two critical points in the overall process: rating of the initial events and the actual invoicing/cycle billing process. Service providers must minimize the number of error files so that items not easily retrievable (i.e., from two to three sources), are not dropping out during the process. There must be ad hoc and production reports to enable the correction process owners to stratify and view the data in a variety of ways, so that the root cause and the corrective action can be quickly identified. All suspended records should also contain an error code, indicating why the calls were rejected, and to facilitate key reporting, analysis and corrections. Just dumping the records into the "bit bucket" to let someone wade through, one at a time, is a waste of valuable resources and limits the likelihood of timely recycling.
Another cardinal rule of effective error file management is: Don't change the record manually to make it billable. Countless hours and personnel time are often wasted by fixing records on an individual case basis rather than determining the true root cause(s) and fixing the problem at the source. If the process is working correctly, then the records can be recycled and will come out of suspense naturally. From an auditing standpoint, it is a bad idea to manipulate the event records; they should stand as they are, with the actual data as created.
How Do We Balance the Checkbook?
The most commonly asked question surrounding the billing process is: Did I bill for everything I processed? And if not, where is it? This has been historically referred to as "carried to billed" measurements, but in today's environment where so many companies are reselling the services and products of another, measurement becomes even more complicated. If you are not the primary service provider, this is now a three-legged stool, carried to billed to invoiced.
In the internal environment, quality billing processes must ensure that all event records are added, subtracted and continuously accounted for to assure that no records are lost in the revenue stream. In a cycle billing environment, this assurance is more complex but still achievable. All additions to each of the cycle "to bill" files must be accounted for each time event records leave the rating process, to be sure that the new overall balance reflects all new records processed. At the billing time, there should be a comparison against the invoice register to ensure that all records expected to bill have been correctly extracted and billed from the event files. This should also follow through to the end of the process. Have all customers who were billed actually had an invoice produced, and were the mailing totals equal to the customer totals? Were the receivables from the current billing posted back to the AR system correctly, and does the new AR total now reflect what was just billed? Whether a system is simple or complex, it should be subjected to this type of balancing. There are countless legendary stories in the industry to prove the need for conscientious attention to continuous quality auditing, balancing and controls in this process.
How Are We Doing and How Will We Know?
A quality billing process demands appropriate measurements and metrics that provide the entire organization with a method of evaluating the revenue process. These measurements and objectives must be continuously analyzed and reviewed. If none exist, industry standards should be used as a starting point, with improvement objectives set from there. Also, remember that "you get what you measure," so these targets and objectives must be carefully set. For example, is speed more important than accuracy? Which are you measuring? What tools are in place to support obtaining the data necessary for the measurements? Examples include: standard reports produced to facilitate trending and performance and ad hoc reporting capabilities against both current and historical information to research unusual fluctuations in performance. Is the data provided on a timely basis, or after it is too late to do anything about it?
It is also important that everyone in the organization understand the measurements, his or her role in contributing to the outcome and what that outcome is. "Give them the answers before they ask ... good or bad!" Results should be regularly released to a wide audience, both inside and outside the billing process. This type of communication goes a long way in eliminating the mystery and "black hole suspicions" that typically surround the process in most organizations.
So Who Does All of This?
Unfortunately, in most organizations, the individual pieces of an end-to-end billing process are fragmented between many groups and functional areas. It becomes a political football; everyone wants to be in charge, but no one wants to be responsible or accountable for the whole thing. Whenever there is a problem, each group is sure that their piece is working just fine and that it is obviously the fault of some other group.
In true quality billing organizations, the end-to-end process is managed as a whole. It does not matter as much where this function reports-IS, Finance, etc.-as long as it exists somewhere. From a personnel standpoint, the individuals who manage this process are typically business users vs. IS types. Excellent candidates for this group usually come from customer service (they have seen everything that can happen on a bill!) or finance, or individuals with a technical orientation. This revenue group also requires the tools necessary to perform these functions. The questions to ask include these: What kinds of revenue reporting tools, e.g. data warehouse, user controlled reporting packages, access to complete records for both billed and unbilled traffic, currently exist within the organization? Are these user controlled and easy for a non-programmer to execute? What types of reports are really necessary to provide a clear, concise view of the billing process, and where are they in the current priority list of "things to do"?
"Perfect billing" is an expectation of both the customer and the service provider. At times it seems an elusive goal, but is one that must be pursued continuously. Doing the right things correctly the first time reduces costs, maximizes revenue and increases customer satisfaction on all fronts. Implementing the appropriate quality billing processes, procedures and organizational structures is critical to staying ahead of the competition.
Linda Gimnich is a Principal with The Management Network Group, Inc. (TMNG). TMNG is a multinational management consulting firm that specializes in leading the development and implementation of competitive communications company strategy and operations. David Ellingsworth, a senior consultant with TMNG, Inc also contributed to this article. Comments may be directed to Linda Gimnich at: TMNGLinda@aol.com.
Email this article
Add a comment
Printer version
Order reprints
RSS Feed
Bookmark article