Vortex Blogging Day 2

I had some time to reflect on the content and conversation I had yesterday, here's a quick rundown:
1) unanimous agreement that IBM is the best positioned vendor (except for Schwartz, of course)
2) agreement that the stack as Moore lays it out makes sense, but also agreement that situation on the ground gets a lot more complex than a powerpoint slide
3) most people I spoke with liked Schwartz's presentation, however upon further questions it appears that they like his style and command of the facts. At any rate, he alone did not change their view that Sun is the first vendor to get voted off the island.
4) the vendors are not answering the tough questions to the satisfaction of the audience, the conference attendees really want the vendors feet held to the fire
5) telcos are a wildcard in the landscape

IBM

- their pitch is really straightforward, have a robust infrastructure that works, open standards based integration, and services along technical and vertical market lines that helps their customers be successful.
- IBM's microprocessor business seems a little out of place in the IBM of today
- good question: is on demand something I do or some set of services that I buy? Answer, IBM wants to help their customers be on demand (huh? so is it something I buy or something I do?). Actually, he almost answered the question, it's a services opportunity for IBM. Surpise, isn't it.
- their position against M$oft is that they support open standards, hetergeneous systems, and most importantly, everything works. Not a bad position.
- heaping praise on Cisco... I do think that IBM and Cisco have an unspoken agreement to steer clear of each others turf.

Q&A

Q: (Geoffrey Moore) yes it's a heterogenous world, yes it's complex, but it's too damn expensive. Rational, Tivoli, Websphere, Lotus, etc., "you guys are making it more complex, I want you to design out the complexity". What are you doing?
A: consolidate the complexity in storage and servers using our Bladeserver.

Q: follow up Q, is there a master architecture you are gravitating toward
A: if there is a master, it's the one that you are presenting.

Q: Chuck Phillips said that customers aren't interested in SOA
A: customers like modularization. No customer is saying they like huge monolithic architectures (good answer). SOA is the only credible way to build modularization.

Q: It looks like Oracle-PSFT deal is going through and represents a land grab for customers. Your strategy is not to compete with your app partners, does that strategy have to change?
A: applications are a fragmented business, that's why we don't want to get into that. By providing the integrating fabric that manages and interconnects, we'll have a better chance than going out and building or buying application businesses

Q: promise of grid computing includes a lot of things, what is IBM's core strategy around grid and single most element of value?
A: virtualize the infrastructure in order to hide complexity.

EMC

- as a side note, they sent a marketing guy while all of the other companies sent some heavy hitter executives. This is disappointing.

It's a bunch of marketing soundbites, so I'm not going to invest the time in covering it. Going for coffee instead.

okay, I have to chime in here because the EMC guy just made the comment that customer don't want to go back to the monolithic stack (parroting the remark that Irving made). But then he went on to say that customers want everything to work. Isn't is amazing how the more things change the more they stay the same, I could have heard this pitch back in the best-of-breed vs. integrated suite debate 5 years ago. Yes of course customers want stuff to work, that's a no-brainer. They also don't want to be locked into a vendor because their experience with that is that they don't always get the things they want at the price they want it at, but they do get things that work.

[rant] forget SOA, forget open standards, forget networked applications. What I hear customers saying is that they have too many servers in their data center, they feel that they are losing control of how their users take advantage of IT, and they are still spending a bundle on professional services to make things work, and even when it does, it's a fair bet that it won't work as well as they expected and it will be a lot more expensive than they planned on. But the truth of the matter is that all that complexity isn't going to go away because enterprise customers are expecting their IT to do more all the time, so vendors should be looking at ways to make that complexity more predictable and easy to digest, as well as more predictable from a cost standpoint. [/rant]

Q: (Geoffrey Moore) EMC has earned a place at the table, and ILM is going to give Veritas fits. But the criticism is that this is the bridge too far, other vendors will tolerate you but eventually try to design you out. I just don't see how you can keep your place in the market?
A: We're focused on doing what we do really well within the stack, we're not looking to expand outside of our layer in the stack.

The Future of Enterprise Software

Dave Kirkpatrick, Fortune Magazine, is moderating
Shai Agassi, SAP
Dan'l Lewin, Microsoft
Marc Benioff, Salesforce.com
Marten Mickos, MySQL
Irving Wladawsky-Berger, IBM

Mod: Dan'l, explain what you do in relation to MySQL and Open Source
DL: other than interoperability, probably not a lot. It depends on how you define open source aside from a licensing model.
MM: mysql is a bridge for Microsoft into the open source world due from our support for Linux. If Microsoft really want wants to compete against LAMP, they should come out with WAMP.
Mod: would SQLServer support Linux if they were on their own?
DL: good question, but there are a limited number of things that we can do irrespective of any groups focus. Wouldn't rule it out, but won't comment on it either.

Mod: Irving, how do you interact with SAP?
IW: in some cases we compete, but we are friends and work incredibly closely with a range of companies. A huge amount of SAP's business is on IBM servers and DB2.
Mod: IBM's deal with PSFT, is that part of their defense against Oracle?
IW: we really want as many apps on websphere and our middleware
Mod: but clearly that deal would not have any life after an acquisition by oracle, how can it be seen as anything other than a strike against oracle
IW: it's a straight business deal, we don't know what will happen with the acquisition but this is part of our strategy

Mod: talk about software as a service. Can you work with Marc's company?
SA: we have an understanding of the cycle across a significant number of processes and deliver it to your company. There is a strategy to deliver it as a hosted service, but it's your environment. Marc's company delivers there predefined processes to your company, and for SFA there are a number of processes where that fits. I don't think that fits the broader market. We believe there is a requirement for a tightly linked application environment with applications that enable flexible applications that are right for your company.
Mod: (to Marc) IBM calls it on demand to the customer, how does Shai's desicription apply?
MB: The internet itself is becoming more capable and persistent, that is the platform. When SAP began building R/3 the platform was client/server. Salesforce is not an application, it's a representation of what you can do with the new model, the internet model. Our tools are not constrained by what we build into them, you can link in Microsoft tools or IBM tools. We focus on integration through web service protocols.
SA: if you look at where Marc has more impact, it's Siebel. Our customers are looking at solving an end-to-end problem.
MB: the internet is providing a platform for propogating web service onto heterogenous devices and legacy apps. You have to buy into the idea that this is a new platform, a new philosophy for applications.
Mod: you are using the idea of the internet platform much more broadly than it has been used historically.
IW: but that's where the new integration standards are going, working towards virtualization. What processes virtualize nicely is the question, we don't know but the market will figure it out.
Mod: Irving, at that point how will a company like SAP transform?
IW: SAP has been working toward services, so at any time they can deliver it.
SA: we see what the customers want. Customers don't go around buying single web services from disparate sources, they want solutions that are assembled using web services. There are 4 different approachs, Marc's is a collection service, Microsoft has dot-net where they are building a platform, IBM is uilding the technology and won't build the applications, and finally SAP is saying we have the objects and we're putting integration technologies in place to pull them together.
MB: each of us has the idea that our platform is what the customer will want
Mod: is that a viable approach that so many companies believe this?
MB: yes, because it gives customers choice and lowers their cost. The lowest cost, lowest risk approach will win, that's why Microsoft is here.

MM: as a representative for the open source community, we have developed a lot of software but it tends to be very technical so the end user is probably not the same customer that these other companies are selling to.
Mod: how does the macro issue of open source relate to these companies
MM: open source is the #1 laboratory for development, but it's not the #1 channel for delivery.
DL: what are the leading platforms that have come out of the open source community
MB: (something about Apache text searching)
IW: all of the protocols are build as open source reference and vendors can build it into their products
DL: but the academic community has always been a lab for R&D

Mod: Yesterday, Shane from HP was talking about HP's platform. How many on this panel believes that?
(nobody really agreed, although Dan'l made a good point about the customers will define what their platforms are).

Q: so you are all platforms and all have great futures, who is going to lose
IW: the IT market is large
MB: the customer is going to be the ultimate winner
SA: crunch happens at different levels, in our business you saw the best-of-breed vendors collapse. It's not a zero sum game, but forces will work to shift the market and innovations in it

Q: how do you simplify interconnecting, not just integrating
SA: each of us collaborates on standards, we're trying to take away the pain of customers using their data in different applications. We can't work on those platforms because of support and economics.

Q: (geoffrey moore) it's all well to say that it's great, but customers are not saying that it's great
SA: there is a delta between solving it in the lab and deploying in the field
MB: our customers all upgrade when we integrate
SA: but if customers build something on sforce and upgrade, their customers don't necessarily upgrade when they do.

Q: expand on the differences between Salesforce and SAP. Marc, do you believe customers will get ERP and financials through your model?
MB: yes. This is the commodization of enterprise software. For a lot of companies the client/server did not work anyway
SA: there are buying patterns for salesforce.com, risk averse tactical solutions. The thing is that SAP does do large and very small solutions. It may be boring, but the model of utility computing is in fact boring but it's profitable boring.

Gotta go, have to be at a board meeting in LA at 2:30, so no more Vortex blogging today.

Comments

[Nick Gray]

Cool notes, thanks - Read them on my Palm during a lunch break.