The basics of lowering software costs are simple. Every company, regardless of size or industry, can achieve superior results by following the example of best-in-class companies. So, how do some companies typically spend less than half of what others spend on software costs? They work on managing product inventory and usage, obtaining the best available software pricing, and don’t buy into the myths of Software Asset Management (SAM).
Consider the origins of IT Asset Management and, more narrowly, SAM. The need for it emerged when changes in mainframe hardware technology, software licensing, and vendor consolidation left many companies trying to figure out how to satisfy increasing business demands despite a decreasing budget.
In a 1996 keynote speech titled “Orchestrating Order From Chaos,” Gartner research director Joe Pucciarelli defined the first iterations of SAM as “a new and evolving system of managerial goals, strategies, and implementation tactics.” Hope appeared on the horizon when IBM announced Parallel Sysplex License Charges (PSLC) pricing on April 6, 1994. Yet, the following year, many grappled to understand the implications of Computer Associates’ purchase of LEGENT Software. Nevertheless, just six years later, software pricing had become so complex that the top five Independent Software Vendors (ISVs) had more than 250 unique pricing tiers. Bill Snyder, in a 2002 Meta report titled “The Shifting Sands of Mainframe Software,” predicted that this confluence of issues would drive OS/390-z/OS ISVs to reduce overall costs. The problem was simple: perceived diminishing value on mainframe software. Costs rose while many software features remained the same.
Today, vendors offering solutions to perceived SAM problems haven’t resolved the problems. In fact, customers face the same issues as before. They aren’t spending less on software and a large chasm between vendors and customers remains. On both sides of the fence, finger-pointing reigns. Amid the chaos, users and vendors have devised their own strategies to reduce software costs, with few proven results. Jobs have been created to follow SAM practices, while others have been lost when results fell short of SAM promises. This has happened not because of a lack of effort, but because companies bought into the “eight myths of SAM”:
Myth 1: Low software costs depend on data center size.
Reality: Although larger data centers often have the muscle and reputation to negotiate steep prices cuts, discounts aren’t what make a company best-in-class. Often, these companies don’t have the ability to manage hundreds of products running on dozens of large LPARs. Many large data centers struggle to get a handle on their product inventory usage and must rely on discounts to manage their software costs. Conversely, small data centers are often more efficient in managing their costs and tend to have an easier time understanding their inventory, usage, and product value. Our research indicates that the average large data center (one with more than 5,000 MIPS) has more than 300 products, compared to 92 products and data centers smaller than 500 MIPs, so it’s not surprising that smaller data centers manage their software costs better. The relationship between large data centers and low software costs is statistically insignificant.
Myth 2: Low software costs depend on having good negotiators.
Reality: Some of the best negotiators have the worst cost structures. They may negotiate 50 to 90 percent discounts on artificially high-cost scenarios. What a company may perceive as a good negotiator can, and often does, work against them. Many vendors want to build relationships and offer discounts based on relationships. When a company’s approach is to try to get the biggest discount possible, a vendor will often stiffen in its position and may even raise the base price offered to the customer and then offer a discount, only getting the customer back to a “negotiated” list price. No one benefits from creating a hostile negotiating environment. In reality, however, a well-managed data center can pay full list price and still end up with a lower-cost data center.
Myth 3: Low software costs come from good vendor discounts.
Reality: A discount is relative to some cost structure using a commonly accepted set of assumptions, including hardware configuration, software configuration, and pricing metrics. Any change in these variables can dramatically alter the baseline cost the vendor uses to determine a discount. Although a discount may look good as presented by the vendor, it can be above list price based on actual usage. Discounts are often determined by a given region. Some territories commonly give (and know how to get) substantial discounts, while other territories are known to offer their customers “discounts” that are actually above list price. Our research indicates that more than 80 percent of savings opportunities are directly attributed to managing product inventory and configuration and have nothing to do with discounts.
Myth 4: Low software costs come from a large number of Enterprise License Agreements (ELAs) with best-in-class terms and conditions.
Reality: Although ELAs often helped data centers during the growth of the ’80s and ’90s, they often hurt companies during slower years by generating costs associated with unused inventory and capacity. A common problem associated with ELAs negotiated prior to 2001 was that data centers were growing 20 to 35 percent annually, and thus wanted large capacity amounts embedded in their ELAs. Since 2001, data center growth has slowed dramatically, leaving many companies with excess capacity. As old ELAs were renewed, they were negotiated with similar product configurations. Companies often don’t allow adequate time to review their inventory mix and configuration to understand the current value of their data center’s core products. It’s not uncommon for an ELA to have a “waste factor” of more than 50 percent to pay for products and capacity not essential to the data center. ELAs are commonly sold for a great discount, only for customers to realize that if they were to purchase the products that were providing the highest usage (more than 90 percent), and limited the use of the products they keep to where they were being used more than 90 percent, they could reduce costs by more than 50 percent. ELAs are padded with bundled-in products and discounts in much the same way that a 20-course meal might be offered to someone on a diet. The reality is that best-in-class companies have fewer ELAs and buy only what they need, when they need it.
Myth 5: Low software costs come from using best-in-class processes.
Reality: Good processes are essential, but software vendors don’t reduce your invoice because you have a good process. In a study conducted on a random sampling of more than 100 companies, there was found to be no correlation between processes and costs. Companies known for their best-in-class business processes may excel with their processes, but they frequently fail to achieve the same results when applying them to SAM. SAM processes are designed to support financial objectives, which are seldom understood before the processes are implemented. Business processes are typically transactional in nature and assist in reducing labor costs associated with change management, data collection, inventory management, invoicing, etc. Some of the lowest cost data centers lack any formal process in managing their software. Conversely, some of the best processes deliver savings, but fail to address core business problems in managing software costs.
Myth 6: ELA discounts and “terms and conditions” are restricted by SOP 97-2, SOP 98-4, and SOP 98-9.
Reality: The American Institute of Certified Public Accountants (ICPA), Financial Accounting Standards Board (FASB), Generally Accepted Accounting Principles (GAAP), and the Securities and Exchange Commission (SEC) don’t tell vendors how to price software; they only govern how to recognize the revenue. There are no accounting rules that restrict a vendor’s ability to negotiate price and contract terms. Discounting, in particular, isn’t addressed in any way in the accounting pronouncements and therefore can’t be used by a vendor as an excuse. License agreement terms and conditions can certainly impact the vendor’s timing of revenue recognition, and this is a real issue for the vendor (and its shareholders). SOP 97-2 was issued Oct. 27, 1997, and together with subsequent clarifying pronounce ments SOP 98-4 and SOP 98-9, governs the revenue recognition for licensing, selling, leasing, or otherwise marketing computer software. In its most basic form, 97-2 states that revenue can be recognized when the following four conditions are met:
- Persuasive evidence of an arrangement exists
- Delivery has occurred
- The vendor’s fee is fixed and determinable
- Collection is probable.
While the SOP goes into great detail regarding all four conditions, none of these provisions limit a vendor’s ability to discount. However, it’s much more difficult for a vendor to refund or waive fees that were set in a contract because SOP 97-2 calls into question the probability of collection of future contract fees. Other terms, such as cancellation provisions or refund rights, can also impact revenue recognition. Vendors shouldn’t hide behind these rules in failing to grant a customer favorable terms and pricing. The reality is that the accounting rules govern revenue recognition, and not the business practice.
Myth 7: IBM’s Workload License Charge (WLC) pricing will obtain the lowest possible IBM software costs.
Reality: When IBM announced WLC in October 2000, many viewed it as the software savior for the mainframe industry. The idea that a company could pay only for the capacity where products are running wasn’t really new; major ISVs had actually applied it for years. But IBM was offering the public what it had asked for. While it is a good concept in theory, it doesn’t offer what many think it does. When an IBM representative or outside consultant offers to reduce a data center’s IBM software costs by converting to WLC, many are quick to jump on board without executing a full financial analysis of all of IBM’s pricing options. Although it’s true that some products decrease in cost with WLC, others actually increase. The challenge and responsibility of the SAM team is to analyze the thousands of pricing configurations offered by IBM to see if WLC is the lowest-cost alternative. That may seem a daunting task, considering the December 2004 IBM price book listed 124,478 line items of software prices (representing a growth of more than 5,000 pricing line items in one year). It becomes challenging to find a software pricing configuration option that offers the lowest cost possible, thus making it easy to settle for WLC. Yet, for some companies, WLC will actually cause the total software costs to increase compared to current pricing options. The reality is that WLC may be advantageous for some companies and reduce their costs, but for the majority, it’s rarely the best solution.
Myth 8: Best-in-class companies save the most money.
Reality: Measuring best-in-class is often elusive in the absence of any formal metrics. SAM managers who have spent years diligently managing their software costs are often surprised to find out that they aren’t the only ones who have worked harder and smarter. It’s often the quiet SAM manager, unrecognized by management, who is doing the best job controlling costs. We often dream of following in the footsteps of the individual or company that is admired for all the savings realized. They boast of the millions of dollars they’ve beat out of software vendors. Yet, if they were truly best-in-class, how could there be so much money to be saved? Improvement in savings is important and impressive, but not necessarily indicative of elite status. Once you reach elite status, it becomes more difficult to obtain incremental improvement. That may be one reason that most best-in-class companies are rarely seen or heard of in SAM circles.
Although there may be a proper role for SAM, traditional SAM methods are not a panacea for lowering software costs. Best-in-class companies spend less than half of the average data center, their SAM managers are rarely seen or heard, and they manage product inventory and usage, obtain the best available software pricing, and avoid the eight myths of SAM.