Isn't there an initial investment in hardware?
That depends on what you are running on right now. You could deploy this on the same hardware you have right now. Or you could deploy it on additional hardware, if you wanted to do a safe migration.
Let me put it differently. The hardware costs, when we measure today, to run a full front end, I think we measure single dollars per user right now. It's completely immaterial when we are talking about saving investments in integration services that are $2,500 a day.
FOR MORE INFORMATION:
Featured Topic: What is NetWeaver?
Article: Yankee: NetWeaver can cut TCOBut that's an important piece, right? Because most likely a user is going to have a non-SAP application when he buys SAP's application.
Yeah, but we didn't want to tax every user. So we could have brought this [XI] in, but then we'd have to raise the price for every user.
And the third model is if you want to buy a CPU model. You can actually buy NetWeaver by CPU, which really goes to a large, unnamed community. I want to put out a portal site with shipment data from all my partners. I have 100,000 partners, but their information is coming in every three weeks. I don't want to pay for 100,000 named users. So I buy two CPU licenses. I create the site. I create the portal. I connect it with BW. I put it out there. I need only two CPUs. Then, if they start using it more, I may buy more CPUs down the road. We'll also see that CPU model for ISVs that decide to build 'powered by NetWeaver' solutions.How do you license NetWeaver?
You have three ways that you can get your hands on NetWeaver. If you are a mySAP licensee, a seat of mySAP contains a seat of NetWeaver.
The second way is that you can actually go out and buy a user seat for NetWeaver. Let's say you have R/3, and you don't want to upgrade to mySAP, but you do want to extend your R/3 solution. You've got your R/3 system and [you've] decided to go out and build manufacturing self-service on your own, something that gives you visibility into the shop floor. You can actually buy NetWeaver seats and go and build. Those users get access to the portal, to XI, to BW, to KM -- all the pieces, basically.Is it all the pieces, really?
There are two pieces that have separate prices. One is Master Data Management (MDM), which has an engine price, which comes in three volumes. It's like a T-shirt, in small, medium and large, depending on the number of objects you manage in MDM. The second piece is XI engine. You only have to pay for XI engine for integration with SAP and non-SAP. If you're doing integration of SAP to SAP instances, you don't need to pay for the XI engine. That's only if you are doing integration of SAP to non-SAP. NetWeaver arrives at a time when people are saying, 'We have no money.' And you're talking about a long-term investment, a strategy that pays off after an initial purchase. How do you overcome that?
We don't expect customers to come in and do a big bang. This isn't mainframe to three-tier client-server, where you threw away the mainframe and then started fresh. This is an evolution, and we expect them to take years as they go from one process to the next, but they go at it from an architecturally sane point of view.
Tomorrow morning? No And now how far to you have to go?
This is the announcement for the rest of us. You'll see things like our installer, something we call the NetWeaver Rapid Assembler. You take a DVD and you put it in your system, and it installs the knowledge management. It installs the portal, and it installs the app server. If finds the ERP system you have and connects them all -- in a half hour.
It comes ready to go. And when you are done, you have a portal that runs your CRM information in a portal, and your ERP in a portal, and your users have been moved into NetWeaver, and everything is ready to go.
You start by doing that, and you get immediate benefits. This is sort of the miracle. You don't expect that from SAP. You can put in a CD and everything runs. We expect that miracle will open the Red Sea for the crossing. You see something like that and it's very powerful. Before you know it, you have generated 20,000 lines of Java code -- and you do it in a minute or two. These small miracles get you through more and more understanding of what NetWeaver does for you.Then you will become an evangelist. But it takes time. It's not something we can do overnight. Eventually?
Eventually, we believe, everyone will want to go to the business suite. We're programmers. We're engineers. We're optimistic. You won't start a programming business if you're not optimistic. But the reality of the business is that customers move at their own pace. We still have customers running R/2. It doesn't happen overnight. Some will do it really, really early on. And some are waiting until the train has passed them. We heard that the other day that you were talking to an audience, and you asked them who had heard of NetWeaver. Everyone raised a hand. Then you asked who understood it --and three people raised their hands. Today we're hearing from you that customers are getting the message, that you're being understood. Which is it?
I think what we hit so far are early adopters. I think the early adopters get it. They see the value and they demonstrate it. They were willing to jump in, pick up the pieces and get it going. Why has this been such a tough education process?
You always see it in major shifts. I mean, you will go back five years from now and you'll look at this date, and you'll say, 'This is as big as a shift that happened between mainframe and three-tier client server.'
When we announced R/3, we shifted a technology foundation. We shifted an architecture. Most people didn't understand that this was [an] industry event. It defined the whole industry for the next 10 years. PeopleSoft copied that model. Oracle copied that model. Everybody copied three-tier client server after that. It took them three or four years to figure it out, but they did go that route. Now we're doing exactly the same thing. Now everybody is arguing whether we can be successful or not, but we see this as the next architecture for the whole enterprise.
If you look at most of the enterprise application vendors, they don't have a platform that is tightly integrated into the application. They have a platform that is either not open, not broad enough, or they don't even have a platform. If you look at Siebel, they need to build to three different platforms, which is the ultimate definition of insanity in this business. Or if you look at Oracle, they have a platform and applications, but they are not built one on top of another, so you don't get the benefit of TCO that comes from putting the two together.