A recent Sandhill opinion piece pointed out how virtualization is disrupting database, middleware and, potentially, storage vendors. We believe the extent of the disruption will be felt even more broadly. And, based on our 20+ years of consulting on software pricing and licensing, we're convinced the future is bleak for B2B software vendors whose business models are based on hardware-based pricing metrics and perpetual licenses.

Hardware Licenses Harder to Estimate and Count
Vendors using hardware-based metrics like CPU-count or performance-based like MIPS or "power units", will be challenged by uncertainties over what the quantities of licenses are (or should be) and how to fairly track and monitor the software in use.

When hardware enclosures in a data center each contained one, single-core CPU, it was easy to estimate the number of licenses required. Counting "CPUs" meant counting "boxes" and it was easy for a customer to estimate how many licenses were needed and a sales rep to quote a price.

But estimating the number of licenses is more difficult for customers now that CPU architectures have grown more complex. Multiple CPUs and multiple cores per CPU within an enclosure are the rule and not the exception these days. This makes it harder for customers to estimate license requirements while making it harder for vendors to monitor licenses in use.

Trust Me Licensing Becomes Less Acceptable
The "trust-me" licensing that has been used for many years ("Trust me. We'll comply with your license agreement and won't use more than we pay for.") will become less acceptable. When companies were willing to pay and sales people got the money upfront, trust-me licenses were fine.

Today, with everyone focused on costs and revenues, vendors think customers are underpaying for licenses in use and customers think they are over-licensed and paying too much. In these economic times, both parties will be more interested what is actually being used or paid for. With multiple cores and especially virtualization, license monitoring and compliance auditing will become more difficult and expensive if it can be done at all.

With virtualization, one might even argue, the amount of hardware in use can't be counted and, aside from running the software, the hardware may be irrelevant since it is "in the cloud".

More Powerful Processors Affect Sales of Hardware-Based Licenses
As Moore's Law marches on and processor power increases, the cost of hardware continues to drop. This makes the stable or rising prices of software licenses look expensive relative to the declining cost of the hardware thereby putting pressure on price levels.

With more processing power available, fewer processors or cores are needed to run software that serves a given population of users. Even though application usage may be growing, the growth in new licenses sold will fall behind the growth in usage. This means development costs will be amortized over fewer licenses thereby affecting a vendor's return on investment.

Hardware-based licensing has existed since the beginning of computer time (i.e. the 60s and 70s) and may continue. But vendors doing business this way may do what the buggy whip manufacturers did - evolve into manufacturers of special purpose whips and riding crops. Like the manufacturers of whips, software vendors using hardware-based licensing may also find alternative, less savory sales outlets or customers.

Perpetual License Fees vs. Economic Value, Risk
Customers will continue to focus on IT spending even as hardware prices continue to fall and virtualization lets budgets stretch even further. Furthermore, in these uncertain(!) economic times, every customer wants to make their cash go further and avoid risk.

As customers looked for ways to meet their objectives while spending as little as possible, SaaS vendors attracted attention with their low price points and small entry-level configurations. This new competition for budgets made the price and the upfront payment associated with the perpetual license look unattractive.

A customer could spend less and still meet their objectives so the economic return of the SaaS alternative looked pretty good. Customers could also make their scarce cash reserves go further.

SaaS-licensed products and their recurring payment stream and lower entry prices also shortened the delay between cash outlays and benefits realized. The large, upfront payments associated with perpetually licensed software and the typical delay before realizing a commensurately large return made the economic returns of a perpetual license look more risky.

More Costly Sales - Even to Existing Customers
As prospects resist purchasing product arising from either a reluctance to spend a lot of money up front or the difficulties in estimating license requirements, sales cycles will stretch. Licensing that isn't aligned with customer business metrics and easier-to-buy alternatives like SaaS will only add to the delay.

If direct sales people respond to delays with more discounts, negotiations could further slow down the sales cycle as prospects stall and negotiate harder. Sales to prospective new customers may therefore generate less revenue at a higher cost of sales. The overall effect could be more deals lost, lower rates of success and discounts that might rise sharply.

Finding new sales prospects in an installed base won't be much easier either and could be more costly than in the past. Improvements in hardware performance will likely continue and erode the need for more hardware-based licenses.

In the face of scarce budgets, the maintenance and support fees associated with perpetual licenses will come under much closer scrutiny. For years, customers have been questioning the value of maintenance and support. Now these expenses will be even more visible, particularly if they are compared to the price of a SaaS offering that comes without an upfront license fee.

While this scrutiny may happen anyway, customers will be even more careful when they are unsure of their license needs and forced to lay out large payments.

Uncertain Future with Failure as an Option
Software companies using hardware-based metrics or perpetual licenses are in for difficult times. Companies whose business models are based on both will be in for really difficult times. In either case, they may be living on borrowed time.

Here's how we see the future shaping up for vendors whose B2B applications use hardware pricing metrics and/or the perpetual license model.

• Prospects will be harder to find and finding them will cost more
• License sales will stall as hardware performance continues to improve
• Estimating license requirements along with intense price negotiations will stretch sales cycles
• The number deals and the close rate will decline
• Sales transactions will be smaller and cost of sales will rise
• Perpetual license revenues will suffer due to SaaS-based competition
• Hardware-based metric and difficulties in estimating licensing needs will jeopardize maintenance and support revenues

This is not an attractive future. At best, these vendors may be able to delay their demise. But even if they do, they won't be able to avoid an untimely and, in our judgment, an unnecessarily messy end.

Difficult Mission But Not Mission Impossible
If you have either hardware-based or perpetual licenses or both and want to do something other than wring your hands, here's some advice.

Start by understanding what customers want in the way of a license model including metric, term and payments. Then change your business practices to more closely align with what they want. Some of the characteristics of a desirable license model include...

• Economic returns occur as close to cash outlay as possible
• Product packaging fits with usage patterns and available budgets
• License metric is perceived as fair and aligns with critical business metrics
• License requirements, and associated costs, are easily and quickly estimated without too much digging
• It is one thing to understand what customers want and another to act on it. But taking action is not so difficult if you go about it systematically and in stages.

What Should You Do Starting Today?

If you use perpetual licenses, the easiest place to take immediate action is there: Stop using these licenses. Offer shorter term-licenses (annual subscriptions) to new prospects instead. (You may also want to bundle support and maintenance with the subscription license fee.)

Don't be reluctant to offer short-term licenses with their recurring fees. Forget the old argument used against subscription payments - that they hurt cashflow and valuation.

First, you may attract enough new customers and do more sales transactions and actually improve cashflow. Chances are the upfront payments of the perpetual license model kept some prospects from becoming customers. Don't let that continue.

Second, software company valuations are in the tank, so this may be the ideal time to abandon the large upfront payments of the perpetual license. More cashflow will lead to improved valuations if that's important to you.

Tomorrow's To-Do List
As you attract new customers to the subscription license offering, think about repackaging your products. Can you create smaller, entry level packages without any added development cost or provide fewer services at even more attractive prices? Even if you have to shift to a different sales model to accommodate smaller sized sales transactions, these actions may still be worthwhile due to increased sales volume.

At some point, you need to think through how you might transition your installed base to the new shorter-term license model before actually offering these options to your installed base. Some customers have been known to convert a perpetual license into a subscription if current maintenance and support fees and future products are priced properly.

Finally, make sure your existing invoicing and billing systems can respond as customers and the product mix shifts to short-term licenses, smaller "chunks" of products and more frequent transactions. Billing systems that are based on low volumes (e.g. a big initial invoice, an annual invoice for maintenance and support, and occasional invoices for additional products) may break down.

But no matter how well you do any of the above, all they can do is buy you time - time you will need to develop a licensing metric that is not tied to hardware. You must find a license metric that is based on the value customers get from your application, instead of the hardware on which it runs. This is the only way to avoid all of the problems caused by hardware-based licensing metrics and the problems virtualization will bring.

To summarize, companies whose business model is based on a hardware-based pricing metric and perpetual licenses can expect even rougher-than-normal times ahead.

Find a metric that is tied to customer use and the value they capture from using the software. Eliminate the perpetual license model and its associated payment stream. The economics will become even more unattractive.

Like everything in business, success is not guaranteed. But if you change from a hardware-based metric and perpetual licensing, as we suggest, you might not get rich in these tough times but you are unlikely to go broke or out of business.

Find out more at the upcoming Software and SaaS Pricing Workshops on March 11 before OpSource's SaaS Summit and on April 16, after Softletter's SaaS University. For more information see www.softwarepricing.com/Workshop.cfm.


Jim Geisman is principal and founder of Software Pricing Partners (formerly known as MarketShare). He is an acknowledged expert in software licensing, packaging and pricing..