Vendor consolidation expands user choice?

One of the more obvious and overstated observations regarding enterprise application markets is that they are consolidating. That's the reality, but taken in isolation this notion can mislead users into thinking their choices—when it comes to enterprise solutions—are shrinking.

The truth is we are entering a new phase in enterprise application markets. The name of the game is change, but what's happening is very different from what's gone on before.

First, let's consider the nature of the current "consolidation."

Unlike the consolidation that took place in the late 1980s—when a company such as ASK acquired NCA so as to kill a competitive product—most acquisitions today are made with the intention of keeping the current product solution alive and reasonably well fed. SSA Global, for example, has made a good business of buying so-called "dead" products, re-energizing them, and deriving revenue from them. SSA Global says it plans to be a $1-billion company within a few years. It will not be meeting that goal if it kills the product solutions it purchases.

In another example, regardless of the initial posturing made by Oracle, if its proposed acquisition of PeopleSoft does occur, it's a fair guess to say Oracle will maintain and enhance the PeopleSoft/J.D. Edwards product lines for quite a while.

The same pattern is seen, and will be seen, in other instances and markets.

The fact is that while the number of vendors may be consolidating, the number of products is not. This presents users with a novel set of opportunities. It implies that if a solution has critical mass—i.e., a few hundred customers—and is integral to key business processes, it will survive regardless of which vendor manages the solution. The acquired solution must become material to the growth of the new vendor or the aggregator of products that did the acquiring.

As buyers of enterprise applications consider technology purchases, therefore, they are now free to focus on the viability of the product as much, or more, than the viability of the vendor. Given this change, the decrease in the number of products will slow, although the decrease in the number of vendors will continue at the current pace.

At the same time, new trends will increase even further the choices of those deploying IT-based solutions. You don't need to buy enterprise applications anymore. Leasing solutions is back, as is building your own.

Building or leasing applications offers users a new set of choices as to how they'll implement enterprise applications. New technology, as well as a desire on the part of buyers to minimize complexity, cost, and risk is moving them toward these alternatives.

One of the most exciting possibilities is a fundamental change in the build-versus-buy equation.

Readers may find it strange that a packaged applications doyen such as myself is talking about building rather than buying applications. But three factors will get users to build software rather than buy it: open source, offshoring, and new development tools.

The speed with which open source is being adopted, and adapted, by companies is startling. Use of open source began with core technologies such as the Apache Web application server and Linux, but is quickly moving into the applications space in areas such as content management (Zope, Red Hat, OpenCMS); ERP (Compiere); and CRM (Anteil, Ohioedge).

Compiere, for example, claims more than 630,000 downloads of its software have been made. Even if exaggerated, it implies several thousand companies are somehow using its intellectual property for some manufacturing processes.

These applications may not be ready for plug-and-play prime time, but two of my clients already have swapped out purchased content-management solutions for their open-source counterparts. To start, the downloaded solutions may not have had all the bells and whistles desired, but with a little customization, the price was right and the clients got what they needed.

The historical reason companies gave for purchasing applications was that custom applications were expensive and time-consuming. Offshoring considerably changes the cost dynamics of that equation.

Rather than pay internal IT, or consultants, between $50 and $200 per hour (fully loaded), how does $5 to $15 an hour sound? How would that change a build-versus-buy strategy decision? One offshoring company, Aztec Software, is combining open source code and offshoring expertise to offer its customers a powerful combination.

Finally, new tools that use Web services technology promise to make customization even easier. These already-introduced tools will eventually allow creation of service-based composite applications based on customized and widely available components.

Such tools are not the exclusive domain of U.S. companies. An offshoring leader such as Tata Consultancy Services has its own set of software tools to speed application development. And global company IBM, for example, recently announced that it's building new J2EE banking applications using an Indian developer. It also announced it will build custom, industry-oriented software solutions.

When you think about it, if a user examines just two of the three capabilities—open source, offshoring, and new tools—the notion of build rather than buy gets more interesting. It also dramatically increases choices and solutions available to solve a business problem. If all three are tapped, building applications is an option that must be considered for many business process needs. At the very least, the new dispensation helps users bargain down prices asked for from software vendors.

Leasing or paying a subscription fee for applications is at the opposite end of the spectrum from building applications. But it's already common for HR and payroll applications. In fact, a variety of sales and deployment models for enterprise applications are taking hold in the market. A curiosity only a few years back, many sellers are considering adopting these new pay-as-you-go models today.

A key reason is that these type deployments minimize the risk of failure for the buyer. Receiving only low up-front payments, sellers are motivated to deliver workable solutions that can be widely adopted by their customer bases.

One of the more successful vendors taking this approach has been CRM vendor Salesforce.com, formed in 1999. Estimates say it exceeded revenues of more than $100 million last year. Having filed for an IPO only last December, these are the best figures available. It has more than 9,000 customers, and its per-seat price starts at $65 per month.

Logistics vendor Descartes Systems Group also has moved from a traditional license fee model to a subscription model of deployment. Using baseline- and transaction-based pricing, it's near profitability, although it has taken more than four years and lots of cash to make the transition.

Larger companies already are starting to recognize revenue—for accounting purposes—based on this model. Both IBM and Computer Associates (CA) sell products using a model called "on-demand." On-demand computing permits a buyer to pay only for the technology it uses at any given point in time, even when it has a larger amount of it installed onsite.

In one of the more extreme examples, a CA airline customer pays for certain software based on passenger miles flown. CA's bill goes up and down depending upon how successful the company is in loading its planes with people, even though CA's software has little to do with this metric.

In the summer of 2003, IBM signed an exclusive agreement with toy maker Lego to deliver all its hardware systems, replacing those of competitor Hewlett-Packard. Key to the deal was that Lego pays IBM only for the computing power it uses; Lego will not need to purchase extra systems to ensure IT works well during the peak Christmas season. Instead, it will use IBM hardware "on demand," with expected cost savings at as much as 30 percent.

A great byproduct of proliferating choices is a consistent flattening or dropping of prices for all but a few enterprise applications markets. For example, at Boston-based AMR Research's recent Strategy/21 conference, Mike Greenough, CEO, SSA Global, said average selling price for his company has dropped by 40 percent over the last three years and continues to do so. Large customers that used to pay more than $3,000 a seat of SAP, Oracle, or PeopleSoft sometimes pay less than $1,000 today.

While niche software markets have been somewhat immune from this squeeze, they're experiencing the same pricing pressure as buyer expectations change.

Savvy buyers find they have more choices than ever, at lower price points. But with these choices they have decisions to make. Aspects of the enterprise software market will continue to consolidate, but by looking beyond the simple headlines, buyers will find a proliferation of choices to permit them to deploy the best solutions at an increasingly lower price.

Erik Keller's new book, Technology Paradise Lost, examines massive changes occurring in the buying and selling of IT products and services.