Subscription Based Biz Model In Software - 4th May 06 - Dhj

Is Subscription Software Pricing The Best Option? Yes: CIOs can lock in predictable costs, upgrades, and better service with subscription software licenses by Mike Conlon, University of Florida November 2004, Issue 37 When CIOs license software on a subscription basis, they get predictable, low costs; the latest features; a strong bargaining position with vendors; and good vendor service. At the same time, companies can avoid a "sunk-cost" mentality and the risks often associated with deferring maintenance when purchasing software. Subscription-based licensing wraps acquisition and maintenance into one predictable bill, and CIOs can negotiate contracts to cover multiple years, providing price protection. Though a subscription agreement includes financing charges, it also grants the right to deploy software upgrades. In contrast, when a company buys and maintains software, there are usually version jumps not covered in the maintenance agreement. A single upgrade obtained through a subscription will typically offset the finance charges. Consider, too, that subscriptions let CIOs lower total cost of ownership by avoiding security and operational problems inherent in operating older software. By including all upgrades and patches in a subscription, CIOs gain access to new features. This is critical when supporting complex, rapidly changing environments. In a subscription scenario, all of the vendor's developers also produce new features for the customer's business. As with a maintenance agreement, you can terminate your subscription. However, a termination, unlike a maintenance agreement, means a large cash loss to the vendor, and the loss of a customer. Vendors know that winning business back after a subscription is canceled is unlikely, so they tend to work hard to retain their customers. CIOs can take advantage of this strong bargaining position in negotiating the terms of their subscription agreements and ensuring improved levels of service. Of course, software gets old fast. When businesses purchase software, they can easily fall into two traps: A sunk-cost mentality: After companies spend lots of money to buy software, they believe they need to keep it and make it work. Software licensing lets them avoid feeling locked in and gives them the flexibility to move forward if the software isn't a good fit. The deferred-maintenance argument: "We ran our old system for years and patched it ourselves, so why should we pay someone to introduce new bugs through patches?" companies say. Networked systems, security concerns, and software complexity require constant upgrades to installed software. When companies buy software, they can be lured into false complacency regarding maintenance. A purchase implies that the software is an asset that can be depreciated. However, unlike a physical asset that loses value predictably, software may lose value precipitously based on environment changes. Its future value is often unknown.

By licensing software by subscription from a good vendor, companies can ensure the software's future viability. The subscription model forces the vendor to recognize that spending operational dollars on its clients' software subscriptions must provide value now as well as in the future. Mike Conlon is director of data infrastructure and research associate professor of biostatistics at the University of Florida. No: Subscription software licenses often prove to be more expensive and provide less value than perpetual licenses by Todd Radabaugh, Sky Financial Group November 2004, Issue 37 Like low-carb diets, subscription software licenses are all the rage. "Step right up and get your software," vendors say. "No money down!" CIOs should think twice before signing that contract. In the end, you may pay more for less. Renting nonessential software has advantages, but for critical applications, a perpetual license is often better from the buyer's perspective. Many vendors push subscription pricing because it evens out their revenue flow. My company recently purchased an automated Web-application testing solution. Our quality-assurance (QA) group planned to use the software every day to perform rigorous testing prior to new-application deployment and upgrades. We invited three vendors to submit proposals. Two steered us toward subscription licenses; the third offered a choice between subscription and perpetual licensing. The subscription-pricing proposals boiled down to a per-day charge. The perpetualpricing proposal meant "you own it." We were skeptical about subscriptions because our QA manager noted that a pay-as-you-go model might, ironically, become a disincentive to test. A three-year, total-cost-of-ownership (TCO) analysis showed that a subscription license would cost two or three times more in the end. Fortunately, the vendor offering the perpetual option had also won our technical evaluation. One common argument in favor of subscription licenses is that if things don't go well, you can stop paying the vendor. I don't buy it. First, if you do a technical evaluation, check references, and stick to reputable vendors, you should expect that things will go well before you buy. Paying double or triple the lifetime cost to give yourself an out is a waste of time and money. Another argument is that subscription pricing lowers your up-front costs. This view is shortsighted. Our analysis showed that the lifetime costs could be much higher. Plus, you probably manage your expense budget, not your company's cash flow. So, if your company capitalizes big software purchases and charges depreciation expense back to your department, you're better off with a perpetual license and a lower annual depreciation expense—not a larger annual subscription expense.

My advice is to first consider your use of the product. If it's a critical application that you'll use often, a perpetual license likely provides better value than a subscription license. Think of it this way: If you play golf every weekend, it's best to buy your own clubs. But if you hit the links only once a year, rent instead. Second, examine TCO. The total cost to us of the subscription-license proposals was roughly equivalent to the vendors' list prices. This isn't Dec. 31, 1999; no one pays list price. Third, don't let vendors steer you toward a scenario that fits their needs and not yours. Make it clear that license terms are an essential part of the evaluation, not an afterthought. Insist on flexibility in license terms and develop partnerships with your key software vendors that benefit everyone. Todd Radabaugh is VP of application support at Sky Financial Group, a $14.3 billion diversified financial holding company based in Bowling Green, Ohio.

Subscription Software Is Sweet By Ben McClure April 5, 2004 We heard it back in the 1990s: The software industry will move from selling one-off licenses to subscription licenses. Now, with software sales down, we're hearing the same mantra again in 2004, and Fools should listen up. The reasoning is simple: In a maturing industry, selling licenses through a subscription model bolsters vendors' revenues. Paying a monthly fee to use software is cheaper and less risky than buying software outright and hiring staff to support it. So, in harder times it's easier for corporate customers to swallow the subscription model. At the same time, vendors spend less on high-cost salespeople and invest more in serving customers with call centers and technical support. Those higher revenues and lower costs go straight to the bottom line. Of course, investors will like the smoother, more predictable earnings profile that the model offers. The customer pays for licenses over time, and the software vendor recognizes revenues over the life of the contract. Software is no longer akin to a big piece of capital equipment, purchased at a high cost for long-term use. Instead, it becomes like a basic service, used and paid for every day. The subscription model levels out peaks and valleys in income and cash flow that hit software valuations. Just look at its impact on Wind River Systems (Nasdaq: WIND). For the maker of software systems for communications, automotive, and other devices, subscriptions are a saving grace. Wind River's total revenue fell by almost 9% in the fourth quarter. But subscription revenues jumped from $1.3 million to $6.8 million, keeping Wind River Systems from slipping into the red for the quarter. For companies that sell products based on "open source" software, subscription-based support is the natural way to go because those companies cannot depend on proprietary licensing for revenue and must rely on support services. Linux open source specialist Red Hat's (Nasdaq: RHAT) fourth-quarter revenue growth came mostly by way of subscription sales, which jumped 79% over the prior year's fourth quarter. Sun Microsystems (Nasdaq: SUNW), struggling to keep revenues and earnings up, has overhauled its software strategy for more subscriptions. Sun says it will charge customers a per-user fee for some of its software, eliminating complex pricing that applied to hundreds of different products. Another stock benefiting from subscription sales is Synopsys (Nasdaq: SNPS). The company makes software that enables microchip makers to design complex integrated circuits. Now into its third year of subscription sales, Synopsys has lowered its prices by 15%. More than 80% of the company's bookings now come from subscriptions. With the

chip business finally coming out of a painful cyclical slump, Synopsys is well-positioned to squeeze more profits via its well-formed subscription sales model. Market research group IDC projects that about half of software vendors will move from a one-off license payment system to a subscription-based model in the next 12 months. But don't get too carried away. For the time being, the old model will probably remain the norm. Heavy hitters like Microsoft (Nasdaq: MSFT) and Oracle (Nasdaq: ORCL), while they pay lip service to the subscription model, have big interests in maintaining the status quo. Discuss the subscription software model and its benefits with other Fools on the Red Hat discussion board. Fool contributor Ben McClure hails from the Great White North. He doesn't own any shares of companies mentioned here.

Software as a service' model a threat to traditional vendors; Applications that are accessed via the Internet carry lower upfront costs than licensed software and allow for frequent upgrades. Sector heavyweights have taken notice.(The Insider) Workforce Management; 1/16/2006; Frauenheim, Ed

Byline: Ed Frauenheim The phrase "enterprise software'' leaves a sour taste in the mouths of many business leaders. For years, business applications for workforce management and other tasks typically have meant costly upfront licenses, lengthy installations by pricey consultants and ongoing maintenance issues. But there's a newer flavor of software delivery that promises customers a much sweeter experience. So-called "software as a service'' involves companies accessing standard business applications over the Internet. Advocates tout this method as having lower upfront costs and fewer hassles related to managing computer systems. Another benefit, backers say, is rapid changes in features. Frequent upgrades allow businesses to tackle new challenges-say, a tightening talent market-with greater agility, notes Kathy Barton, senior vice president of marketing and product management at Raleigh, North Carolina-based Peopleclick. Barton, whose firm offers recruitment software over the Web, says the key is that a single software code base can be used by many customers. Up to now, she says, organizations typically have run their own copy of the software and even customized

that code, meaning lengthy testing might be required before installing an upgrade. "We can add changes to the system and make those available to our clients much more quickly than you could with the license model,'' Barton says. Software as a service, also referred to as "on-demand'' software, is growing in popularity and recently made headlines when tech icon Bill Gates acknowledged its importance. In the world of workforce management, on-demand software poses a threat to dominant traditional vendors Oracle and SAP. But those big guns are developing strategies involving software as a service, and the smaller vendors touting the approach face some hurdles. Among them: the relative lack of flexibility in software that is rented to many clients at once and a recognition that poor customer service can ruin a vendor in short order. Licensed software over? For years, organizations have installed most of the business management software they use on their own computers. They typically pay software vendors for a license to use the program, face a one-time cost for installing the software and write another check for yearly fees to maintain it, which can include fixing bugs that crop up. Altogether, large companies can spend millions of dollars to run HR and other applications in this way. In many cases, however, these massive projects have failed to provide expected results. Stories abound of troubles such as systems that fail to work properly and cost overruns. With the widespread emergence of the Internet about a decade ago, another way to deliver software became possible: The code and the computers that run it could reside elsewhere, with businesses using the Web to input data such as the health benefit choices an employee makes during open enrollment. Overall, the adoption of software provided over the Web had been underwhelming until the past few years. Things began to change with the success of companies like Salesforce.com, which offers a product for tracking clients, and Google, which has introduced Web-based services such as e-mail and instant messaging and has emerged as a rival to old-guard software giant Microsoft. In early November, Microsoft chairman Bill Gates effectively endorsed the shift to software as a service by previewing two Internet-based services, including Office Live, which aims to help small companies do business online. Roughly a third of the $3 billion annual market in so-called human capital management software is currently delivered over the Web, estimates Paul Hamerman, an analyst at research firm Forrester. That ratio could rise to 50 percent by 2010, he predicts. "It's clearly growing,'' Hamerman says. "It's become a proven, mainstream model.'' On-demand software has evolved over time. In the late 1990s, it was common for a vendor to offer to host a remote copy of an application dedicated to one client-a practice

known as the "application service provider'' model. Peopleclick, for example, offered this kind of recruiting software from 1997 to 2003. That's when it switched to what's generally called software as a service-meaning a single instance of the code runs for more than one customer. This feature, dubbed "multitenancy,'' also is at the heart of software offered by San Mateo, California-based SuccessFactors, which provides a range of workforce management applications over the Web. Multitenancy is the reason SuccessFactors can upgrade its software each month, says Rob Bernshteyn, the company's senior director of product marketing. He likens the frequent enhancements to the way Google or Yahoo regularly offers vast numbers of users new options, such as making phone calls over the Internet. Customers "can take the upgrade or not,'' he says. Client Quintiles Transnational likes the way SuccessFactors upgrades its software so often, using customer input. Quintiles-which is based in Durham, North Carolina, and offers clinical research, sales and other services to pharmaceutical and health care companies-signed a contract with Success Factors in 2004 for performance-management and goal-setting software. To meet European data privacy requirements, Quintiles needed an option to limit the ability of managers to view the personnel data of lowerlevel employees. SuccessFactors fulfilled the request in about two months, says Tim Toterhi, director of global learning and development at Quintiles. Quintiles pays a monthly fee to access SuccessFactors software for some 6,000 employees, a number that is expected to grow to more than 15,000. Toterhi says the lack of upfront license fees was one of the reasons Quintiles signed its three-year contract-a typical length for software-as-a-service agreements. Toterhi declines to state the exact cost of the service, but says that "it was the best bang for the buck.'' Other vendors pitching HR software as a service include Wayne, Pennsylvania-based Kenexa, San Francisco-based Taleo, Orlando, Florida-based Workstream and Waltham, Massachusetts-based Authoria. The heavyweights of the field also are paying attention to the software-as-a-service trend. Oracle, for example, offers to provide hosted HR software, but clients still pay for a license to the application. Another way the Redwood City, California-based giant is tapping the trend to take software management off customers' hands is by working with an outsourcer. Oracle says its HR software is used by Gevity to serve more than 8,000 small and medium-size businesses. Oracle archrival SAP does not directly offer human capital management software over the Web. It also has an outsourcing strategy: SAP has arranged for its HR software to be used by several major outsourcers, including ADP and Convergys. Smaller vendors argue that SAP and Oracle will have a hard time moving fully to a software subscription model, suggesting that those companies' quarterly revenues would take a big hit without new license sales. Forrester's Hamerman, though, expects

both giants to come out with additional software-as-a-service products in the coming year. Price points On-demand software deals often are priced based on the number of people using the application, and may involve some initial setup fees. Workstream, for example, says companies signing a three-year contract to access its complete line of applications for 10,000 employees can expect to pay $1 per employee per month. Combined with implementation costs and a monthly hosting fee, the total cost is about $3 to $4 per employee per month. Michael Mullarkey, Workstream's chief executive, says software as a service allows for a dramatic price cut compared with the licensed model. A typical large organization might spend $2 million on a software license, another $4 million to get it up and running, and then another $1.2 million over three years to maintain it. Compared with that price tag of about $7 million, similar software over three years from Workstream would cost a business about $1 million, Mullarkey estimates. Licensed enterprise software "never paid for itself,'' Mullarkey says. "The return on investment made wasn't there.'' Forrester's Hamerman, however, says software over the Web isn't necessarily a better bargain. Renting an HR application may avoid a license fee and save on computer hardware and software maintenance, but big companies especially may end up saving money with the traditional model, he says. That's because per-employee pricing can add up when you have tens of thousands of employees using benefits management software or other applications, he argues. Using software over the Internet raises other potential red flags. Yahoo, for example, questioned whether it could zap employee performance data across the Internet before it signed up with SuccessFactors in late 2005. "The concern was security,'' recalls Libby Sartain, Yahoo's senior vice president for human resources. "We said, `We're going to have performance reviews online. Can we do that?' '' Yahoo had specialists dubbed "the paranoids'' check out SuccessFactors' security safeguards, which include data encryption. Yahoo has arranged to have 10,000 employees access SuccessFactors' software for performance management and goal setting. Yahoo also is looking into tapping the vendor for its succession-planning software. Even if companies can feel comfortable with sensitive data sent over the Internet, they may find their service provider doesn't give them exactly the features they'd like. SuccessFactors, for example, offers 15 colors for its Web interface, and companies have to choose one that's closest to the hue of their corporate portal. Nor does SuccessFactors

offer to tailor its software for particular customers who may want a highly specific feature. Bernshteyn counters that companies can "overcustomize'' their software, and says SuccessFactors' product allows for a large amount of variability on key business matters including job titles. The more streamlined approach, he says, allows the vendor to keep costs down and frequently churn out upgrades valuable to most customers. At their service? Another potential obstacle to a satisfying software-as-a-service experience is embedded in its very name: customer service. Most software companies aren't accustomed to emphasizing customer support services, says Kevin Dobbs, senior vice president for marketing and business development at Workstream, which offers a range of HR applications over the Web. But shoddy support spells big trouble to providers that offer their clients contracts as short as one year in duration, he says. "If they don't renew, the model blows up,'' he says. Mark Lange, vice president of human capital management at SAP America, says the support challenge is precisely why SAP decided to forge partnerships with outsourcing companies steeped in the service field, like Affiliated Computer Services. "It is very difficult to deliver best-in-class functionality and best-in-class service,'' Lange says. "We decided, `Hey, look, let's go with the best service providers.' '' In addition to claiming lower prices and faster upgrades, on-demand providers also say they can get started quicker-meshing their service with a company's internal systems in a few months rather than the years that on-premise software implementations can take. Even if the on-demand approach goes down easier than licensed software, some companies still might not be completely happy with what they're served. A chief reason given for failures in traditional software installations is customers' unwillingness to change corporate habits to match the "best practices'' coded into the application. To make on-demand software succeed, organizations will probably still have to tackle tasks like improving business processes and training employees. No matter what flavor of enterprise software a company chooses, that part of the recipe hasn't changed. MORE ONLINE: Software products combine learning and performance functions: workforce.com/performance CAPTION(S):

UP TO THE CHALLENGE: Peopleclick's Kathy Barton says frequent upgrades to her company's on-demand software make it easier for clients to address new business needs. * Rob Bernshteyn, SuccessFactors * Libby Sartain, Yahoo * Mark Lange, SAP America COPYRIGHT 2006 Crain Communications, Inc. This material is published under license from the publisher through the Gale Group, Farmington Hills, Michigan. All inquiries regarding rights should be directed to the Gale Group.

A 20-year-old overnight sensation: on-demand software arrives with a flourish.(PUBLISHING TRENDS) Searcher; 9/1/2005; Berinstein, Paula You know how the software industry always seems to reinvent services information professionals devised long agog They're at it again. Subscription-based software, on-demand, pay-as-you-go, software-as-a-service, metered services, hosted software, utility computing, provisioning--whatever you call it, we've used it for decades, for some of us even before SDC and Dialog first offered dialup database searching. So we're entitled to laugh when the CEO of Jamcracker, a company that hoisted itself onto the on-demand bandwagon in the late 1990s, naively claimed in the December 2004 issue of Software Magazine that the concept of ondemand delivery was revolutionary. Should we bother to tell him, or should we just gloat? Sure, the old and the new ways differ. Our vendors always furnished the data as well as the infrastructure. Nevertheless, the best practice remains the same: don't buy the horse, buy the ride or the right to ride. It may be the darling of the technorati now, but in the world of mainstream computing, the much-hyped on-demand model got slapped around a bit before finally taking root. Around 2002, ASP (application services providers), aka hosted software vendors, were supposed to proliferate like kudzu, but instead just withered on the vine. A plethora of problems did them in--little things like reliability, security, scalability, customization, integration, and identity management. At last, however, on-demand software (and its many aliases) has finally arrived, or at least, the pundits say it has. Market research firm IDC is projecting worldwide sales of hosted software to grow at an annual rate of 26 percent over the next 2 years--about the same as the annual rise in Southern California housing prices--to $8.1 billion. Fast growth, yes, although with a long way to go when you consider that the enterprise software market totaled more than $150 billion in 2004. Halsey Minor, CEO of Web services provider Grand Central Communications and founder of CNET Networks and other companies, and Marc Benioff, CEO of the much-

touted Salesforce.com, are betting the farm on pay-per-use services. Minor, also a founding investor and board member at Salesforce.com, has created the $50 million On Demand Venture Fund with his own money to finance on-demand startups. Benioff claims enterprise software is dying out. A recent InfoWorld cover screamed, "Will these troublemakers put IT out of business?" How Software Is Licensed Now To understand the new model, you first have to look at how software is offered now. Today, much of the burden falls on the user. You take responsibility. You acquire software and install it on your machine(s). If you're a big company, your IT department customizes and maintains it, integrating periodic vendor-supplied patches and upgrades. Enterprises invest huge amounts of time and money after purchase. Merrill Lynch estimates the costs associated with implementation at six to eight times the upfront fees. If you're a little guy, you probably use whatever you get from the manufacturer, investing fewer of your own resources than the big company, but still slowing down for learning curves and other time sappers. You buy a license, not a product. Software isn't sold--it's licensed. You never own software unless you write it yourself or buy the company that produced it. Instead you pay for the privilege of access, whether or not you ever use it. Enterprises purchase lots of privileges in order to estimate need way before it occurs; often enterprises overbuy to take advantage of attractive discounts. In practice, this model, like gym membership, works to the advantage of the vendor because, as the Gartner Group reports, about half the software paid for is never used or severely underused. You're in it for life. When you license software, you normally pay up front for what is known as a perpetual license. That means it's yours forever. That model is changing. According to a Macrovision/Software & Information Industry Association/ SoftSummit/Centralized Electronic Licensing User Group survey conducted in October of 2004, 33 percent of vendors are now selling subscription-based software, while 67 percent of them offer perpetual licenses. You pay up front. Charging models for perpetual licenses generally fall into one of several categories, but all require payment in advance: by the seat (an individual license), for a group (a certain number of concurrent users), or by CPU capacity. Under the last, your vendor charges according to the total potential CPU capacity of your systems, regardless of how much power you actually use. Problems with the Current Model Even though general protection faults have virtually disappeared, most software is still a royal pain. During April of this year, IT executives at Sand Hill Group's Software 2005 conference in Santa Clara, Calif., told attendees just that. According to Tony Kontzer writing for Information Week, tire IT execs let the software execs have it, citing abysmal quality, insensitivity to customers, out-of-control complexity, outrageous expense, and

security problems. One panelist, John Leggate, CIO and group VP of digital and communications technology at British Petroleum, reportedly suggested a new business model: software fees based solely on business outcomes. Are we angry or what? We are, and for good reason. Our issues include the following: Heavy metal. The first problem with the perpetual license model is that you need a lot of hardware, and the burden goes on and on. Once you've got your system in production, you have to keep installing new metal, which throws your IT people, accounting folks, and sometimes users into a tizzy. Every time a vendor releases a patch or, worse, an "upgrade" (do you really need all those new bells and whistles?), your system, which is usually pretty complex, can develop bugs. You want to quit and switch to something better, but change is so expensive and time consuming that, for all practical purposes, you're trapped. Enough is enough. Now that you're locked in, you may feel at the mercy of your vendor. An article at C/Net News.com from Oct. 19, 2004, on the SoftSummit conference in Santa Clara reported customers are fully fed up with bullying salespeople, arduous upgrade and maintenance work, and charges for features they never use. Taking matters into their own hands, these customers are increasingly using open source programs, cutting down on software consumption, and "renting" software. Potential inflation. Another issue looms for customers who pay according to CPU capacity. Multicore chip architectures are due in 2006. If you upgrade your hardware to a dual-core chip, your performance could improve by 30 percent to 55 percent, but your software costs could double. That's because you pay for each unit of processing power and you're going to have twice as many of them. Oracle, IBM, and Sybase charge this way now. Fortunately, Microsoft has announced new licensing structures for BizTalk and SQL Server, which will base charges on the number of processors, not cores per chip, and Sun is moving toward per-employee pricing and subscription models based on usage levels. Lest it sound as though vendors get off scot-free under current models, let me emphasize that this is not the case. Their products can be pirated. Their income spikes and plummets as customers finally decide to buy, then segue into a holding pattern. Vendors lose potential customers who balk at the risks and inflexibility of all-or-nothing pricing. Income is forfeited because with the pressure to book revenue by the last week of every, quarter, vendors have to offer discounts to get companies to buy fast. Some experts estimate that 60 percent of software sales are made this way. Companies know the routine and milk vendors for every cent that can be shaved off full price. How Software Works Under the New Models If the old models sound less than utopian, how much of an improvement could be gained under the new ones? We can look forward to the following:

More vendor responsibility With on-demand software, applications reside on vendors' machines, not yours, and you get to them through a Web browser. Vendors purchase, maintain, and upgrade the metal so users can cut down on capital spending and get about their business quickly. Vendors "also must fix bugs at the source and make the same clean code available to all users simultaneously. Vendor workload increases; yours should fall off. More than a few consultants and potentially some IT staff will find themselves out of work. New charging schemes based on usage and financial measures. Searcher readers already know about on-demand charging schemes, some of which are happily straightforward, such as Factiva's annual password fee, free searching, and $2.95 an article. We also know and don't love others, including Dialog's convoluted and hard-to-predict DialUnits. Expect more charging based on the number of uses of an application, the amount of time it's used, and the number of transactions. You'll also see offerings based on financial measures, such as revenue. User commitment. To eliminate some of the uncertainty from revenue streams, many vendors still require that customers sign renewable contracts, usually for terms of 2 to 3 years. In theory, because the term is finite rather than perpetual, vendors are motivated to provide an excellent product and stellar service. We'll see. Flexibility. Customers can find themselves a lot more agile under the new models. You can scale your systems easily because you use only what you need, not what you're forced to buy, although you can still remain locked into some term contracts and still face training and learning curve problems. Technology writer Steve Ulfelder said in his February 2005 ADT-mag.com posting that small to medium-sized businesses will benefit the most: These businesses will be able to get in on enterprise software without the cost and the necessity of in-house expertise. Cost savings. You may save money. Then again, you may not. Some experts contend that while the new method may seem cheaper, it could actually cost more over time because you pay and pay and pay. In the short run, on-demand software requires a smaller cash outlay. For example, CRM (customer relations management) software via ASP now runs about $50 to $100 per user per month. If you have 100 constant users, you'll spend $60,000 to $120,000 a year for the software--a lot when you consider that a perpetual license for a large organization costs hundreds of thousands of dollars. Nevertheless, as happens when you take out a mortgage, your cash flow will benefit. If your demand is seasonal or otherwise intermittent, you'll probably come out ahead over time. Loss of control. This one is a doozy. Customers can lose plenty of sleep over these issues: * Data security and privacy. Do we really want our sensitive and valuable information housed under someone else's roof? Can a hacker peek at our stuff, or worse, disrupt it? How much information does the software company get to collect about us? (Usage patterns provide inside information on a company.)

* Reliability Can someone else run our data center as well as we can? Will they stay in business? Are the partners they choose worthy? * Integration. Can hosted offerings work with our other applications? * Usefulness. Does this one-size-fits-all software really suit our business? * Control. We want our power and empire. * Charging. We don't think this metering scheme is fair to our company. * Interface. Browsers are OK for some applications, and we don't want to be locked into a proprietary interface, but browsers limit what we can do. (Acrobat and Flash plug-ins solve some problems, however, and Opera and Firefox are now starting to incorporate SVG, Scalable Vector Graphics, an XML-based language). * Legal hassles. Because the model is new, it takes time and money to resolve legal issues, and contracts can fill volumes. New vendor lifestyles. Vendors' lives will change in both positive and negative ways under the new models. No longer will they have to scramble to book big deals every quarter. In some cases, they will garner a more steady and predictable revenue stream, although still experiencing variations as customer usage patterns fluctuate. Their products will he far less susceptible to piracy. Vendors may also attract new customers previously unwilling to take large risks. However, vendors will have to invest more up front, and because they won't he able to hit and run, they will have to treat customers well or lose them, a novel idea in the software world. Vendors that sell both hardware and software could suffer the most because users will need less hardware to perform the same amount of work. While changing over to the new way of doing things, vendors and their investors will start to feel the pain. Reported revenue will fall and share prices may follow. One analyst says that companies could be penalized even if they have a good backlog of deferred revenue. Why Now? Why is subscription software finally taking off? The reasons are both technical and economic. Software sales have slumped over the last few years. The sluggish economy, catalyzed by the tech bust, has hit the industry hard. But another culprit is the very nature of the business. It takes a lot of time, money, and manpower to develop software. Brutal competition can kill a company overnight. The successful firms have survived by locking customers in or generating loyalty. Some have risen to the top because they got to market first and some

because they genuinely offer the best product. Consumer-level companies attract customers by offering low prices. However, to stay on top, vendors have to cut some throats. Sometimes those throats belong to users. Up until now, manufacturers of business software and operating systems have been able to bully business customers into buying more than they need. Their profits depend on user overconsumption and dependence. We all complain about the constant need for up-grades that don't give us anything useful but that we have to buy anyway to stay compatible. We all hate the planned obsolescence and the monopolies that allow manufacturers to hold us hostage. Most of us don't have the time or expertise to convert either to another vendor or to open source alternatives. In addition, software has traditionally been sold as a tangible item; believing we had to own something we could "touch," we paid more for it. Downloaded software has caught on to some extent, allowing vendors to decrease their distribution costs and prices and speed time to market. But downloads can seem interminable, and users worry about installing potentially disruptive programs. Many downloads are abandoned before being completed. Until recently, we've been stuck with the box or the download. However, new technologies and technical paradigms have allowed us to enjoy benefits such as common standards, scalability, ease of integration, and alternatives and have changed everything. Here's what you can be thankful for: * Open source. Open source arose as a rebellion against Microsoft. Now it's making inroads into the OS business and applications too. The open source Firefox browser is grabbing Microsoft Internet Explorer's market. In April 2005, Firefox's market share had risen to 10.28 percent, while IE tell 1.78 percent from the previous January, to 83.07 percent. Linux's market share in servers now comes to approximately 24 percent; Windows is down to 59 percent. In January, Business Week reported on a Forrester Research survey in which 52 percent of business users surveyed said they are replacing Windows servers with Linux. * Bigpipes. Broadband and streaming capabilities are finally maturing as prices fall and quality rises. Software can now be delivered as a stream rather than a lump you install (and that vendors worry about being ripped off). Streaming software allows publishers to offer time-limited trial versions with all features enabled. It attracts eyeballs to Web sites, allowing purveyors to make revenue from advertising. It can even act as a stopgap between the purchase and receipt of boxed software. * New architectures. Web services--Web-based software modules--and SOA--an architecture that creates a collection of services that communicate with each other--allow customers to mix and match the features they want. Hardware can reside outside the organization, delivering economies of scale. Of course, the new models won't work for every application. Imagine having to pay forever and ever for heavily used programs such as Microsoft Word, Outlook, or Excel, or a mail program such as Eudora. You could easily end up forking out much more that

way than by just buying outright. And, as online searchers know, even with software you rent once in a while, you can still lose: Running meters can intimidate and learning curves cut into productivity. A Wish List Because the trend is just getting underway, we users can still get in on the ground floor. Vendors, are you listening? Now that the tables are turning and you need to make us happy, how about considering a few requests: Lease or buy. Give us the option to purchase or subscribe. If we subscribe, our set-up fees shouldn't break the bank and we don't want to pay for the "privilege" of terminating. We're happy to pay for what we use and we want you to succeed, but we're not your banker. We'd like a low-cost learning option that is limited by time or some other measure. This way, you assume most of the risk for a change, but if you do your research first and develop stuff we want and can use, you will minimize your exposure. If the software is must-have, we will still purchase rather than rent. A la carte. Let us buy by the feature. Why not offer a menu of capabilities and let us pick and choose which ones we want? Systems and menus would be simpler and more useful to individual customers, who could still receive information on other options at any time. We like regular and "lite" versions, such Adobe Photoshop and Photoshop Elements, which allow for varying needs and skill levels. Say-so. Let us provide input before you design the software. Do your homework before turning a screw! (Where have we heard about the value of research before?) Users should contribute regularly, from development through production and implementation to follow-up and maintenance to development again. Executives in customer companies can justify employee time spent this way in dollars saved down the road. Partnerships. Vendors and customers can experiment with new models. Maybe a group of users gets together and forms a consortium to fund advanced projects. Maybe a group of non- but potential users does the same. Why should developers get to make all the decisions? When the software becomes commercial, the consortium shares in the profits, if, of course, there are profits. Wouldn't you love to commission Adobe or heaven forefend, Microsoft, to develop something fabulous for you? How about having a piece of Amazon's, eBay's, or Netflix's software that could be sold to firms around the world? (Yes, if you're a stockholder in one of those companies, you already do, but why not a more direct involvement? In essence, users could become angel investors.) Whither Software? Despite their pain, users are still acting schizophrenic. The Macrovision et al. survey cited earlier found that enterprises prefer perpetual licenses over hosted software two to one, although larger companies demonstrated more enthusiasm for change than small ones. Forty-four percent of companies with more than half a billion dollars in revenue prefer subscriptions, while only 33 percent of those with less than $50 million do.

These customers may soon have little choice. The survey also found that 52 percent of vendors expect to offer subscription-based software as their primary model by 2006. That's a substantial change from the 67 percent of software now purchased through an upfront license with ongoing maintenance fees. Under the new model, software manufacturers will have to look for new sources of revenue. About 20 percent of their revenues presently come from maintenance fees; a full 80 percent derives from all upfront fees. Don't be surprised if your tech support costs soar and companies nickel and dime you for every little addition. Hefty early termination fees might be in the offing as well. So speak up now. by Paula Berinstein Consultant Berinstein Research RELATED ARTICLE: On-demand software sampler. Here are some companies that offer on-demand software now. Some products are hosted; others involve obtaining updates over the Web via subscription. Salesforce.com http://www.salesforce.com The darling of the business press is known for its customer relations management (CRM)-hosted software starting at $65 per user per month. Clickability http://www.clickability.com The San Francisco company produces cmPublish Web site content management software as a service. Starts at $749 per month for a full installation. Autodesk http://www.autodesk.com The maker of 3-D software for designers and visual effects professionals provides updates and enhancements, support, maintenance, and training via annual subscription. McAfee http://www.mcafee.com

The company's Managed Services--McAfee Managed VirusScan, McAfee Managed Mail Protection, and McAfee Managed Desktop Firewall--are available via a 1- or 2-year subscription, with a 25 percent discount if you opt for the longer term. You pay less than $35 per year per license per product under a 2-year subscription. Symantec http://www.symantec.com Norton AntiVirus, Norton SystemWorks, Norton Internet Security, and other products include a complimentary renewable 12-month subscription that provides virus definition, Web filtering, and intrusion protection services. Annual subscriptions run in the $25-30 range. IBM http://www.ibm.com Big Blue features a panoply of products available by subscription, from e-business hosting services to developer Works to administrative tools such IBM Director. Paula Berinstein is producer and host of the podcast, The Writing Show [http://www.writingshow.com]. She can be reached at paula@compulsive creative.com. COPYRIGHT 2005 Information Today, Inc.