Enterprise applications work! That is 2003's key message. If you read the doom and gloom in the trade press concerning failed implementations and financially troubled packaged software vendors then it is probably a surprising message. Nonetheless, our research with users suggests that enterprises of all sizes have plans in the pipeline to extend existing or implement new enterprise applications. The reason is simply that these users face challenges that mean they must drive new business initiatives through their organisations. Enterprise applications can be a great enabler of change.

Despite this good news, 2002 was a tough year for the Enterprise Application Vendors. Their revenues fell, in Europe and the UK, as well as globally. It has become a buyers market. High discounts on software licenses are available.

Three general trends can be seen:

" Businesses who bought Enterprise Resource Planning (ERP) in the 90's are returning to the market. Now they have digested the initial deployment they are ready for the next stage of exploitation.
" Enterprises can become more agile by equipping staff with mobile computing and wireless connections or by tracking products with automatic identification tags. They want to link this to back office systems.

" Many mid-sized companies today still have simple stand alone departmental enterprise applications for corporate accounting, manufacturing etc. They now want to integrate these into a single system.
Overall, SAP remains the dominant player, both nationally and globally. Recent merger and acquisition activity in the following pack of vendors has strengthened SAP's hand in the short term. The familiar names of J.D. Edwards, Oracle, Peoplesoft and Siebel are contenders for the remaining top five positions world-wide. In European manufacturing, the Europe based vendors, Baan, IFS and Intentia are also among the top players.

Figure 1shows the market shares of these vendors in 2002 Source: Cambashi

Enterprises extending their system behave differently today than in the ERP boom of the late 90's. They want to extend their ERP backbone from accounting, distribution and manufacturing into new areas like sales and marketing; procurement; and joint planning with suppliers and their distribution channels. In the 90's, the primary driver was replacement of systems that would not cope with the Y2K change. Implementation was often driven by IT departments and an objective was minimal changes in operational procedures.

Even these limited objectives enabled significant operational improvements. The new systems processed transactions much more efficiently than the legacy applications they replaced. The enterprises' managers could now see immediately an up to date and consistent financial view of the enterprise. ERP provided one user interface, one database and one set of business rules. Before, much management time was wasted waiting for figures and arguing about what the figures really were rather than on deciding action plans.

There were well publicised implementation failures. When we interviewed users who had problems with their ERP deployment we often found that the key reason for failure was that management had failed to consult widely enough. The enterprise did not have one set of business rules. The ERP deployment exposed this and problems had to be resolved on the fly. In those enterprises where everyone's common purpose is aligned that was simple. In other enterprises, politics could disrupt everything. Not for nothing is the spoof explanation of the ERP acronym Early Retirement Probably (if I have to put this system in).

ERP systems simply keep the score, managers are paid to improve the score. Today, managers have identified business initiatives they want to drive through their organisations that will improve processes and generate profits. New Enterprise Application deployments have to support these business changes.

There are many differences today from the purchases of the 90's, in the way the systems are bought and in the nature of the applications chosen.

Today, buyers are more educated and more cautious in planning an enterprise applications deployment. Rather than leaving decisions to the IT department, many managers make the buying and deployment decisions. As a result, the process is lengthier but better planned. The chances improve of meeting project objectives.

The IT department does have a significant role in limiting or even reducing the on-going expense of new hardware and licences. There are many recent advances in computing infrastructure architecture and the tools to manage complex data centres. Until now, Enterprise Applications' use of infrastructure has been profligate with redundant systems for development and back up. Often systems software licensing is on a per server basis. A server and storage consolidation project using new technologies can often recover its investment costs in under a year.

There are typically two objectives for the applications chosen:

" Extend the visibility of operations both geographically and to new departments.
" Provide assistance to managers trying to decide between options to optimise the resources.
These two thrusts are linked. There is a danger that a manager's optimisation decision becomes a sub-optimisation if they can't see the overall effect of a decision on the enterprise, because enterprise applications are not connected.

Few enterprises have completely rolled out a single ERP system into every department and location. In some departments, such as sales and marketing, there is little computer assistance beyond a contact manager and email. In others, such as logistics, existing stand alone applications are only loosely connected into the ERP backbone. Sometimes, mergers or acquisitions have brought in different applications that conflict with the existing deployment.

The debate between choosing an "integrated suite" and "best of breed" continues. The integrated suite vendors insist that the advantages of full integration outweigh any shortcomings in functionality. Actually, I can't recall them admitting any shortcomings in functionality but that's not what their users tell us. The "best of breed" vendors insist that their focus on the needs of specific industries or specific roles in the enterprise permits optimisation and progress in driving particular business initiatives. In the longer term, object oriented systems and standards might make these choices obsolete. In the short term, each user has to make a trade off decision. Where a process is critical to an enterprise's value added it may well need to use a highly customised or "best of breed" solution. Fortunately, the new generation of Enterprise Application Integration tools have greatly improved the interfacing between these applications and the ERP backbone.

These issues are particularly stark for those enterprises that can see advantages from adopting new forms of computing technology. Different industry sectors have different opportunities:

" Food processors can ensure retailers don't run out of stock during trial promotions if workers in shops use tablet PCs with wireless connections to update stocks;

" Retailers can automatically re-order when a high value item is checked out at a Point of Sale Terminal;

" Process Plant Operators who monitor heat exchanger performance on-line can let a fitter realise that it will soon need maintenance attention and order spare parts before a failure;

" Construction companies can reduce idle time if they upload progress from a PDA on the building site and dynamically reschedule deliveries;

" Pharmaceutical companies can use Radio Frequency Identification (RFID) tags to trace particular batches of drugs without waiting for medical staff to write up notes

What all these initiatives have in common is that they depend on back office systems being synchronised with what is happening in the field. You can do this with almost any system but it often won't yet come out of the box.

In a recent interview by Douglas Haward of Computer Weekly, Henning Kagermann, Chief Executive of SAP was asked "How important is wireless technology?" and responded "It's still just taking off. There are problems with too many devices, standards not being there, and networks charging too much for airtime." When asked "What about radio frequency identification (RFID) chips?" he replied "Some clients are using this technology, but there aren't yet products that we can deploy to 20,000 customers. Maybe in two years' time."

There are specialist applications for all these initiatives and many more. Rather than wait many users are deploying specialist enterprise applications and using systems integrators services to make a reasonable integration with their ERP backbone.

Middle sized companies face most of the challenges of their larger brethren. They simply have less resources to address those challenges. As there are often fewer levels of management, individual departments and business units are left alone to manage by overall financial objectives. As their markets globalise and become more competitive, they need to seek out synergies between formerly autonomous units: for example, setting up common procurement between different locations to obtain better prices and deliveries. Enterprise Applications are more attractive and Enterprise Applications software stands ready to help.

Those middle-sized businesses that have strong balance sheets are showing signs that, in mid-2003, they are more prepared to invest than larger companies. These companies have managed to contain costs and are making small profits. The management teams have gained confidence that they have controls, which allow them to respond to volatile demand. They want to rebuild margins and recognise that better enterprise applications will help them to do so.

The problem for these companies is to find partners to help them with the deployment. They perceive big management consultants and international systems houses as too expensive. They can't get attention from the Information Technology vendors direct sales forces.

Many Information Technology vendors, both infrastructure and packaged software, have tended to focus on large enterprises - partly because they sell with an expensive direct sales force. From the vendor's point of view, large probably means at least £500m in sales. Size is not the first characteristic industrial users think about when classifying themselves - it is more likely to be what they do, or the market they serve. An enterprise with £50m in sales will almost certainly have some multi-national operations and will regard themselves as a decent size business.

Today, quite a few of these businesses have not yet been served by the traditional Enterprise Application suppliers. They will probably need help, and value added resellers may be the route for that to happen. In 2003, Microsoft Business Solutions is beginning to develop its product and reseller network from the Navision and Great Plains businesses acquired recently. The solution set is evolving from basic accounting to a wider range of enterprise applications. For example, they recently introduced a Customer Relationship Management application. If Microsoft develops an ecosystem around these resellers and its products, then we can expect to see another round of explosive Enterprise Applications growth.

Mike Evans

First appeared in Conspectus in Sep 2003

Other articles from Cambashi:

Engineering Applications in 2003 - boom or gloom?

back to top