From EnergyPulse: (June 02, 2004)

Given two competing solutions to the same problem, the simpler one is best. "Occum's Razor", William of Occum, 14th century logician

ESG serves more than 80 clients in over 100 deregulated energy markets, and when discussing their billing requirements, we are often asked whether we can support "complex pricing and billing". Particularly in the electricity market, the conventional wisdom declares that marketers must offer complex pricing and billing features to be successful in the competitive supply business. Marketers need go beyond the conventional and understand

• who typically benefits from complexity and who usually wants to avoid it
• which aspects of complexity must be most carefully managed and why
• how Occum's Razor can cut away inefficiencies and improve cash flow

The economic function of price is to reconcile what customers are willing to pay for a service with the cost of supplying it. While there are exceptions, most energy customers prefer simple pricing they can easily understand. Complexity is usually created by the supplier to help manage cost and maximize profit. It is a rare customer who gets satisfaction out of MDCQ calculations or demand ratchets but utilities have used them for years to manage their exposure to capacity costs.

Expectations are higher for competitive marketers but they face the same pressures. Complex factors determine the cost of serving customers and marketers must account for them in the pricing scheme.

In the best case, a pricing package meets the complex needs of the supplier but is presented to the consumer in a simple package. One consumer industry we all know offers a fixed price for contracted block volumes in both on and off peak periods on a pre-paid, take-or-pay basis with post-period settlement of excess volume at a premium per unit price. They call it "500 free minutes with 1,000 night and weekend" and cell phone users seem to love it.

Even when a complex price scheme efficiently hedges supply costs and can be sold to customers, there are operational factors to consider. Pricing complexity comes in different forms which affect billing operations in ascending order of importance as follows:

1. Multiple or complex formulae for the calculations
2. Multiple data sources providing inputs to the formulae
3. Likelihood that data sources will be subject to restatement requiring cancel/rebill activity
4. High volume data sources with potential validation problems
5. Data sequencing and cross validation requirements
6. Manual processes or user discretion in utilizing data
7. Number of plans where the billing process differs from plan to plan (accumulated complexity)

Complex formulae need only be programmed once and good servers make short work of the most daunting calculations as long as all the required input data are available and valid. Automating the acquisition and validation of complex input data is the real challenge in complex pricing schemes. This is particularly true for energy marketers because they depend on utilities or ISO's for much of the critical input data. Those institutions have not yet achieved the standardization or reliability of data exchange that we take for granted in financial services or telecommunications.

In this type of environment, cash flow is at risk if complex price schemes are not carefully analyzed with respect to the billing processes necessary to support them. Ill considered complexity in pricing schemes inevitably creates costs that are a drain on the business.

• Complex schemes require business analysis, requirements definition and test scenarios with user acceptance criteria.
• New data sources require the design, configuration and testing of new acquisition and validation routines.
• New calculation methodologies may require modifications to the billing engine with attendant regression testing to ensure the stability of existing schemes.
• Personnel require training in new manual processes and exception resolution.

• Billing processes that require manual intervention or user discretion eventually increase FTE count when the business grows. Spreadsheets are not free.
• High volume data, multiple or interdependent sources increase the likelihood of validation failures (Occum understood Murphy). {Exceptions/Meter} x {Hrs/Exception to resolve} x {$$/Hr} = $$/Meter. Exception management costs scale with your business unless you can reduce the rate of exceptions as you grow. Working capital costs
• Waiting for required data delays getting bills out and cash in.
• Resolving exceptions delays getting bills out and cash in.
• Manual processes (especially errors) delay getting bills out and cash in.

• Delays in billing, high levels of cancel/rebill activity and anomalies in manual processes create customer dissatisfaction and billing disputes.
• Call center activity heats up. Disgruntled customers wait for a collections call to air their frustration, delaying payment and increasing the likelihood of bad debt.

• Customer dissatisfaction leads to lost business. The monthly bill is the only regular interaction most customers have with their marketer. When it is late, confusing, or frustrating to manage, the customer will go elsewhere.

Accumulated complexity renders otherwise supportable pricing schemes dysfunctional by layering together schemes that require divergent or even conflicting business processes. This magnifies all of the potential costs described above and over time the billing process can grow into a tangled mess. Some marketers only confront the problem when their pain reaches this point. Better late than never, but the situation is much easier to avoid than to untangle.

The operational risks of pricing schemes can be minimized by carefully analyzing the underlying operational requirements before launching a price offering.

• Decompose and document each formula in each step of the calculations.
• Identify the data source for each input variable (utility EBT data, ISO downloads, published indices, utility websites, etc.).
• Examine the data structure and presentation protocols of the various sources and identify necessary validations based on how they will be used by the price scheme. This is particularly critical for utility provided EBT data where actual utility practice can vary within documented "standards" for the market. Interval data and usage loops for multi-metered accounts deserve special attention as they are often a weak link in the chain.
• Create a timeline of when the various data feeds are available and determine the frequency of restatement and how long after original posting it is likely to occur.
• Determine what effect data sequencing, interdependence, and restatement have on different steps in the calculations.
• Evaluate the reliability of data flows and the capacity of data providers to respond to problems. There is no substitute for production experience with the data flows in question.

With a clear analysis of where the price scheme creates the greatest operational risk, you can design a billing process using Occum's Razor according to the following principles:

• Get the cash sooner rather than later. Any data flow or calculation that delays the billing process must mitigate greater risks or add greater value elsewhere. If not, rethink why you need them.
• Avoid pointless precision. Sensitivity analysis should demonstrate that additional data and calculations materially add value. If not, leave them out and use an acceptable approximation.
• Strive to automate. In addition to reducing labor costs (and manual errors), it forces a rigorous definition of the billing process and uncovers weaknesses in the data flows. If a billing process is difficult to automate, investigate whether it is unreliable or poorly designed. If not, it needs to generate enough margin to cover the costs of manual support.
• Live within the capabilities of your data sources. If a particular data flow from the ISO is unreliable, depending on it is asking for trouble. If you cannot do without it, make appropriate provisions in customer contracts (e.g. specify an estimation methodology or an alternate data source, and limit your obligation to reconcile with a materiality test).

The design objective is to minimize the operational risks of a complex pricing scheme while retaining the benefits. An elegant solution uses only the degree of complexity necessary to satisfy customers while protecting the marketer's margin. Success is measured by a reliable billing operation that improves both cash flow and customer retention.