SAP has remodelled itself as an open company, promoting interoperability and a service oriented architecture. But is providing rivals with an entry to its customer base a good idea? Angela Eager reports.
The SAP of today is a different company to the one of five years ago. Where it was once an applications company it is now deeply embroiled in the middleware space, re-architecting in order to move to a service oriented architecture. It has also learnt a thing or two about marketing and communications, enabling it to better handle the introduction of new technologies and concepts such as NetWeaver and Enterprise Services Architecture (ESA).
"A few years ago people questioned what SAP was up to when it was talking about NetWeaver and platforms and architectures because it was outside the applications space," says Graham Kingsmill, UK MD.
Graham Kingsmill, SAP UK MD
"The reality is that as things develop the application space becomes more open and the pressure to deliver more and more for less and less is there. So you have to continue to evolve and ask, "Where is the business coming from in the future?" We intend to sell to many more organisations and have created a pick-and-play type environment where you do not have to buy everything up front. You can buy as your demands grow, so agility around the platform is there."
But is SAP taking on too much? Last year and again this year the company decided to sacrifice margin in order to push forward with its R&D; program. Yet Oracle maintains that developing applications, tools and infrastructure all at the same time is not a viable strategy.
SAP is not denying the effort but is confident that it is manageable and necessary. "Are we biting off too much?" asks Kingsmill. "It is certainly not too much, but we are biting enough and as we see more opportunities unfold and more demand from customers we will take that on board."
Although SAP is in an enviable position, with impressive financials and an upwardly mobile market share, it has had challenges, not least getting partners and customers to understand what it was doing and why. Early on there was less belief in the vision and more suspicion that NetWeaver was just a ploy to drag more dollars from SAP customers, but that has changed as the strategy has unfolded.
"SAP has spent time on articulating its vision. Prior to that NetWeaver was a collection of technologies and it was not clear how it fitted into the vision. Now ESA has put it into context. There is a lot of awareness and people understand where it is going and are thinking of it from the viewpoint of an upgrade strategy," says Melvin James, enterprise client director at Diagonal Consulting, a company that offers both SAP and Oracle services.
SAP is fully committed to the concept of an SOA, encapsulated in its ESA design. It is enabled through technology such as xApps composite applications, the existing NetWeaver process and integration platform and its evolutionary successor the Business Process Platform (BPP). This is due for delivery late this year and is touted as a complete application and process development, deployment and management platform. NetWeaver/BPP is core to SAP's long-range plan to transition to an SOA, based around the blending of fine-grained business processes, as opposed to large rigidly structured applications.
By the end of this year SAP plans to offer an Enterprise Repository with services that users can access to create composite applications, plus about 1000 pre-built composite applications. The aim is also to make an all-in-one offering available that includes the generic BPP platform plus the my SAP applications suite. By 2007, the ESA road map should be complete.
The company is also talking about building industry-specific business process platforms.
December 2005 saw the inception of SAP's Industry Value Network for Banks, which is a verticalised approach to the Business Process Platform strategy. In mid-2006 SAP hopes to create a repository for banks that encompasses the most relevant processes, coupled with guidelines for creating a flexible environment. It plans to release a BPP for each of the 28 vertical industries it supports, to be released over the next year and a half.
SAP is able to accommodate this development because it has a platform-based strategy, according to Kingsmill. "While we have a vast array of product capability, the architecture and platform allows customers to understand them. We do not do everything, so we do a lot in the context of legacy solutions and specialised applications. We have to have a platform on which everything comes together."
The idea is that the platform is not a future concept but something that is put down at the outset, on which everything comes together: legacy, current and future applications.
Everything SOA, and therefore SAP is about, relates to the concepts of openness, flexibility and agility. In the SOA world, in theory, a customer will be able to swap out a SAP component and replace it with a module from a friendly value-add SAP partner, or more dangerously with something from competitors in niche areas or from rival Oracle.
By simultaneously opening up and chopping down the size of its applications, SAP is creating a vulnerability. It is providing competitors with an entry point into its customer base at a time when small incremental sales are core to the company's revenue growth.
SAP's SOA dilemma is how to offer pick-and-play interoperability without undermining its own business. NetWeaver is an acknowledgement of this danger and an attempt to belay it. If SAP cannot prevent its customers calling on non-SAP applications, it can at least control how it is done by providing the platform to orchestrate everything and use it to add value such as the planned vertical BPPs.
Furthermore, if the company provides the middleware stack it can optimise it for SAP modules, subtly putting rivals at a disadvantage without being seen to decry interoperability. It is essentially a question of platform lock-in and Oracle is in exactly the same situation.
SAP is adamant that its commitment to interoperability is deep rooted. "We have not locked ourselves away in terms of the architecture, saying IBM is wrong and we are right. If you want to run SAP on that architecture you can. When you look at the NetWeaver stack you can see clearly that it has the ability to support other middleware, other architectures, in the same way, unlike Oracle where they clearly push customers towards one database environment," says Kingsmill.
As far as the issue of optimisation and performance is concerned he is no less adamant. "It would make no business sense in the short term to do that," he says, referring to fears that there could be a performance disadvantage when using a third-party stack. SAP's relationships with Microsoft and IBM are two of the strongest it has. "The reality is we are trying to develop an ecosystem that gives customers choice. We have an opinion in terms of recommendations but at the end of the day it is about giving customers choice."
SAP focuses on the benefits that can be gained from an all-SAP stack such as the ability to offer one-off customised development projects for customers, while still guaranteeing upgrade compatibility within standard maintenance terms. This is just one example of what can be achieved through the use of a platform that accommodates everything.
"The platform is something that has got to accommodate everything. It is not the old idea of a platform: a proprietary platform where everything fits together providing everything you use has our label. The new architecture stack caters for the past, it caters for the present, and also the future," says Kingsmill.
There are other benefits too. As Paul Smitherman, SAP client director at Diagonal Consulting, points out, those who buy into the SAP stack will benefit from embedded functionality such as portals and analytics. While this can be beneficial it could generate internal conflicts within the customer base as SAP's integrated and embedded functionality benefits are lined up against entrenched applications.
Market consolidation has had an impact on SAP. The decision to increase spending on R&D; in order to bring ESA and its supporting technologies to market earlier than planned was a direct response to Oracle's PeopleSoft and JD Edwards' acquisition and its Fusion plans.
But the company is playing a long-term game where potential benefits are concerned. "Customers feel a bit duped. They are genuinely taken aback by what has happened. They may not change their allegiance but they are nervous," says Kingsmill.
Organisations affected by consolidation are reacting in three ways: some are choosing to make a change, others are remaining where they are, but around 60% are taking a wait-and-see stance. SAP estimates that an enterprise business system relationship lasts about 10 years and says that the wait-and-see organisations will make their vendor decision over the next five years. That is the conversion market it is after and it is prepared to wait for the opportunities.
Another of SAP's ambitions is to expand its customer base from 30 000 to 100 000 by 2010, and there is only one way to do that: succeed in the mid market. The company does not have a good reputation in this sector and there is a lack of trust: the one attribute SAP recognises as vital in securing enterprise level deals. Nevertheless, the company thinks that this time things will be different.
Diagonal's James confirms that there is a high level of investment in the SME market and that SAP is gaining traction. One of the differences is that SAP is approaching the mid-market from a global perspective. Previously, SME initiatives were country based, which meant the local subsidiary was responsible for maintenance of the package, which led to problems and made it look like SAP was dabbling in the market.
With a global operation Smither-man believes SAP can get the necessary volume. However, David Pinches, director of accounts and ERP at SME-oriented Sage UK, disagrees and thinks the global aspect will be a weakness. Referring to Business One, he says SAP has struggled because it was designed as a global product, whereas in Sage's experience local design and customisation are important.
SAP also has to learn how to manage the channel, but according to Pinches this is not something that can be rushed. "They need volume in the channel, hundreds not tens of partners. We have the partners others want," he says. "SAP went for 30 to 40 of our partners but the majority failed to take off or sold off the SAP application business and kept Sage."
David Pinches, Sage UK director of accounts and ERP
Pinches identifies one of SAP's biggest challenges, pointing out that the channel is based on the software replacement model and "it is not as easy to sell SAP to Sage customers. It is a tough market to turn channel partners. It has taken three years to get [Sage's] Line 500 partners to add CRM to their portfolio and we know them and have integration," he says.
SAP is in a strong position. It has the advantage over Oracle in terms of clarity of vision and deliverables, which is being demonstrated in its ever improving financials and increasing market share. However, the uncertainties it is facing are serious enough to prevent SAP becoming too comfortable, as evidenced by the company's decision to maintain its guidance through fiscal 2005 despite rapidly rising revenue. Only when it announced end-of-year prelims did SAP say it expected 2005 software revenue growth to come in at 18% against a forecast of 12% to 14% and total revenue to rise by 13% year on year. However, a lot is expected of it. Despite what looks like an enviable position, SAP is staring risk in the face.