Barely are any of the problems encountered in the day-to-day use of a computer or network the fault of the hardware. The glitches, stops, starts, and crashes are the result of bad software ? programs that are too complex, poorly written, crammed with too many features, or that conflict with other programs.
IT executives are in almost universal agreement: Software is a nightmare. At the Spring 2005 software conference held in Santa Clara, California, by the Sand Hill Group, IT executives laid into software vendors over the embarrassingly poor quality of their offerings, the astonishing complexity of programs, the fatal security problems they’ve experienced, and the outrageous costs. At that conference, the head of IT at British Petroleum suggested that vendors start charging based on business outcomes: You make us money ? we’ll pay for your software.
It’s no wonder that suggestion was made: Last year, more than half of all new enterprise software projects came in late, over budget, or simply didn’t work as advertised, according to CFO IT. That incompetence on the part of software vendors will cost the business community an estimated $84 billion this year alone.
Software, it seems, is one of the few business models in which a firm can keep its customers perpetually dissatisfied and angry ? and still make money.
But that model is about to change, according to an article in Red Herring. Software as a service is fast becoming the new model that’s challenging those vendors to get their act together.
How does this new model work? The vendor keeps the software on its own computers. The customer pays a monthly fee to use that software over the Internet. The vendor is responsible for maintaining its own hardware to run the software. It’s also responsible for keeping the software running and installing any upgrades or patches.
In effect, all the customer does is provide the data on which the software will operate, such as data relating to inventory control or payroll. The data is processed remotely by the vendor, and the results are sent back over the Internet.
This is a far cry from the traditional model, in which IT departments play an endless game of catch-up, having to buy new hardware to support new software ? and then having to upgrade the software every time the vendor discovers another bug or adds another feature that the customer may not even want. After a time, a client company has sunk so much money into the system that it can’t switch, even if a better product comes out.
Fed up with that model, more and more companies are moving toward software on demand. It’s said to be saving them as much as 90 percent over the cost of a traditional installed base.
In fact, a report from Nucleus Research showed that 82 percent of those who were using on-demand software achieved a positive return on their investment ? compared with around half of businesses that licensed software under the old model.
As a result, licensed software has been hit hard since its double-digit growth of the 1990s. It’s down to less than 5 percent this year, according to the latest research. While traditional packaged software still represents 90 percent or more of the revenue in that sector, on-demand software is growing fast and is expected to represent up to 4 percent of global sales by 2009.
The Sand Hill Group, a software consulting firm in San Francisco, is calling “software as a service” the biggest thing to come along since home computers. It’s growing at a rate of 21 percent a year and racked up $4.2 billion in sales in 2004. That should rise to more than $10 billion in 2009.
The biggest bite that on-demand software has taken out of the traditional market has been in customer relationship management. Traditional applications in that area lost 4 percent in 2004, while software as a service rose between 5 and 9 percent. This has forced the big vendors, such as Oracle and SAP, to begin introducing their own on-demand product lines. Even Microsoft, which dominates the market for software licensing, is offering its packages on-line now.
One of the companies that has been a driving force behind the changing software model is called Salesforce, founded by Marc Benioff in San Francisco. Benioff, formerly of Oracle, has managed to push growth at Salesforce up 245 percent to $176 million last year. It now has more than 300,000 subscribers paying about $65 a month for its on-demand CRM services. At a time when IPOs aren’t what they used to be, Salesforce saw its stock, introduced at $11 in June of last year, shoot up to $20 this summer.
Among the other firms that are emerging as players in this new market are:
- RightNow Technologies, where earnings have grown 130 percent between 2002 and 2004, going from $27 million to $62 million. - NetSuite, which offers enterprise resource planning software on-line, and has grown 250 percent each year for three years running.
Meanwhile, the software establishment and the upstarts are waging a war of words over what the future will hold. Microsoft, Oracle, and others are predicting that a comfortable mix of on-premise and on-demand software will make up the business landscape. But CEOs from companies like Salesforce and RightNow Technologies say that in five years all software will be on-demand, and there will be no installed bases.
But most analysts do not subscribe to this scorched-earth view of the software industry. Their more moderate view is that on-demand software will gradually increase its share, especially in areas like customer resource management. Two of the big issues that new start-ups don’t like to talk about are security and scalability.
Security is an obvious problem with the Internet. Companies naturally worry about their customer and company data, and a breach could send the reputations of both vendor and customer tumbling.
In terms of scalability, at the moment, software that works over the Internet is mostly serving small- to medium-sized companies. Those systems aren’t yet able to handle billion-dollar companies. They don’t have the rich features or the ability to be customized for the needs of large corporations.
In addition, the new model of on-demand software is going to require adjustments on Wall Street. Traditional software companies have operated on a boom-and-bust cycle of earnings. They may record millions in sales in some quarters when a large base is installed, and far less in others. With on-demand software, the income would be smaller but steadier, and no one is yet sure how this will affect stock prices.
In light of this trend, we offer the following six forecasts for your consideration:
First, software as a service ? or on-demand software ? is here to stay. For example, Oracle recently hired 600 new engineers to make sure that it is a player in this new game, and within the next five years, Oracle expects its on-demand model to outstrip its traditional software sales. In short, every enterprise software maker will have to begin offering hosted applications on demand, even if it retains its old model of software sales and installed bases.
Second, expect a spate of new IPOs from software-on-demand companies powered by new venture capital being pushed into this market. Because everyone needs software, this could become one of the hottest areas for investment and growth in the next five years.
Third, expect to see a flurry of acquisitions and partnerships, as companies such as IBM muscle into this lucrative new area. SmartOps Corporation, which provides enterprise-class supply chain optimization solutions, recently announced a partnership with IBM. On-demand software is growing at 21 percent overall, at a time when tech growth in general is sluggish in the single digits. With the independent software companies growing at 16 percent and IBM having just purchased Corio, we can see the large companies snapping up the start-ups to stake out their position in the on-demand market.
Fourth, the start-ups that can master the software-as-service formula now will be the big winners a decade from now. Remember that most paradigm-shifting technological changes have started small and grown big. The fact that on-demand software has so far penetrated only the small-to-medium market is not a sign that big business won’t move in that direction, too. And those big companies that can make the change ? such as, perhaps, Oracle ? will share in the rewards. Some of the traditional vendors that resist the move, or stumble, may face serious financial difficulties, or even find themselves being acquired by the upstarts.
Fifth, the technical problems now standing in the way of large-scale adoption of an on-demand model for software will evaporate with technological advances by 2015. With 128-bit encryption a standard on every laptop these days, security issues in the future will be what they’ve always been: The human will be the weak link in the loop. Some applications will have to remain in-house as a result of that weak link, such as design work for auto makers, who wish to protect their new product lines from prying eyes.
Sixth, the problems of scalability will be solved through technological advances already underway in multi-core processors and ever-faster, fiber-based and wireless broadband. For example, Salesforce just announced an investment of $50 million to expand its data centers to serve the largest customers. In short, most standard business applications will go on-line, and those companies that do the best job at optimizing costs, efficiency, features, and customer satisfaction will be the big winners.
References List : 1. Information Week, April 28, 2005, “IT Execs to Vendors: Your Software Stinks,” by Tony Kontzer. ⓒ Copyright by CMP Media LLC. All rights reserved. 2. CFO IT, Summer 2005, “Software as a Service,” by Norm Alster. ⓒ Copyright 2005 by CFO Publishing Corporation. All rights reserved. 3. Red Herring, September 19, 2005, “Is Enterprise Software Dead?” ⓒ Copyright 2005 by Red Herring, Inc. All rights reserved. 4. CFO IT, Summer 2005, “Software as a Service,” by Norm Alster. ⓒ Copyright 2005 by CFO Publishing Corporation. All rights reserved.
