If IT is to be perceived as valuable, it has to have a price. Usage-based chargeback is the best way to build a price-to-value relationship for IT services, and it is one of the cornerstones for running IT as a business within a business. As long as IT has a solid understanding of its operating costs, it can use pricing as a strategic tool for improving alignment with the business by giving executives better understanding and control over IT resources. Different models, with different classes of service, can be used to drive more cost-efficient consumption of IT and to achieve more effective matching of service to business need. Four basic methods for pricing IT value are described below.
The simplest chargeback model, subscription pricing is a pay-per-use model in which pricing is per unit of time, which is much easier to monitor and measure than consumption-based pricing. The operational cost of the IT facilities is calculated and amortized across a subscription period (for example, one year) and then divided between all the users of the service. Depending on the operating profitability goals applied to the IT organization by the business as a whole, an element of gross margin may be added—perhaps to create a pool to fund IT research or future projects.
* Simple: If, for example, five lines of business were subscribing to a service that cost $60,000 per month to provide, the subscription charge (assuming a break-even business model) would be $60,000/5 = $12,000 per business unit per month.
* No usage monitoring or penalties: It assumes all parts of the business will use the service at the same level on a constant basis, with no penalties for excessive consumption or peak time usage.
* No cost justification: There aren’t metrics by which the actual level of consumption can be measured, calculated and justified to skeptical consumers.
The peak-level approach takes the subscription model and adds a mechanism to monitor and record peak consumption. Consumers are billed according to their peak use, not according to their average use.
* Simple to meter: Only peak-level usage needs to be monitored and recorded.
* Clear cost justification: Easy to show when consumers are using more than the base level resources.
* Penalizes variability: If there are just a few peaks of usage during a given period, the scheme can seem unfair. But shortening the analysis period—say from six months down to one—and the measurement intervals—from weekly to daily, for example—can solve the problem.
If user management is a bigger cost issue for IT than hardware usage, it makes more sense to meter IT by the person rather than the machine. If users are connected to their computers for fairly similar periods of time and have relatively well-understood transactional profiles—for example, bank customer service representatives who work on Web portals—this can be a fair and easy way to charge for usage.
* Easy to implement: Tracking the authentication of individual users to IT services is relatively simple, especially if a single sign-on system is in place.
* Clear cost justification: The authentication records provide the basis for cost justification.
* Ignores system load: If users make heavy demands on systems when they log on, this model shortchanges IT.
In IT environments where quality of service is critical, IT can meter and control usage very tightly using electronic “tickets” that use a short validity period (say four hours).
* Consumption regulation: Ticket-based pricing lets IT control system load to a fine degree, helping to eliminate usage peaks and ensure business continuity.
* Simple: All that is required to monitor ticket pricing is a low-latency (i.e., fast-responding) portal, most probably constructed as a Web service. Tickets provide permission to use the IT service multiple times during the ticket validity interval (a “multiple right of re-entry” solution).
* Strongest cost justification: Of all the models, ticket based pricing is the most powerful in terms of cost justification.
* Pinpoint monitoring: Tickets can be very specific, allowing both sides to monitor exact usage down to the specific application level.
* Ticket hoarding: For the ticket-based model to operate effectively, it’s often necessary to implement “use-by” dates on tickets to avoid stockpiling.
First Class or Coach?
There are other, more complex models for chargeback that bring even more depth to the monitoring and costs. But these four models provide a start. And they can be made more meaningful by layering a system of service levels (and varying costs per unit of service) on top of each model, similar to the airline fare class-pricing model. For example, network access could be offered under the ticket-based chargeback model at three price levels, each with varying degrees of bandwidth, service level guarantees and peak usage guarantees.
The goal in any of these chargeback scenarios is to deliver IT services in ways that present the highest degree of visible perceived benefit to consumers—just like in the real business world. CIOs, meanwhile get the opportunity to manage their own cost structures, away from the prying eyes of consumers.
Chargeback is a way to put IT services in terms that businesspeople understand and value. When IT is bought and consumed like other services, IT can become a business within the business. And that is the path to true IT value.
Mark J. Denne (email@example.com) is a partner with Accenture. To comment on this article go to the online version at www.cio.com/010107.