Note 21: Differentiating charges

Posted: 27 January 2007

Each charge type is used by billers in different ways to charge customers for various products and services. The ability to charge different amounts for different customers and their transactions relies on the quantity and complexity of the information available to make each charge distinct. This note outlines the construction of 'what' is charged for. Transactions-level rating and invoice-level pricing describe 'how' prices are determined for each charge, whether calculated manually or using automatic processes.

Determinants

Determinants are the fields and their values that provide the basis for individual charges to be differentiated. They support the ability to choose one rate or price over another. The biller's scope for offering unique offerings to their customers increases with the number of available determinants.

Four sources of charge determinants are network transactions (i.e. usage), network elements (i.e. details about the network), customer details (e.g. demographics) and marketing (rate) information. The more sources available to determine a charge's price, the greater the scope to charge one customer differently from another (price discrimination).

For example, a measurement of electricity consumption can only have a generic rate applied unless there is some information that allows a distinction between this measurement and any other measurement for this or any other customers. The distinction might be a field on the measurement that indicates that it was off-peak electricity, meter number allowing derivation that the measurement was from an off-peak meter, customer information from the billing system indicating it was from a business premise, or the measurement date indicating that the winter rate applies. Each additional piece of information (determinant) provides the capability of, but does not mandate, differentiated charging.

Examples of differentiation

Customised product offerings provisioned to a customer can provide pricing differentiation. For example, a biller may offer two pricing plans for mobile phone calls. The 'High-use' plan offers cheap phone calls but charges a high monthly network access fee, whilst the 'Low-use' plan has high call charges, but charges only a modest monthly network access fee. Two customers with the same network connections and making exactly the same phone calls will generate quite different monthly invoices based solely on the different marketing determinants provided by their pricing plans.

Examples of differentiation include:

Peak / Off-peak: This information may be provided from the network, or derived based on information known about the customer or service location. e.g. an electricity meter measurement may indicate that a measurement is for off-peak consumption, or the consumption's nature may be derived based on the meter number.
Date: The transaction or (meter) reading date can derive additional differentiation properties such as day-of-week, day-of-year (e.g. Christmas Day), and membership in other date classes (e.g. weekend, summer, school holidays).
Time: Like date, the time of a transaction can derive additional distinctions including measures of peak and off-peak periods (which may differ based on the day of the week, or day of the year), and the appropriate rate when transactions span different periods (calculated using the start and end times).
Location: Charges performed in a remote location may attract a premium, or a mobile phone call performed overseas may be charged differently to one performed at the customer's home location. A calling card may be used when at a location away from home. Location may also drive tax calculations since these vary based on jurisdiction.
Source / Destination: For phone networks, each destination country attracts a different chargeable rate. Reverse-charge calls to your home will be judged as being local, long distance or international depending on their source. A tollway biller may use the source/destination combination to determine a vehicle's journey on their road network.
Channel / Media: Where the same transactions can be performed in different ways then that choice can be a determinant. For example, phone rates can be different for fixed and mobile networks, a regular phone line versus a high-quality fax line, or regular video quality versus high-definition TV.
Response time: A non-network, intangible quality used to distinguish between product offerings sold to customers. For example, customers may pay more for faster service restoration, reduced delivery times, or help-desk response times. The information used to differentiate these charges would be stored against the product-offering or customer, be extracted when supporting the customer, and used when a charge's price is determined.
Customer service: Like response time, customers may purchase market offerings that customise customer service (and the amount charged for it). This may involve the use of special contact phone numbers, individual customer service manager, and priority problem resolution.
Market segment: It is common for consumers and businesses to be charged at different rates for similar products. This can be due to government regulations that places limits on consumer pricing, the volume of business that a corporation may bring to the biller, or market pressures (or lack of them) that influence how a biller generates revenue from their network.
Device: Tollways may differentiate their journeys based on whether they were performed by car or truck. These distinctions would be based on information supplied from the network, or captured ahead of time and derived for billing (e.g. from the licence plate or electronic tag).
Bundled rates: The product offerings a customer has purchased or negotiated can provide details of how a charge should be priced.
Historical information: Details about the customer's previous network use may be used to calculate a differentiated price for a current charge. For example, details of last year's water usage may be used when calculating this year's prices.
Quality: Measures of the quality of the network, or network can be used. For example line quality for phone calls, the heating value of natural gas, or availability for electricity.
Network Options: Transaction fields may indicate where optional features were employed. Examples include where the call price is provided back to the caller (i.e. call back), or demand management was employed to remove electrical supply temporarily. These options, if available, can be flagged to and used (if required) by the rating process.

Moderating pricing complexity

The actual list available to a particular biller in an industry is constrained by their networks' and billing systems' ability to provide the information, and the customer's desire or willingness to be billed based on the different distinctions. Market pressures, industry regulations and customer preferences drive the use of only a small selection of the available determinants.

There is also an operational cost that accompanies each additional level of differentiation. There are more details to be administered, monitored and explained (to both customers and staff) as you move from (say) charging just by date, to date and/or time, to date and/or time and/or customer segment, to date and/or time and/or customer segment and/or monthly spending level. Even when many determinants are available, usually only a few are used in any particular product offering.

There is a balance between re-using a field for new determinant values, and creating new fields to hold those determinants. There is a danger when expediency and lack of design foresight force a small number of fields to provide too much differentiation. At some point, maintenance of the possible combinations held in one field becomes too costly and the distinctions required for new offerings become too hard to implement.

Due to the high-cost and operational impact of deploying new data fields to an existing billing system, the initial implementation of a billing system should invest time and experienced effort to provide sufficient mechanisms and fields to support current and envisaged future determinants. This design investment will likely include fields that are unused initially, but are filled as new transactions, networks or values become needed. Future application development is then performed only when all fields are used and there is sufficient business justification to justify the cost and system impact. At that time, another experienced review should be performed to include as many new fields as considered justified to delay a future application update.

Data definition methodologies (such as XML) may provide solutions that allow new fields (transaction information) to be added to existing interface and file definitions. The extensibility benefits they provide come at the cost of supporting the 'descriptive' overheads of their (XML) definition structures.

An important point to note is that the billing system itself doesn't understand or place any special meaning on the determinants listed above. They are all human interpretations of the fields and their values. The value of 'X' in a field drives billing without the billing system understanding what the field or 'X' represents in the real world. This allows the same billing application to be configured differently and used across a range of industries.

Product Segmentation / Versioning

A set of product features can be combined in different proportions, with added (or omitted) extras, and with variable dimensions such as timing to create differentiated product offerings that appeal to different market segments. Versions can be charged at different rates allowing businesses to capture a larger slice of the available revenue from a broader range of customers.

"Information Rules" - This book outlines how the pricing, compatibility and versioning policies that businesses apply influence the take-up of their products and services. It looks at the 'problem' and 'solution' of product versioning from both the business and customer perspectives. [ISBN: 0-87584-863-X / HBS Press / Carl Shapiro & Hal R. Varian / 1999]