Organisations' preferred approach to software licensing falls into three fairly evenly divided categories, although it varies according to the type of application being licensed.
By submitting your personal information, you agree that TechTarget and its partners may contact you regarding relevant content, products and special offers.
In a survey of more than 300 IT and business executives conducted by AMR Research 23% of respondents advocated the highly variable "usage-based" or "on-demand" approach and 26% wanted "site licensing", where they pay a flat fee, regardless of how many workers use the product or how much they use it. Another large segment (20%) preferred a middle road involving a one-off licence fee based on the number of users.
The research found some correlation between licensing preference and the stability of the business. Organisations that are highly cyclical or growing rapidly tend to favour licensing that is tied to usage, and more stable companies prefer site licences, where there are no surprises.
In general, chief information officers prefer less variable licensing approaches because it makes planning and budgeting easier. They dread the idea of having to go to the chief financial officer in the middle of the year for more money because some business activity has triggered a software usage metric.
By contrast, departmental or divisional user executives prefer the idea of only paying for what they use and are typically unsympathetic about the IT department's budgeting problems.
The concept of paying for application software based on actual usage is more than 30 years old, but it was largely replaced by perpetual licences after computer "time-sharing" died out in the 1980s. Recently, large infrastructure suppliers such as IBM and Oracle have thrown their marketing muscle behind the concept, and Salesforce.com has signed up nearly 250,000 subscribers for its customer relationship management products.
Now widely referred to as on-demand computing, this is an approach where the user owns no rights to the software, but simply pays a monthly or annual fee based on a formula that is usually related to the number of people who use the application and how much of it they use.
Although there has been a lot of press coverage about on-demand computing, few major applications are offered this way. Many people confuse "hosting" with on-demand, but in most hosting situations the user must buy software licences before arranging for the supplier or a third party to run the applications for them.
In the AMR survey, 9% of respondents were paying for some applications on an on-demand or usage basis, although it represented less than 3% of revenue for the enterprise application market.
However, when users were questioned about how they wanted to purchase applications in the future, a different pattern emerged. Many companies that are currently licensing applications based on user seats believe that in future they will be paying for applications based on usage.
The fact that 23% of executives polled indicated that their preferred method of buying enterprise applications was on-demand is a wake-up call for software suppliers, who have mostly been taking a wait-and-see approach.
The preference for on-demand varies considerably by type of application, but in each segment the percentage of companies that want to switch from user-based licensing to on-demand has increased significantly. Even in enterprise resource planning, where perpetual licensing based on user seats has been the accepted standard for more than 20 years, 14% of executives surveyed said they would like to switch to on-demand.
Application software providers have been flirting with schemes that would allow users to pay for their applications based on actual usage, but few are ready to offer a full on-demand option. Even the high-profile success of Salesforce.com has not convinced leading providers to move away from perpetual or term licensing.
Part of the problem is that the software industry has grown up with a business model based on collecting large licence fees at the outset of a contract and then charging 15% to 20% per year for maintenance and support. If a supplier suddenly switched to usage-based fees it could see a dramatic fall in revenue. In the long run these companies might generate the same total revenue per customer with much better predictability, but that initial revenue plummet frightens them.
Suppliers also struggle with the problem of compensating and motivating their sales reps in an on-demand environment. The typical enterprise application salesperson is used to getting huge commissions from a small number of large licence transactions. Gradual monthly payments based on a customer's usage fees just do not excite them.
User companies expressing an interest in usage-based pricing will probably not be sufficient to get the major application providers to enthusiastically adopt a whole new model. Smaller suppliers or new entrants are likely to offer on-demand pricing as a way to differentiate their products and lower the barrier to entry by eliminating the large, up-front expense for buyers.
In the CRM market, for instance, the rapid growth of Salesforce.com, which offers only on-demand pricing, has forced even a well-established player like Siebel to introduce an on-demand product.
In the enterprise resource planning and supply chain markets, none of the leading suppliers seems to be ready to take the plunge yet. Oracle has a programme it calls Oracle On Demand, but it is really a hosting service where customers still have to buy either perpetual or term licences. Many suppliers are using value pricing approaches where they tie licence fees to variables other than the number of users, but these still require the customer to buy licences and are not truly on-demand.
Jim Shepherd is vice-president at AMR Research