by Christopher Koch

2010: The Future of Software

Dec 15, 200313 mins
Enterprise Applications

Scenario One A Land Where Giants Rule

By 2010, CIOs will find themselves hostage to a few monopolistic vendors that keep software expensive and complex

Software vendors are like dinosaurs?they live in an environment that encourages them to grow really big to survive.

The bigger the vendor, the safer CIOs feel. It’s not that CIOs are happier with the behemoths?in fact most of them hate the endless cycles of upgrades and complex maintenance and support agreements that the big vendors offer. But a big vendor promises more longevity and suites that are often easier to integrate than the best-of-breed applications from many smaller vendors. The lumbering beasts rule not because they are better or faster or more efficient than their smaller counterparts, but because they solve some?though far from all?of the CIO’s integration problem.

And that’s why CIOs will keep shoveling food to the big beasts and starving the little ones through 2010. The only other segment to get a healthy feeding will be outsourcers and services vendors that offer software bundled together with integration, upgrades and customizations as part of the price. Call it hyperconsolidation of the software market.

Unless Web services lives up to its promise?and we predict it won’t?integration will remain a huge burden that drives up the overall cost and complexity of IT. The trend toward outsourcing will continue as many companies decide that their CIOs cannot solve the integration problem on their own. Web services will remain tightly controlled by a handful of large vendors that have no economic incentive to fix the problem. That would simply free customers to choose smaller, more nimble competitors.

Open-source software will not be the answer to integration problems either, though it will continue to drive down prices in selected areas of the software infrastructure. In fact, open source may turn expensive databases such as Oracle and IBM’s DB2 into commodities by 2010. Smaller vendors will move toward the open-source model because it will lower their marketing costs. Rather than devoting 15 percent to 50 percent of revenue to selling, they will simply give it away, build a user base through word of mouth and then sell services and add-ons.

If open source takes hold in the big corporate infrastructure, enough momentum could build for an open-source integration technology to emerge that would rival today’s Web services model. But solving the integration problem would be so complex that it’s unlikely?unless someone with the technology savvy and leadership abilities of a Linus Torvalds emerges to head the effort.

Net result: Big company CIOs in 2010 will find themselves increasingly beholden to an outdated economic model for purchasing, installing and maintaining software. Fewer, bigger vendors will mean higher prices, fewer choices, higher switching costs and increasing vendor lock-in. There is a model that could bring the dinosaurs to heel: It’s one where vendors would sell applications as specific, highly configurable components that upgrade automatically and integrate with any type of system at no cost and minimal effort. Companies would pay for these applications only when employees use them.

This vision, however, will not arrive by 2010 for purely economic reasons. Wall Street might throttle the stock prices of vendors that try to swear off the big up-front license fees that pump up quarterly revenue in favor of the slow drip of a pay-by-the-drink model. And the big vendors simply have too much invested in the status quo to open up their markets to smaller competitors. (For a different outcome, see “Open Source Slays Goliath,” Page 58.)

“For customers, [the enterprise software business model] is like being held hostage,” says Julie Giera, vice president and research fellow for Forrester Research.

The Hostages

CIOs hate the current model of buying and selling complex big-business software such as ERP and CRM. That’s because the cost of integration is borne entirely by the buyer. The smaller vendors often offer more capabilities than their bigger, integrated counterparts, but because there is no standard technology to link software from different vendors together, the little guys are a riskier choice. Assembling many vendors together in a best-of-breed strategy can cost more to install and take 25 percent to 37 percent longer to change and test than large, integrated packages from one vendor, according to Forrester.

Thus, CIOs put their careers on the line in a desperate guessing game of trying to determine up front how much an enterprise package will really cost the company?if businesspeople will accept it, and if it will truly have any beneficial impact to the bottom line after months or even years of disruption and change. In a 2002 survey of 600 companies that had purchased CRM software, Gartner found that on average 42 percent of the licensed software was sitting on the shelf unused, accruing additional license fees in support and maintenance charges.

Despite such waste, customers of enterprise software are locked in to their vendors with a padlock rusted shut by years of neglect. Moving from one vendor to another is so costly as to be prohibitive?just 1 percent to 2 percent make the move willingly, according to Meta Group analyst Barry Wilderman.

While application vendors still refuse to share the burden of integration with their customers, services vendors are learning that by sharing it, they can not only pump up their consulting revenue but their software revenue too. For example, Computer Sciences Corp. (CSC) already has such a pain-sharing program for the insurance industry called Virtual Development. Customers get a copy of the software source code that they can modify but only with the help of CSC consultants. CSC agrees to incorporate all customizations into the next version of the software, relieving a huge workload for customers. No more costly rewrites of customizations each time the vendor upgrades the software. By incorporating all the changes, CSC builds a software package that could become an indispensible industry standard (or an ungodly mess if not managed carefully). Customers are protected from becoming overly dependent on CSC because they can take the source code to someone else to run it for them.

With its huge advantage in services and outsourcing, IBM could complete the hat trick by acquiring enterprise software vendors such as Siebel Systems and Oracle, and bundling guaranteed integration and upgrade services with the software. Whether the Department of Justice would agree remains to be seen. For CIOs, it would mean one great big throat to choke. If this happens, pure application vendors such as SAP and Microsoft will be at a tremendous disadvantage against companies that combine software and services together under one roof.

Innovation Dries Up

Of course, this is just consolidation on a broader scale and does not resolve the issue of integration. Software and services vendors don’t just eat their own anymore?they become omnivorous, and the line between software and services blurs completely. Will a services vendor that owns lots of software have any more economic incentive to create a set of universal integration technologies than a standalone software vendor? Probably not.

By 2010, there will be fewer vendors taking more corporations hostage. The unsolved problems of integration and IT complexity will drive acquisitions of smaller software vendors by bigger vendors eager to create one-stop shops for CEOs who want to outsource IT.

Such hyperconsolidation will probably not end IT complexity but will hide it in the shadows of fewer dinosaurs. That will do little to foster IT innovation. With CIOs locked in to services vendors that also offer a stable of software solutions, it will be even harder for small vendors to get a meeting. The dotcom bust taught CIOs a hard lesson about investing in startup vendors with good ideas and lousy balance sheets. Likewise, venture capitalists are much more leery about simply throwing money at anything that has the word technology in the business plan. We predict that a handful of companies will control more than 70 percent of the Fortune 500 IT budget in 2010. Small vendors will play to small companies exclusively ?to the great disadvantage of big companies that won’t have access to the innovative technologies that startups are developing. Small companies will be happy to nibble away at chunks of software rented to them over the Internet for a reasonable monthly fee. After all, neither vendor nor customer has to worry about integration.

CIO as Contract Manager

Hyperconsolidation won’t help the CIO role, either. Reducing CIOs’ strategic options makes CIOs less strategic and more like vendor managers. Vendor consolidation will pull power and strategic control away from the CIO and make the role seem less defensible to skeptics inside the company. With fewer vendors taking over bigger pieces of the IT infrastructure?and outsourcers vying for the rest?strategic knowledge and future technology planning will increasingly reside outside the company. CEOs and CFOs may wonder why they can’t simply meet with the few remaining vendors themselves instead of having a CIO do it for them. And any arguments that CIOs make about maintaining multiple vendor relationships to avoid lock-in will fall on deaf ears. Without universal integration technologies to open up software to true price competition, what’s the difference between paying steadily rising fees to separate vendors or simply giving it all to a services company?

In this kind of environment, open source will be the only economic weapon CIOs have for keeping the lid on big vendor prices and contract terms. CIOs should set aside portions of their budgets to fund open-source pilot projects inside their companies if for no other reason than to create a negotiation tool to hold over the vendors’ heads. CIOs who invest the time and money necessary to assemble low-cost open-source alternatives to enterprise software packages could win back the hearts of skeptical CFOs and CEOs. It will be a long, slow process that stretches well beyond 2010, but top CIOs will no longer be unaware of how much open-source software they have in their infrastructure. Open source will move from bottom-up to top-down.

Indeed, by 2010, the strategy that will differentiate top CIOs from their more tactical peers will be one that focuses on a vendor-neutral architecture and less vendor reliance?not more. It will be a difficult task, but CIOs who focus on creating an infrastructure that is low cost, easily maintained and doesn’t rely on a handful of vendors to function will have the upper hand in price negotiations with vendors and the ability to adopt innovative new solutions more quickly and easily than those CIOs who are locked in to a vendor’s software release schedule. CIOs who merely pick vendors from a list will have a more appropriate title: contract manager.

Scenario Two Open Source Slays Goliath

By 2010, European and Asian governments will lead the way toward the adoption of open-source software, and American CIOs happily climb aboard

Hyperconsolidation might not rule the day if governments and companies decide that it is too risky to become reliant upon a handful of vendors for their IT. By 2010, nervous governments in Europe and Asia could instead shift their purchasing and development dollars to open-source software. Overseas anxiety about vendor lock-in could be compounded by the fear that the single vendor will be American and might put European and Asian needs second. Many European and Asian companies might follow their governments’ lead by choice and by necessity?if their governments mandate open source for communication, they will need to comply.

Such a trend could force sweeping changes in the model for buying and supporting software. It could even spur a revolt among CIOs who decide that they have had enough of the endless cycles of complex licensing agreements and upgrades on enterprise software. Instead, they could adopt the open-source licensing model and turn everything they’ve paid for out to the market for free.

“What if corporations say, ’We don’t believe in licensing agreements anymore, and we’re going to open source all our [legacy] software and distribute it for free’?” asks Paul Maeder, managing general partner of venture capital company Highland Capital Partners.

By 2010, open-source alternatives for complex enterprise software are encroaching from all directions. Customer unhappiness, combined with the commodity threat of open source, will drive big vendors to completely revamp their sales, pricing and installation models. Some will make older versions of their software available for free and sell support to make money. New software will come with integration and upgrades included and be paid for on an as-needed basis. But some big software vendors won’t be able to make the switch, and they will be swallowed up by services vendors that offer clients paid support for open source as well as proprietary software. CIOs will once again be running development shops, but this time the work will be open source as often as it is proprietary.

Under this scenario, CIOs will be architecture experts who take a hands-on role in creating cheap, standards-based infrastructure from which they will build highly customized IT-enabled business processes based on open-source standards.

“We’ll see a model where you don’t pay for the software but the services,” says Nick Gall, senior vice president and principal analyst for Meta Group. “Open source and commoditization is a bottom-up process. It will move slowly up over the next 20 years to the top of the stack. It will be a slow, painful process for vendors.”

There are patterns emerging today that could make this scenario possible. Like operating systems before them, complex enterprise software?core ERP functions and CRM?are sitting still like bunnies in an open field and make easy targets for open-source duplication. Already, small, very limited versions of open-source ERP and CRM exist. Some of the companies that have adopted them have invested in custom development projects to add functionality to the packages and have agreed to freely incorporate the new code into future releases.

A market will emerge to service these applications. Small developers will get paid by the line of code, but also for servicing and supporting the stuff they’ve written. If their client decides to make the code available for incorporation into the open-source package for redistribution, they will be able to sell services to other companies that adopt it. Open source will become a respectable route for startups and individual developers, and proprietary add-ons that need service will attract venture capital money to build real businesses.

Could it happen? Sure. “All you need is one good set of code out there” to act as a foundation for building the complex software systems, says Jeremy Allison, developer of Samba software. Open source avoids the biggest barrier to entering the software industry: marketing and sales. Open source needs no sales and marketing budget, only a good development leader, quality software and word of mouth for adoption. The open-source enterprise software is not free, but it is cheaper, and services vendors that install and run it for customers are happy to contribute paid developers to the cause. Innovation will flower because it will be much easier to get new projects going and to sell add-ons for existing open source. To separate the promising software from the bad, good CIOs will be more in demand?and more valued?than ever.