Software is Moving to a Consumption Model

When I bought my first mobile phone – a giant brick of a phone from Motorola – I had to pay for my service by the minute. On the one hand it seemed fair to pay for only the minutes I used; On the other hand it led to some surprises when the bill arrived. This was consumption pricing at its purest. I used minutes, I paid for them.

Over time, mobile carriers have steadily moved away from per minute, consumption pricing to subscriptions that give unlimited or large fixed amounts of services. This has coincided with the addition of new services to consume including SMS messaging and data connections for internet access. Consumers seem to like this model better since it yields fewer surprises when the bill arrives. It eliminates the problem of teenagers spending $400 on tens of thousands of text messages (which used to happen just a few years ago). The downside is paying for more than you use or need. My family, for instance, uses only a fraction of the data plan, despite financial incentives to use more data.

Software, especially software for the SMB market is heading in the opposite direction. In the past, the prevalent option available for software was perpetual licenses. Whether they were individual licenses or site licenses, it was “all you could eat” forever. Maintenance fees were something else entirely since they were designed to pay for continued support and upgrades.

Unlike mobile carriers, software companies are moving away from perpetual licenses to subscriptions and, I believe, a consumption model. The software consumption model will have customers paying monthly fees only for the licenses used each month. In time, we may see software companies only charge a customer a fee when the customer actually uses an application not just on a monthly basis.

There are a number of reasons for this shift away from perpetual licenses, at least from a customer’s perspective. With perpetual licenses, customers pay for licenses they don’t need. That may because of bundling (paying for bundled applciations even though they are rarely used or used by few end-users) or because of changes in a workforce. This is an especially important problem for companies with seasonal or fluctuating workforces. When I worked in the business process outsourcing field we would often see sizable swings in our workforce either because of holidays or changes in our customers’ need for our services. Often perpetual licenses are non-transferrable, hence they cannot be resold like a piece of equipment. Working capital could easily get tied up in software purchases that lay fallow after a change in the workforce. In seasonal businesses this means buying license that you only need sometime, often for a few months a year.

Cloud services, especially SaaS, changes the equation. By hosting the application with the vendor, changing licenses is a matter of turning on and off access. This makes it practical to sell subscriptions that could be turned on or off at any time in the month. Monthly subscriptions, in turn, enables companies to only buy the subscriptions they need. If my workforce might fluctuate either with business or seasonally, I can make the choice to spend a little more per license but not have to pay for the entire year. The ability to pay-as-you-go fits in well with industries that have flexible staffing models.

The obvious next stage for software pricing is a consumption model that bills you automatically only for the license that you consume in a month. This is the ultimate in flexible spending. Each month a company receives a bill for the applications that companies have actually used. If your workforce shrinks, you stop paying for license that no one can use. Infrastructure as a Service is already doing this, charging for network bandwidth, storage, and processor time used by applications.

This may make it harder for software companies in some ways. Billing by consumption can get complicated but cloud applications already track when software is used and when it is not. Besides, telecommunications companies have done in for years. A keen advantage to this model is that it will become readily apparent which products are used, and hence worth putting resources into and which are not. Bundled or multi-faceted products can hide that information. The bundle is successful even if individual components are not. It will also allow companies to build temporarily personalized bundles based on what brings value at this moment in time.

The biggest reason to do move to consumption pricing is that the idea is out there already and customers will want it. Customers want only to pay for what they use if for no other reason than it will seem more fair. That and it may save them money.