How to eliminate the complexity of software licensing with utility pricing.

Software licensing seems to grow more and more complex every year, and as software piracy mushrooms, software companies and watchdog groups are aggressively pursuing violators with fines and lawsuits. This presents a unique challenge for IT managers who already expend countless hours and dollars to ensure compliance with increasingly complex licensing schemes. Fortunately, solutions such as software utility pricing and data volume-based pricing may help solve the problem.

Current Software Licensing

IT managers must try to manage the increasingly complex software licensing landscape while simultaneously reducing their cost structure. In order to develop a successful licensing plan that reduces costs, improves service levels and eliminates non-compliance risks, IT must negotiate a tricky process:

Baseline Licenses. They must identify hundreds or thousands of software licenses (including various versions, upgrades and patches) from multiple sources for individuals, departments, business units, branches and worldwide locations.

Track Changes. They must understand the licensing consequences of hardware changes such as adding, eliminating and/or consolidating servers, clients and tape drives.

Prepare for Audits. They must prepare for inevitable software license audits from internal auditors, software vendors, the government, etc., and prove they are in compliance with all licenses.

Traditional Software Licensing Schemes

All too often IT deals with a confusing system of inconsistent licensing pricing schemes. It is not unusual for the same software package to be priced differently depending on the number of seats, the number of network servers, the types of server and the number of CPUs. Sometimes there are different prices for different operating systems and, in some instances, even for different "flavors" of the OS. As a result, licensing prices at some software companies run to hundreds of pages.

To help customers negotiate this licensing labyrinth, many large software vendors have resorted to so-called "license configurators" to guide customers (and, in many cases, their own salespeople) in understanding and selecting the correct software license.

Another option that has emerged in recent years is software asset management (SAM) software. SAM software helps companies keep track of software versions, upgrades and patches to ensure the company remains in compliance with all software agreements by pinpointing any misuse within the enterprise. While SAM features are being incorporated into existing software solutions, SAM is a new expense and, in itself, adds another layer of complexity. As great as SAM may seem on the surface, SAM software still falls short of the Holy Grail of true simplicity.

On the customer side, many companies seek to alleviate confusion by taking the path of least resistance: purchasing an enterprise license. While this ensures the company will remain in compliance, it will likely result in overpaying for licenses that are never used.

Another tactic is "gaming the system." Since there are so many caveats, exceptions and changes in licensing schemes, enterprises and their software vendors maneuver to extract the most value from each other--an exercise that does little to increase value or productivity.

Simplifying the Licensing Process

For many IT managers, simplifying complex licensing schemes is essential for maintaining their sanity. As a result, there is a trend towards a virtualized IT structure that allows companies to access software on-demand and pay only for what they actually use. Although this is not yet a reality, new schemes are emerging for aligning IT infrastructure costs with actual IT usage--a critical first step in simplifying the process. Two key models are utility pricing and data volume-based pricing.

Utility Pricing

The "utility" in utility pricing refers to the metaphor of a gas or electric company that turns on the power and charges companies only for the amount of energy they use. Because there is no need to overbuy hardware or software, a company enjoys maximum pricing flexibility. To date, most of the thought and effort regarding utility pricing schemes have concentrated on hardware for the simple reason that hardware lends itself more easily to the concept. Hardware costs are created the moment the metal is pressed or the sand is turned to silicon. Hardware revenue is realized when the physical unit is shipped to the customer. The focus of utility pricing for hardware has been on virtualizing the environment: moving to blade severs, pooling physical resources and building out grids.

Software, on the other hand, has virtually no marginal cost to a software vendor--these days vendors don't even have to burn CDs. However, because software revenue is realized only when it is licensed to the customer, the focus on software utility pricing must be on creating methods that reflect utility or usage.

For software to achieve a true pay-as-you-go model, it must overcome several major obstacles. First, utility pricing requires resources that may not be present in legacy systems. Second, and more importantly, the company must have an infrastructure agile enough to adjust as storage and processing needs change. Finally, the vendor community must buy into the concept as a way to simplify purchasing and improve customer relationships.

The best example of utility pricing is in the storage industry. Storage hardware is already purchased based on usage; when an enterprise needs to use more storage, it purchases new disks or tapes. Usage is normally determined by the amount of data that needs to be stored--a metric known as data volume.

Data Volume-based Pricing

With data volume-based pricing for software, IT no longer has to deal with the number of nodes or seats, with adding or reducing the number of clients or with tracking multiple versions of the same hardware. Data volume-based pricing is the first pay-as-you-go model that offers true licensing simplicity. To be successful, however, it must be:

* Easy to use. The mechanism for tracking data volume must be embedded in the software. Tracking it should be as simple as looking at a fuel gauge: a bar indicates if data capacity is full, empty or in between. Pricing must be equally simple, and the mechanism for expanding or contracting storage space must be direct. Although one could argue about organizational fit, the Sun Microsystems storage initiative, which charges $1.00 per gigabyte under management per month, is an excellent example of this approach.

* Easy to measure. In the storage industry the generally accepted metric for utility is data under management measured in gigabytes, terabytes or petabytes.

* Fair. Open source software-as-a-service and other emerging models demand licensing that is less complex and simultaneously more fairly priced for small to medium-sized businesses (SMBs). In backup software, for example, solutions must ensure that companies opting for data volume-based pricing do not pay for rapidly growing archived data.

* Fully featured. In the past, vendors have offered stripped-down versions of their software to make it affordable for customers to start and grow to the full-fledged version. However, if the software does not meet customer requirements or does not offer the essential features they need, it is usually doomed from the start.

Paradoxically, data-volume based pricing for storage software also benefits companies that sell hardware, e.g., OEMs, system builders, VARs, because it allows them to add software to their boxes with little or no upfront cost. Costs are incurred only as the software is actually used, enabling differentiated software solutions to be added to the hardware and empowering the end-user to choose how to maximize value of the machine.

Is IT ready for a new paradigm?

Although new software licensing models are emerging that promise to truly simplify the licensing process, some IT managers may be uncomfortable about changing their existing licensing policies. A legacy system may be complex, but it is familiar. Changing that system to a new model requires new procedures, new plans, and a new way of doing business.

Those who embrace simpler licensing models, however, undoubtedly will experience a strange feeling: one of their major headaches has vanished.

Dave Elliott is Director of Marketing and Alliances for Arkeia Corporation (Carlsbad, CA)

www.arkeia.com

MEASURE           TRADITIONAL LICENSING           NEW UTILITY LICENSING

Key Cost Drivers  Number of servers or CPUs       Use of software
                  Type of servers and OS version
                  Number of seats
                  Instances of software
KeyBenefits       Traditional, well accepted      Simple to manage
                    method                        Equitable to vendor
                                                    and customer
Key Issues        Growing complexity              Agreement on utility
                  Managing changes                  metric
                                                  Purchasing mindset
Success Criteria  Determine a baseline            Ease of use
                  Manage change                   Level of integration
                  Continual auditing and            of licensing
                    monitoring                    Fully featured
                                                    offering

Figure 1. Comparing traditional licensing models with utility pricing
models.