Last in a series. It would be unfair to identify "Microsoft software licensing: Seven deadly sins" without offering like number of remedies for them. To recap: Software Assurance is an upgrade addition to Microsoft volume licensing announced 10 years ago that generated stiff resistance from customers. The company removed any-time upgrade options, replacing them with annuity programs where customers annually pay 25 percent of a server license's cost, and 29 percent of a desktop (primarily Windows and Office) software license. Customers must choose whether or not to get Software Assurance when acquiring new licenses. Customers get the right to upgrade but not the assurance Microsoft will release new software during the contract period.

Seven Remedies

Software Assurance has many flaws, but Microsoft does need a program that rewards customer loyalty to its products by giving them discounted upgrades and that provides ongoing support comparable to what other sotware vendors offer large customers. Here's what I would do to mitigate some of its worst features.

1. Let customers purchase Software Assurance at any time, not just when they purchase a new license. By prohibiting that today, Microsoft reveals insecurity about its own products: "We can't count on people wanting the next version, so we'll force them to buy the pig before they know how sick it is."

Software Assurance could still be a two- or three-year contract, but at least customers would be guaranteed to get one upgrade if they can buy Software Assurance when they know Microsoft will release new software they value. Since everyone who purchases Software Assurance under today's kludgy program will probably continue to purchase it under better circumstances, and more customers would purchase it if they were guaranteed an upgrade, Microsoft would probably sell more SA. Even a small increase in new license purchases would make up for any loss in Software Assurance revenue due to other factors.

Since customers still pay somewhere between 75 percent and 87 percent of the full license price for an upgrade, even with Software Assurance, Microsoft isn't giving away a lot even if every customer does get an upgrade -- particularly since they should get an upgrade if they pay for one, which Software Assurance doesn't guarantee today.

Microsoft would also see much better adoption of new products, as a bet on Software Assurance becomes a sure winner. Customers would also be more likely to take advantage of some little-used SA benefits, like e-learning materials and TechNet, which make sense as self-help resources for deployment and training on new products they have just purchased. Those materials don't get much use among the good-enough crowd.

2. Reduce the Software Assurance term on renewal. To avoid having to renew SA for another three years to get an upgrade that might be only a few months away, SA renewals should be for only a year at a time. This poses some risks that customers won't renew. But that just means that Microsoft needs to keep performing -- producing compelling products in a timely fashion to earn customer loyalty. Locking customers in and making them overpay for products they don't want only pushes them away. Reducing the term also gives Microsoft more flexibility in adjusting prices. If the price of a product goes down or there's a change in the product lineup, Microsoft can adjust Software Assurance prices on succeeding bills to reflect the change and ensure that customers don't overpay.

3. Reduce the price. Outside of an Enterprise Agreement, Software Assurance "attach" to new licenses is very limited. Version upgrades used to cost between 40 percent and 65 percent of the full license price. That would reasonably put the cost of SA over three years at 15 percent to 20 percent a year, more in line with general industry standards.

4. Stop charging more for Software Assurance on the desktop than on servers. Software Assurance benefits on servers are better than those for desktops (one tech support incident for every $20,000 spent, rather than one for every $200,000 spent, for example), and yet SA on servers is less expensive (25 percent rather than 29 percent). That doesn't make sense. The 29 percent customers pay for Software Assurance on the Windows OS is particularly egregious:

  • Software Assurance on the Windows upgrade includes no tech support benefits.
  • Many companies don't do company-wide OS upgrades on existing computers because they require formatting the hard drive, reinstalling applications and personalizations -- in short, a massive amount of lost time and user productivity.
  • Every time customers purchase a new computer, they get the latest version of the OS -- so most of the time Software Assurance only gives them another license for a product that they already own, and since they can only use one (the OS license can't be transferred to another computer), the second one is a waste of money.

5. Reduce and simplify the benefits. With customers guaranteed an upgrade, Microsoft no longer needs to give customers a random bag of tchotchkes to make up for possible disappointment on upgrades. Fewer, more focused benefits that help customers get more out of their licenses are good. At a minimum, the minuscule Office tech support incident can be ditched. It makes Microsoft a laughing-stock and most customers spend more money keeping track of it than they save on tech support.

6. Get creative on Software Assurance terms, if letting customers buy it at any time is ruled out. Software Assurance costs should reflect risk. Someone signing up for SA in the 12 months after a product is released has the highest risk of disappointment. They should be paying 15 percent a year, not 29 percent, to reflect the risk they're taking. Someone adding Software Assurance to a new license three years after the release of the current product is much closer to the release of an upgrade and can pay more. It's a bit complicated, but any customer who can read a current Microsoft price sheet is qualified to figure this one out.

7. Offer a discount on Premier Support based on Software Assurance spending. One of Microsoft's Premier Support executives admitted to me a few years ago that SA was his number one headache. Why maintain parallel support programs? There is already a way for Software Assurance customers to convert their SA incidents into Premier Support incidence, and I always encourage customers to do the conversion. But why complicate the customer's inevitable journey to Premier Support that way? Microsoft could greatly simplify the process and add a lot of momentum to the Premier Support brand if it used Software Assurance as a shortcut to Premier Support instead of making it the long way around.