AdSlot 1 (Leaderboard)

Software development – Blue sky thinking.

In many ways, 2003 is not the best year to start peering over the horizon, asking what business systems could be delivering in three-to-five years. Today, companies are so focused on cost and sweating the assets – particularly their IT assets – that any software systems vendor who wants to talk futures is speedily shown the door. However, it is often the case that just when everyone has their eyes on the here and now, questions about the likely shape of future events have a particular pertinence.

Of course, long-range research and development does not come out of nowhere. In any industry sector there will be specific drivers, and the business software industry is no exception. There are a few major trends that business applications vendors such as SAP, Oracle and PeopleSoft have had to react to, and those trends help shape the future.

The most obvious of these has been the arrival of the internet and the web, which is still dominating change in this sector. Take Microsoft’s .Net web services vision, which is about as internet-dependent a notion as one can get. In reality, many Fortune 1000 companies are still in the process of re-engineering their business processes to take advantage of the anytime/anywhere access to information introduced by the internet. Its implications for extending the enterprise outwards, opening it up to clients, suppliers and staff, are enormous and best practice is still being defined in a variety of areas.

The internet changed the dynamics of the software industry from a client-server model to a web-based model, according to Simon Harrison, chief technology officer at SAP. By comparison with earlier technology, the internet is cheaper, faster and more convenient since users can access it simply use a web browser.

On the down side, however, this ease of access raised a host of security issues, from hacking to fraud to computer viruses. Everyone is still grappling with these issues and part of the R&D being conducted by vendors attempts to address these issues.

We have seen the emergence of an evolving multibillion-dollar industry in user authentication and directory services – how one recognises users in a world where anyone can dial in from anywhere, and how those users are matched to their access privileges. However, much of this new functionality around security is being provided by specialist third parties, or being built as a bolt-on layer to existing business systems. By far the largest slice of business systems vendor R&D spend – up to 2001-2002 – was taken up with a major redesign, or total rewrite, of their software to shift to the new internet world. The aim was to produce e-suites instead of clunky, client-server suites of applications.

Until last year or so, a second and rather different strand to their R&D was aimed at adding functionality to software suites, according to Andy Sutton, VP of Oracle’s UK Financial Services Business. If companies did not have an HR module, they worked to develop one. If they were weak in manufacturing, they worked to strengthen this element in their suite, and so on.

“If you look at Oracle and other vendors going back several years, our customers were clamouring for ever-increasing functionality from business applications,” he says. Sutton argues that the need for more functionality has largely subsided, or undergone a fundamental change. By and large, enterprises now have sufficient functionality. The focus now, Sutton argues, will be on getting better value from what companies already have. “If you are a software vendor, phoning a client today with the message that you’ve got a new applications module is not going to get you in the door,” he says.

Vendors can’t completely neglect adding functionality, but that is taken as just a ‘go-to-market’ given. If you don’t have it, you won’t be competitive, but new functionality will not drive sales. So where are the vendors spending their research dollars?

For Harrison, a key design driver has to be making software simpler and easier to implement and use. “Users want systems that are easy to learn and attractive to use, so there is inherent encouragement to make the most of what the system can deliver,” he says.

In other words, after decades of building increasingly complex systems that took armies of consultants and millions of dollars to implement, vendors now have the task of eradicating much of the difficulty. Things have to become simple, intuitive and obvious, not just at the user interface but also at the level of systems implementation. This is difficult to accomplish and only the fact that we are now getting amazing amounts of computing power in affordable (ie, sub-£500,000) enterprise servers makes this a feasible goal.

Simplicity also equates to much shorter implementation cycles. Achieving simplicity, Harrison points out, means that SAP will be in a position to address mid-market and smaller companies with big company systems and functionality. SAP has been threatening to reach down into the mid-market level to compete with mid-market systems vendors for a long time, without conspicuous success. Success here will put a tremendous squeeze on these vendors and could change the dynamics of the package software market in the next few years (see also Decisions, page 5).

“We are spending a lot of time with end-users now as part of our development process, trying to give them a process flow that moves smoothly along different task lines, so the system takes them along in easy stages. This is where our work is going,” Harrison says.

Sometimes, the future of a market is shaped as much by the arrival of a new player as it is by innovative R&D, and sometimes the two go hand in hand. The entry of Microsoft into this market with its purchase of two mid-market vendors, Navision and Great Plains – one strong in Europe, the other in North America – looks more than capable of being such an event.

Andy Smith, general manager of Microsoft Business Solutions, points out that MBS is now hard at work developing a next-generation business suite, due for release in about three or four years. “We’re working closely with our .Net team and SQL Server database team as we develop this new vision,” he says.

Smith points out that the all-existing enterprise business applications, even those rewritten for the internet, are based on ideas that are some 10 years old. They don’t draw easily on the huge information resources of the internet itself, he argues. “Success in future will not just be about getting a zero footprint client, or being able to deliver information on a personal digital assistant. It is about using the internet intelligently and this is where our .Net vision come in,” he says.

.Net is, in essence, about a component vision of the world. The idea is that any application can be broken down into applets or component parts.

So a system that needs, say, foreign exchange rates can pull in a small piece of software from elsewhere on the web that delivers a specific currency price in real time at the point that a user needs it. The applet will be there as a ‘web service’.

“We are taking our .Net vision as the base layer and MBS is working on a business framework, or the idea of writing web service components that perform functions, such as opening a new customer account, or printing an invoice,” Smith explains.

The idea will be a kind of Legoland build that will enable MBS systems integrators to point-and-click an application together for vertical markets. If it succeeds, it could transform the business systems landscape.

In parallel to the R&D efforts of software vendors, the hardware community has been working on rolling out 64-bit computing platforms. This might not mean much to the lay person, but the key thing about 64-bit computing is that it removes the internal memory constraints of 32-bit servers.

This means that vast databases can be loaded into memory and it makes complex, multipart queries of business information easier to resolve.

As businesses push for better return on investment on their existing spend, implementing business analytics and business intelligence projects are a fast way to deliver this ROI. All the leading vendors now offer business analytics that aim to compete with specialist business intelligence implementations. There might not be much new R&D here, but there is certainly loads of project work to be won by vendors and consultants over the next few years.

Another crucial development that has had its roots outside the world of the package suite vendors is XML, a protocol for exchanging data between incompatible systems. Basically, all the major vendors have come under pressure from their clients to get their software to open up. The thrust of XML initiatives will be to explode the integrated business suite approach altogether. When every system integrates easily with every other system, there is no need to buy every module from one vendor.

A similar case can be made for Microsoft’s web services vision. When packages are entirely reduced to plug-and-play components, it is easier to add in different components.

It is possible to imagine a smart computerised wizard building a treasury system for a company from scratch, using a directory of components, with no one vendor getting a decent purchase price for the finished product.

That would be a total transformation of the market and a real empowerment of the user – to the detriment of the vendor.

We are, of course, years away from this kind of vision becoming a reality, but any view of the future of the business systems market needs to keep an eye on these trends.

Related reading