META Report: Avoid Paying Extra for Licensing Fees

News Item: Big 5 consulting firms are reporting that many of their recent contracts are to complete installations for software that companies licensed months ago or expand installations already in place, rather than begin integration of new software licenses. Therefore, though the integration business is doing well, this is not an indication that software sales are strengthening.

Situation Analysis: Users can stretch their IT budgets much farther if they refrain from the common practice of buying large numbers of extra licenses at the start of projects, long before those licenses are needed for an eventual rollout. Although licensing fees amount to only 25%-40% of total project costs, these fees can have an important impact on ROI, particularly when bought months ahead of actual need. Furthermore, adding extra, unneeded licenses can increase maintenance costs. However, this is a false economy. In most software areas outside the desktop, competition is strong enough for users to negotiate favorable unit costs at the time of the initial acquisition on adding extra licenses – but they do not have to lock in those costs by buying the licenses a year ahead of time (with application lock-in, once a solution is picked, buying extra licenses without contractually fixed prices can be expensive). This is particularly true in the application server arena, where vendors are delighted to have new customers develop applications on their platforms, knowing this will mean large sales of add-on licenses when those applications are ready for rollout.

We recommend that users leverage asset management disciplines (e.g., contract negotiation, five-year financial analysis) to prevent turning these “discounts” into premiums. Best-practice organizations utilize a conservative, internally developed rollout plan with a net present value analysis (including maintenance, support, and upgrade fees). This analysis is then compared to the vendor proposal to understand the “true” discount, which is often nonexistent.

Waiting is particularly advantageous in markets with falling prices, such as enterprise application integration, where users may actually get a better deal by delaying purchasing licenses. It will be less advantageous in markets where prices are stable, such as enterprise resource planning.

Buying ahead is often a symptom of a dysfunctional budget strategy. In too many organizations, financial officers regard a positive year-end budget balance as a sign that the department in question received too large a budget in the present fiscal year – and an invitation to slash that group’s budget for the coming year. This encourages executives to use up their budgets, often with premature or unnecessary expenses, to preserve their budget levels for the next year – which is the opposite to the behavior that financial officers should be trying to instill. Therefore, some IT groups may choose to use a budget surplus by buying extra licenses at the end of the year with the expectation that it will need them sometime in the future.

When they do roll out new applications, users should seriously seek site rather than individual licenses for applications that will have high densities of users. The point at which it becomes more cost-effective to purchase a site license depends on the pricing of the individual vendor, and users should establish a not-to-exceed price at the time of the initial acquisition. When weighing these alternatives, users should consider the cost of managing individual licenses, as well as the price of multiple individual licenses, rather than a single site license. A major advantage of a site license is that the IT organization can be sure any amount of use at that site is licensed.

Changes in licensing practices (started by Microsoft and followed by other vendors) will force users into a portfolio approach to software deployment, taking into account an assessment of qualified end-user and business requirements, IT deployment capabilities, and budget resources. This approach will enable users to predetermine when they will play (buy products) and when they will pass (skip product versions and pay to get current at a later stage).

User Action: Users should be careful about buying large numbers of extra individual licenses when purchasing new software for application development. If they are concerned about the price of expanding their licenses once the application is ready to roll out, they should negotiate the price for additional licenses upfront, during the initial purchase, but they should not actually buy those extra licenses until they are needed.

Other Recent META Reports

Novell Faces a Critical Year

The Nirvana IT Organization

Value-Based Collaboration Strategies

Protecting Internet Assets

When following this strategy, users should be careful to negotiate the license price for new versions as well as the current version. When bringing out a new version of a product, vendors have been known to claim that the negotiated license price applied only to the original version.

An even better solution is to move to alternative licensing models. A license-on-demand model is often an excellent choice during development and initial rollout. Once an application is in general use, IT organizations should consider moving to a site license rather than maintaining very large numbers of individual licenses.

META Group
analysts Dale Kutnick, David Cearley, Val Sribar, Jack Gold, Daniel Sholler, Ashim Pal, David Folger, William Snyder, and William Zachmann contributed to this article.