But Henning Kagermann, chief executive of SAP, the $10 billion market leader, says new competition, changing economics and fast-moving technology are driving the 34-year-old company to rethink how it builds its software and how to sell it.
SAP is forging ahead with a new services-based architecture, and archrival Oracle is right behind the company. The big question is whether SAP's more than 33,000 customers will buy into the plan--and buy new software--this year, next year, or not at all.
At the company's annual user conference in Orlando, Fla., this week, Kagermann sat down with CNET News.com to discuss why SAP and Oracle are wrestling for services business and how Microsoft has gone from being a prospective acquirer to a respected partner.
Q: You said this morning that SAP is
Kagermann: There's never an end. Such a road map brings us to a certain point. It's like when we had a road map to bring R/3 (applications) to market years ago. That's a big investment. But later on, there are normally so many requirements coming from the market for new features that you are constantly evolving. You have to give customers new concepts, and the risk is whether you can do that, and will it work, and are you first?
Once the concepts are out, then the risk is lower. It's proven, and you get more customers. This is like the transition from client-server to a services-oriented architecture, which is a bigger step and a bigger risk.
Do you think that the move to a services-oriented architecture is a bigger step than the one taken to move from mainframes to a client-server architecture years ago?
Kagermann: Yes. This is a bigger step--because the other one was more a technology step, where you rewrite some of the application, but then they run as they did before. But SOA is really about how you build an application. It has nothing to do with the hardware underneath it. Therefore, the switch is tougher and if you apply some of these new concepts, you are never sure if it's really working, if it's not too slow, all of those things, all of those risks you have.
What about your customer base? How far along are they in moving to a services architecture?
Kagermann: Well, the customer base has to do a lot. I think there has been some introduction in the customer base. But I believe the first big wave will come this year.
SAP has been, more or less, the first company bringing those applications to the market. The first was last year with ERP 2004, but we had no services in. With ERP 2005, we now have over 500 services, so now you can really say this is ERP on a services architecture. Over time, you can imagine that we will have 1,000 or 1,500 services. But with 500, you can already do a lot.
I think customers now see that this is a step where it makes sense to switch from client-server, because they already, with these services, have 70 percent of what they need and the rest comes in smaller chunks, so it's worth doing it.
Are there certain types of companies that are more likely to be moving to a services architecture sooner? Such as telecom or financial services companies?
Kagermann: Yeah. It's more popular in companies that have less packaged software, like financial services companies. But we do something different. We are not just bringing a services-oriented architecture to the market which helps companies to connect their various legacy systems, which is sometimes the IBM approach. Our case is to also rewrite the applications, which is a significant step.
So for SAP, a services architecture isn't just an integration strategy?
Kagermann: No, it's a re-architecture of the software. Software which is designed this way looks different internally. Flexibility is part of the design.
That must impose a steep learning curve on your customers' in-house developers.
Kagermann: Yes, that's true. For the people who implement it, the businesspeople, it is not a big learning curve. It's more learning about what they can do with this new architecture. There are more opportunities, and people have to learn how to use them. But for the in-house developer, it is a big step. It's a different mentality and a different way to design software. So we'll see.
With IT staffs shrinking, do you see the people doing this sort of development within companies as the more experienced programmers? Are there new people being attracted to this sort of development because of the new technology and this new approach?
Kagermann: There are new people coming in. But companies are trying to shrink the number of in-house developers. But I believe for large companies there will always be some in-house development. There will always be some percentage of software that they do in-house. You can't, 100 percent, get rid of it. Otherwise the standard package would be too large and too complicated to meet every company's needs.
Kagermann: The hybrid approach is the key approach. You will see more applications as we move to this more sophisticated model. This model is cheaper in terms of hosting. We believe it is the right approach. It is a little more expensive, but with blade servers and all of this stuff, I'd say the difference (from hosted applications from other companies) is more that the security is pretty high.
The other models of on-demand might be better for companies with 10 or 20 employees. But since we get our customers from the midmarket and higher, it's the right approach for us. We really don't target small companies today.
So those smaller companies might still go to Salesforce.com?
Kagermann: Yes. But over time, I think we will be cheaper there as well. I think this is a larger market, and this is more effective for that. It doesn't help us if we are going after companies with two or three users.
Let's talk about Microsoft: You've gone from
Kagermann: Yes. That we figured out. We know we are too large to partner on all areas. But on the other side, we can partner where we don't compete. We've figured this out--we know we can be tough but fair in competing with them. It's a good relationship.
So no more acquisition talks?
How is the mix of license revenue to maintenance and support revenue changing over time? I know that license revenue was very strong for SAP last quarter.
Kagermann: Maintenance is largest; it has been for one or two years. License revenue is growing faster, and if it continues to grow at the same level, it could come up to the same level as maintenance. But today, maintenance is slightly higher. Services revenue is going down. It's less than 30 percent now, which is good. Services is lower margin.
So why the battle between SAP and Oracle to
Kagermann: It's not a battle for services. It's a battle for software. It's a battle for customers and over who sells more software. We sell much more. Every software deal is a competitive deal for new clients.
With existing clients, it's not so competitive because those clients go for the wider SAP strategy. It doesn't mean they give you the check immediately, but the competitiveness is lower. But for new clients, it's always competitive.
Is your license revenue increasing due to new sales and new customers? Or is that sales to your existing customers?
Kagermann: It's both. Twenty-five percent is new sales, and the rest is to existing customers, which is not surprising, because whenever you sell to a company which is already a customer, it's the installed base, and we have a pretty high penetration at the high end, so we have left to get only the smaller (companies).
Which leads to the next question: Where does the new growth come from, since you already count as customers a very high percentage of the market?
Kagermann: We have a big chunk, but there is still a large market. But if you look at the whole market, we have roughly 12 percent in North America and 36 percent in Europe, so it's not so high that you would say we have saturated the market.
On the other side, we have many new products. We believe that with these new products, we can double the addressable market over time. So whether the market grows or not doesn't matter that much, since we will open up new markets and gain market share in existing ones.