The Rise of Usage Based Billing

Usage, or activity based billing is on the ascendant. In a recent whitepaper, iGR reported that over 28% of SMBs they surveyed are already using usage based billing and of those considering changing their billing practices 49% are actively considering a usage based model. In this article we dig into the pros and cons of a usage model and talk about what usage models really like.

Pros of usage based pricing

Get what you pay for: From the customer’s point of view, the obvious appeal of usage-based billing is that you only have to pay for what you use. This seems fairer and you would expect it to be more cost effective, although, at the end of the day, the measure of cost effectiveness will depend on the actual price points and usage pattern.

Lines up pricing with customer value: usage models are often naturally aligned to customer value. The more a customer uses a service, the more value they derive from it, and so, more they are prepared pay.

It is often important to create different price tiers to cater to users with different levels of price sensitivity. Companies will often take the approach of basing differential pricing on feature differentiation – so lower priced plans get fewer features. If you can identify clear customer segments associated with each feature set this approach can be very effective. However, if the feature differentiation is arbitrary the customers on lower plans may just feel they are getting poor value for money. In this case, a differentiation based on usage can be much more powerful.

Cons of usage based pricing

Unpredictable cost: One of the benefits of a simple recurring billing model is predictability; the customer can plan around a fixed recurring cost. Usage-based billing can be unpredictable – particularly if you do not closely monitor your usage. The worst case of unpredictability is pricing shock. One of my colleagues recently reported that her cell phone bill had come in around 4 times its usual price – quite a shock. She had dialed in to a number of conference calls and had not realized that 1-800 numbers that are “free” on a conventional landline still consume cell phone minutes. Such price shocks can sometimes be difficult to collect on and leave the customer considering alternative providers.

May inhibit use: It is often said that one of the reasons that the Internet in its very early days took off more rapidly in the US than in Europe was because of the fixed pricing model for local phone calls that was prevalent in the US. Americans got more benefit from the emerging Internet because they were not afraid of the possibility of price shock if they stayed connected for long periods of time. If you are overly concerned about the bill you are racking up you may not get the full benefit from the service.

Weaker revenue guarantees: From the point of view of the provider the flip side to unpredictable cost is unpredictable revenues. In a recurring model with a stable churn revenue becomes very predictable. In a usage based model that predictability is reduced.

What does usage-billing look like?

The model that we inherit from the telecommunications world separates the problem into three pieces Metering, Mediation and Billing.

The Metering system collects the raw data of what actually happened. In the case of a telephony service this is Call Detail Records (CDRs). In the case of a SaaS website it might be number of pageviews or downloads, or it might be gigabytes of diskspace used.

The Mediation function is responsible for taking the raw data from the metering system and transforming it into a form appropriate for the billing system. Mediation may involve data transformation, filtering, aggregation, validation, reconciliation and smoothing.

The final stage is usage billing where the transformed data gets turned into invoices. There are two main ways that this can be done.

Price per unit or block: At the end of the billing period we charge the customer for units consumed, these could be minutes, or pageviews, or gigabytes of storage. Often we will charge for blocks of units consumed in whole or in part; blocks could be 100 minutes, 10,000 pageviews, 5 Gb.

Volume pricing: In this model pricing is set by usage band, e.g. 1 – 100 minutes, 101-500 minutes etc. At the end of the period we determine which band the user falls into and charge them as fixed price the associated price for the band.

Hybrid models

Hybrid models combine usage with a recurring subscription. This can offer the provider the best of both worlds, a recurring revenue stream and the appeal to the customer of being charged only for what they use.

In a hybrid model the recurring-billing period may be different from the usage-billing period. It is good practice to keep the usage billing period short to reduce the possibility for major billing shocks. On the other hand it is desirable to have the recurring period to be as long as possible to limit the customer opportunity for churn. So an annual recurring-billing period with a monthly usage-billing period is not uncommon.

In a hybrid model there is often an allocation or endowment of units that comes “for free” with the recurring charge and then a usage-based overage charge, that can be price per unit or volume pricing, that is added to the bill if the endowment is exceeded. The most familiar example is cell phone minutes where you will pay for a 500 minute plan at $40 per month and then its $0.5 for each additional minute.


In this article we noted the trend toward an increased interest in usage or activity based billing, which we believe will accelerate as the tools to support this kind of billing become more readily available. We looked at the Pros and Cons of a usage model and looked at the basic bones of what simple usage models look like. Finally we discussed Hybrid models that combine usage with recurring billing to provide the most flexible approach.