racle has in recent times not so much been throwing off the old to bring in the new, as hanging onto the old while adding the new. It has considerably changed the shape of its business and is now presenting a new face to the world.

The company is faced with numerous challenges, but they all relate to one fundamental change: the transition from an infrastructure technology provider to an applications provider. The business of selling and supporting enterprise business applications is very different to its former bread and butter, as database and middleware software provider.

Yet Oracle seems to be getting to grips with the change, both operationally and philosophically. In EMEA it has close to 4,000 staff dedicated to business applications, representing a significant expansion compared with three to four years ago.

This is because its sales approach is increasingly industry and value-based, because the mantra for the next generation of applications is fairly simple: control the platform and you control the process; control the process and you win the business. The vendor that controls the process has the biggest buy-in from the business and wins influence around that. With a database, middleware and application stack, Oracle is pitching for optimisation and process control.

Oracle has changed in other ways, too. The launch of its Applications Unlimited program, which is a commitment never to shelve a viable application while customers are willing to pay maintenance, indicates that it is starting to prioritise business needs over technology trends. "For the first time we are not forcing an upgrade," says Sergio Giacoletto, Oracle's executive VP of EMEA. "Traditionally software companies forced upgrades on the technology cycle, not the customer cycle. This is a significant internal change."

Oracle is also pitching at a different level. "As a result of the acquisitions we have taken our presence within corporates higher," says Ian Smith, regional senior VP at Oracle UK. "We were database critical but not business critical. Being mission critical means you come in at a higher level. You are within the radar of the board, where previously we were viewed as a database company."

Enhanced reputation
Externally, Oracle is also being viewed differently. It has had a reputation among partners and customers as being difficult to deal with but this is changing. Oracle has come around to the concept of customer and partner feedback, including increasing use of customer advisory panels, which is a move that has accelerated as a result of the acquisitions.

"It has reinvented itself," says Keith O'Hara, Oracle practice director at Diagonal Consulting. He attributes the change to the shift from technology company to applications vendor, and the shift to higher ground in terms of how to sell, and to whom.

The message has undergone a fundamental change too. Oracle is no longer pummelling a straightforward 'single suite' message. Now it is more of a best-of-breed approach, with a modern twist. Where traditionally best of breed has been about providing a broad-based monolithic application suite to serve all business functions, the predominant trend today is a service-based middleware platform designed to make it easier to integrate third-party applications with the ERP backbone.

Best of breed has gone beyond the application to become a platform-level issue. As Giacoletto points out, with both Oracle and SAP having mastered best-of-breed applications and suites, they are now creating underlying SOA-enabled platforms: "So the question is, 'Who will have the best platform?"

"The acquisitions changed the position and the message," says O'Hara, from one where Oracle was preaching a single-source message that was pitted directly against SAP, to more of a best-of-breed stance across a wider area. "SOA is a modification, a change of tack, not an about-turn," he adds.

It is not about offering isolated best-of-breed applications, but applications that fit together, while also maintaining a single version of the 'truth'. "Oracle has the tools to make the single-source message work, even though the functionality is best of breed. The technology was not there a few years ago," says O'Hara.

A critical part of that technology is SOA-based middleware supporting service-enabled applications. If a vendor has its own SOA-flavoured middleware platform, it can offer an extensive suite while also offering users the ability to choose best-of-breed applications.

But there are problems with this approach - older technologies need to be SOA-enabled to integrate with SOA-based systems, which requires work - but what the new style best-of-breed approach does offer is the ability to add and replace, delaying the forced move to an entirely new system.

SOA-based middleware is the entry point to next-generation applications: Fusion Middleware in Oracle's case. So is it also an early indicator of enterprise application vendors' future success?

It is not clear-cut, according to Giacoletto. When it comes to buying intentions, industry specifics can be more important that the middleware, he warns: "Core ERP functionality is standard but there is a lot of difference in the vertical systems. These decisions are based on big functions and business processes. That is why we are increasing vertical functionality."

Business need drives choices, Giacoletto maintains, and for some that results in the prioritisation of the application, while for others it is the middleware. For manufacturers, application functionality is the primary factor rather than the middleware, while the banking sector tends to make decisions based on the middleware because of the need for extensive integration, according to Giacoletto.

As far as Oracle is concerned, the quality of the application and provision of vertical functionality are the deciding buying factors, and now that its major horizontal acquisitions - PeopleSoft and Siebel - have been concluded, it is skewing its acquisitions towards vertical players such as Retek and ProfitLogic for retail, G-log and 360Commerce for supply chain and logistics, and i-flex for banking.

With its applications strategy focussed around the retail, banking, telecom, life sciences and public sector verticals, the company's acquisition strategy is to build a comprehensive portfolio of applications for each industry. Its view encompasses core horizontal functionality available across multiple industries, plus a higher level of vertical functionality, Giacoletto argues. Despite the trend whereby some aspects of middleware are becoming part of the business application, application and middleware decisions are still largely separate, he believes.

SOA commitment
However, like a self-fulfiling prophesy, the more code bases in the mix, the more important the SOA-flavoured middleware platform becomes. Without a solid Fusion Middleware base - and Oracle database installed base - Oracle could probably not sell Fusion Applications.

Oracle is heavily committed to SOA, but SOA is just the latest attempt at distributed component computing. So why does Giacoletto think SOA will work, when predecessors like Corba did not?

"The mistake then was that the level of granularity was too small, it led to an unmanageable number of objects. Now we are talking of a few hundred functions, not thousands, to interface to. At the business-application level, mixed components, both horizontal and vertical, will continue to exist, but it will be simpler.

"For us, hundreds of different horizontal and vertical models mean we do not force customers to upgrade them all at the same time. The SOA approach is [the ability to] upgrade or swap at different levels and the technology layer is simpler and less challenging."

Diverse goals
Giacoletto believes that SOA is at the stage ERP reached 10 years ago, and that it will undergo a similar evolution from components, to best of breed, to integrated offerings. "Over the next four to five years customers will adjust to SOA, as they did to the best middleware and the best suite," he says.

One of Oracle's challenges is balancing its need to compete as a middleware vendor with its desire to use middleware to promote its enterprise applications, and lock users into its application technology stack. Giacoletto does not regard SAP as a middleware competitor because its platform is geared towards its own applications and user base.

"Our competition is IBM. BEA is struggling [although BEA's latest quarter sales growth of 15% seems to contradict that assertion], and Microsoft .Net is too grand a vision. Customers will buy best of breed, or Oracle versus IBM," he argues.

Still, Oracle's middleware platform is well established and expanding. "Some competitors argue that Oracle Fusion middleware is not ready," says Giacoletto. "That is false. It is certified and is shipping now. Over the past two years we have added identify management, business process management and activity monitoring, through acquisition and development." Oracle's ambition is to complete the portfolio by adding vertical functionality at the middleware layer.

It is certainly pushing hard with Fusion Middleware, including Fusion-enabling versions of the current generation of applications so users can be exposed to some of the Fusion architecture and middleware benefits without undergoing a major migration program.

Middleware sales may not correlate directly to future application sales, but as Fusion Apps require Fusion Middleware in part at least, they are a leading indicator. What it does allow, O'Hara says, is "best of breed with the benefit of a single suite".

There is a firm belief at Oracle that the back office is still the key to improved productivity, but that commercial businesses are lagging. Smith points to UK government technology initiatives, which include adoption of standards, shared services and processes, and are based on lessons drawn from IT companies. The key is to "free back-end costs to deliver front-end services," he stresses. "Why can't organisations see that? There are enormous savings in the back office from the use of standards and shared services that can free up resources at the front end, but you don't see commercial organisations doing that."

Oracle has invested heavily to help realise its vision. But will it work? Morgan Stanley predicts that Oracle will grow at high single-digit rates over the next few years, potentially going higher after the release of the Fusion applications scheduled for 2008. But it believes the company will need to step up investment to keep its leading position, increasing its investment in research and development, acquisitions, higher sales and marketing spending and covering rising support costs.

However, in the long term it does not think organic growth will be impressive. Based on a Morgan Stanley survey of current Oracle customers, nearly 80% expect spending on the Oracle E-Business suite to be flat, while 71% see flat database spending. Only 17% said they would increase licence spending on the E-Business suite in 2006. Some of the cautious behaviour can be attributed to uncertainty over Fusion and reluctance to invest in an old technology before the new one is available for assessment. For the next couple of years, Oracle will be fighting for essential upgrades and incremental sales while customers wait for the next generation.

While the overall market is improving, the Western European economy could be better, Giacoletto says, although there are some hot spots. Demand for CRM among telecoms, public sector and banking is rising, and following general market M&A activity companies are consolidating their finance and HR operations and applications, fuelling a market for replacement software sales.

SME challenge
Like SAP, Oracle is also looking to the mid-market for new business, although success here is anything but guaranteed. As O'Hara cautions, going into the mid-market is a bigger step than either company realises. SMEs are used to buying packages and having them work with little IT set-up, and they are unfamiliar with the notion that there could be teething problems because each implementation is unique and needs to be customised.

However, Oracle's vision appears to go beyond just selling a mid-market version of its business application. According to Smith, it is looking at how small and medium- sized businesses operate, and thinking seriously about what Oracle can do to fit in.

CBR Opinion
Oracle is in the midst of probably the largest integration and development project any software vendor has ever attempted. It is facing major operational changes as it adjusts to the subtlety of business application sales, and it is exploring how it can expand its market by targeting SMEs. These activities all have to be managed without negatively impacting its day-to-day operations, or those of its customers. Something may have to give: either the timescale, or the quality of the work. If the company wants to prove that the changes to its business are systemic rather than cosmetic, CBR believes it needs to prioritise quality.