Friday, October 19, 2007

BriefingsDirect SOA Insights analysts on BEA, Google/Capgemini and systems integration

Edited transcript of weekly BriefingsDirect[TM] SOA Insights Edition podcast, recorded September 10, 2007.

Listen to the podcast here. If you'd like to learn more about BriefingsDirect B2B informational podcasts, or to become a sponsor of this or other B2B podcasts, contact Interarbor Solutions at 603-528-2435.

Dana Gardner: Hello, and welcome to the latest BriefingsDirect SOA Insights Edition, Vol. 25. This is a weekly discussion and dissection of Services Oriented Architecture (SOA) related news and events with the panel of industry analysts and guests. I'm your host and moderator, Dana Gardner, principal analyst at Interarbor Solutions. On our panel this week, and this is the week of September 10, 2007, we have Tony Baer, a principal at OnStrategies. Welcome, Tony.

Tony Baer: Hey, Dana, how are you doing?

Gardner: I'm doing great. We’re also joined by Jim Kobielus, principal analyst at Current Analysis.

Jim Kobielus: Hi, everybody.

Gardner: Hi, Jim. Also, Brad Shimmin, principal analyst at Current Analysis.

Brad Shimmin: Greetings, Dana.

Gardner: Welcome. Our topics this week are timely. We’re going to look at the recent BEA conference and some of the announcements that came out of that. We’ll also take a look at some of the issues around management and SOA, something that’s been of great interest to me.

Last, we’ll look at the Google announcement that Google applications, software, and service offerings which are closely aligned with Microsoft Office offerings will be supported by Capgemini, a large enterprise systems integrator (SI) and professional services organization, giving a bit more credence to the virtual applications suite approach.

First, let’s talk about BEA World. Let’s go to Brad Shimmin. Brad, you were there. We had some announcements about registry repository. We had some announcements around rich internet applications (RIA) and partnering with Adobe, and then also some ticklers around something called Genesis. Why don’t you give us the update?

Shimmin: Thanks, Dana. I’d be glad to. I was at the show, and Alfred Chuang was kind enough to do a keynote and to take some questions from analysts afterwards, which was unusual and very good. It was a great experience.

What I found most intriguing from the pretty big list of announcements they made was the under-the-radar partnership they did, not with Adobe, but with a company called Enterconnect, which is an on-demand portal company. They just announced a site called SOAApps.com, an on-demand software and service-based SOA environment. BEA partnered with them to provide the infrastructure for that.

So, any time Enterconnect sells a software-as-a-service (SaaS) SOA solution, BEA is going to get a chunk of that money, because they’ll be rolling out the infrastructure. What I found interesting wasn’t so much that partnership, but the way the BEA is positioning itself as an enabler of SaaS-based SOA environment. Everyone is saying they want to get on the SaaS bandwagon, but BEA is very well positioned to do that, because of what they’re doing with the virtualization software. You guys saw how they announced a new user-based licensing scheme for the virtualization software.

Gardner: Right.

Shimmin: Well, they're going to transfer what they learned from that to the stuff they’re doing with Enterconnect. So, down the road, they're are going to be able to have back-end software that’s going to know how to charge back and account for how the software is utilized. BEA is going to be pretty well positioned for that, and that leads into what you were talking about with this nebulous Project Genesis.

Gardner: Brad, just for the edification of our audience, what do you think BEA means by "SaaS SOA?" Are we talking about the ability to create services and make them available in someone else’s infrastructure, or are we talking about bringing that infrastructure into your enterprise and then being able to yourself create some on-demand business model?

Shimmin: Actually, it's both. They're an enabler, and that's in the positive sense. They have a strong history of partnering with ISVs and other vendors in space, and that’s going to continue for them within the SaaS space. You can look for them to provide the backend infrastructure that other companies can utilize in selling their wares, like they do it for GXS, for example.

At the same time, and that’s why it leads into this Project Genesis thing, Alfred mentioned that what they want to do for all the Weblogic and AquaLogic software is not only virtualize it, but to have that software such that it can be deployed in either an on-premises or a SaaS-based environment.

Gardner: It sounds like BEA, with some across the board approach to the pricing and business model, is trying to allow for these virtualized environments to be acquired by SaaS or on-demand providers with a low upfront cost, but then with a recurring revenue model and a sharing revenue model associated with it.

Shimmin: Exactly. That’s the Enterconnect partnership in a nutshell. Right now, BEA is the only company that’s thinking about those issues and put a pricing issue on this front.

Gardner: That’s a significant departure for BEA. In the past, they've been large capital-intensive, high-performance, and niche-oriented, when it came to their middleware and transactional and application development deployment strategies.

Shimmin: Absolutely.

Gardner: How does that affect them as a company? It’s a tough transition for any company to go from large upfront project-based revenue to a slow ramp-up recurring revenue model. Did they try to ameliorate those concerns amongst say financial analysts?

Shimmin: They had a financial briefing, which I didn’t attend, so I don’t know if they talked to that point exactly. In a roundabout way, they said they're certainly not abandoning that side of their business and see that as a very important thing for them in long-term. I don’t think they are going to try to switch. They're really just trying to build up.

Gardner: They can continue to provide high-performance transactional licensed product to the financial sector, government, and manufacturing, where they have been strong, and, at the same time, focus on this burgeoning on-demand ecology.

There might be enterprises that want to get into a shared services bureau approach internally or externally. More and more, I expect we are going to start seeing organizations like Salesforce, Amazon, and Workday that will provide more application services and composite services and then charge for them on a monthly user basis.

Shimmin: Exactly. They'll be charging the Salesforce.com folks to do this, but they also think -- and this is an interesting and weird thing to me, Dana -- that partnerships like Adobe and Enterconnect will be their ticket into the SMB market.

Gardner: Right. They don’t do really anything now.

Shimmin: Exactly. Adobe is now going to bundle in their application server with their development environment. They've already been integrated in the past, but they're now going to bundle up the pre-license and provide frontline support both ways. So, they have bundles on either side and each company can provide first-level support themselves.

Gardner: It sounds as if, as an attendee and an analyst at the conference, you came away not so much impressed with their product announcements and their technology roll-outs. but more with their shift on business model and their partnerships?

Shimmin: Exactly. Tangible announcements really centered on their registry repository. That's something every company is doing and it didn’t strike me as being unique, except this for the fact that their Metadata Interoperability Framework (MDIF) that they have for the registry repository is something they are trying to get partners to write to, and they already got Skyway on board.

They've have got some moxie going behind them there, but that's stuff that every company should be doing. I don’t really see that as being unique. The technological announcement they made is on a 10.3 server, which isn’t coming out until much later. So, it was more the direction.

Gardner: Before we bounce this off of our panelists, maybe you could offer what you think Genesis is. It seem quite nebulous.

Shimmin: I wish I could provide some clarity, but from what I came away with, reading through releases, and listening to Alfred, Genesis seems to be two different things.

When I read the releases, it sounded like it was going to be some totally new approach or direction. After listening to Alfred and talking to some folks around the conference, it seemed more like an extension of what they are already doing. I'd say it was a wrapper around things like their SOA 360º and WorkSpace 360º, ThinkLiquid, and other initiatives that they have. I was a little disappointed, because I feel that BEA put too much emphasis on these initiatives and they ended up muddying the waters and preventing themselves from making an impact.

The reason why I say that is that Alfred was saying, “Well, this is stuff we already have, a lot of the technology we already have. It's based on a lifecycle approach that is role-based for whoever it is, the IT analyst, the developer, or the business analyst."

So, it really isn’t anything dramatically life altering and new. It'a an extension of where they have already been heading.

Gardner: It sounds like a marketing exercise to try to take some steam out of the competitors in the market, maybe even a little FUD.

Shimmin: Yeah, I agree, but a positive FUD, nonetheless.

Gardner: Maybe it’s, "Let’s race to this goal. We're on our way. Let’s see who can get their first."

Shimmin: I agree.

Gardner: Let’s go to Tony Baer, now. You wrote in your blog that you are somewhat underwhelmed with the BEA announcements, but, given Brad’s perceptions of a shift in business model, I wonder if that’s changed your outlook at all?

Baer: I'll put it this way. The Enterconnect stuff, added to the virtualization strategy on which they've been embarking on for about the past year or so, is definitely the most interesting nugget of all this. The other stuff just seemed to be a lot of big marking buzz words for "We have upgraded our registry repository. Now, we have published our interchange format."

The question is whether that has anything to do with what Object Management Group (OMG) has been pushing all these years, which is their meta-object facility (MOF). Basically, it’s the idea of having some sort of standard to exchange data in and out of repositories, kind of like a like a Holy Grail.

My sense here is that it makes sense for BEA. BEA has always been an infrastructure vendor, SaaS vendors need infrastructure, BEA might as well step up to the plate and supply it. So, that is interesting.

The other thing I read in all this is that I could see that, through SaaS, it would be route towards the SMB market that BEA has never really had, because they have always tended to be coming in at a high price point. There's definitely a response to Red Hat and JBoss, on one hand, and Oracle, on the other. They need some strategy there.

The question is, though, can all this produce the types of revenues that will provide a growth path in terms of absolute revenues, compared to their traditional high-priced ticket business. I am not really sure about that, because the whole idea of SaaS is that it’s subscription and supposed to be cheaper. Well okay, we lose a little bit on each sale, but can we make it up in volume.

Gardner: I share this concern about BEA. Its stock has been bumping along at about $11 a share for several years now. They've been having some ups and downs on their licensing revenue, but nothing approaching the strong growth they had early on, led by the transactional processing capabilities and then their application server development-deployment stack. You're right. It’s also difficult from a sales perspective.

Its sales force has been out there creating enterprise accounts. How are those sales people going to go out and sell and be compensated -- and be motivated -- on this recurring-revenue model. It’s a whole different sales thing. It's hard to steer a big ship in a sales organization that way.

Shimmin: Dana, if I could jump in here. The impression I got from them was that they didn’t want to do that, but want to sell through Salesforce.com in the same model they already have. Here’s the big ticket. Give us all your money. Here’s your infrastructure and then Salesforce.com is the company that uses the model for subscription.

Gardner: In that case, we're really talking about a dozen or so major companies in the world to call on.

Shimmin: Exactly. It’s like the telco play.

Gardner: Yeah, the service-provider play. But, to address Tony’s point, those dozen or so major players are going to be very cost conscious, driving down to a volume business with low margins. Open source is certainly going to be near the top of their short list for consideration and architecture. They're also going to be looking at pure play and best-of-breed approaches. A number of these SOA infrastructure players have already targeted the SaaS and on-demand infrastructure markets. I'm thinking of Cape Clear, IONA and a few others. Let's go over to Jim Kobielus. What's your concern from a business perspective about BEA making this transition?

Kobielus: The concern I have is the same one you articulated, that they are moving away from the software license revenues, which have been their mainstay, towards a wide variety of go-to-market delivery mechanisms for getting functionality out. Now, there is the whole SaaS model, the virtual appliance model, and so on. It’s the same concern that I've expressed to Business Objects in the business intelligence space.

Looking at the BEA announcements, I am impressed with the various announcements that came out of their conference this week. It shows that they are expertly surfing the paradigms in terms of rich Internet applications (RIA), social networking and, SOA governance, and virtualization on demand. I see an exact parallel in all the initiatives to what Business Objects is doing in the business intelligence (BI) space. Business Objects and BEA recognize that their mainstay, the software license revenue model, is slowly dissolving, and they need to have alternate packaging for their existing technology.

So, looking at BEA this week, I see them more or less trying to stay on top of this very turbulent time, when its not clear how software companies will make a living in the next decade and beyond. They are going to need, as I said, multiple ways of delivering functionality to very focused marketplaces, such as the small and mid-size business marketplace.

I like the fact that that BEA has at least put a foot in the SaaS channel now with the Enterconnect partnership. I like the fact that they now have put a foot into the virtual software appliance space. Business Objects has in the last two months, seriously ramped up its SaaS strategy and is going to be working with third-party SaaS companies. Their growth is strong right now, but I think they recognize, as BEA recognizes, that the party eventually will end for traditional software companies.

Gardner: If there was any hesitation by IBM in aggressively moving to this marketplace, perhaps this will reduce any of those latencies, and IBM will say, "Okay, we are on. Let's go for it," and IBM will be knocking on these same doors.

Kobielus: I'll let somebody else on this call speak to that, because I am not really up on IBM’s open-source strategy.

Gardner: Let's go to Brad. Do you think, now that BEA has declared its intention to go after these providers, hosts, SaaS and ecology players, like a Salesforce, that IBM will walk up to the plate as well and also offer a variety of both commercial and open-source configurations?

Shimmin: Absolutely, Dana. IBM obviously has a strong heritage in the open-source market and will continue to apply that as they go forward. Their strategy is a little bifurcated right now with their ESB, open source versus close source, it’s not the same code base, etc., but still I think they're pretty well positioned to do that, if they choose to. They've got the muscle to do it, so they would be able to win in that game.

Gardner: What about Microsoft? Can Microsoft walk into these same organizations with its stack and its integrated, but albeit a window-centric approach, and compete with either an IBM or a BEA.

Shimmin: They already have. It’s not live yet. I think it’s still in Beta mode, but they have their BizTalk Services, which is a hosted SaaS-based SOA, at least for integration, that they are playing out. They have their partnership with GXS for trying to host integration services. So, they're really positioning themselves in the SaaS market. If they can tie BizTalk services to their Dynamics Live, their line of business applications that are hosted, they'll be able to hit the mid market pretty hard.

Gardner: So, as we're seeing with other waves of adoption in IT over the past few decades, the first big push into the market tends to come from highly innovative players that create their own architecture and cherry-picked components. I'm thinking of the underlying architecture of a Google, an Amazon, or a Yahoo! Then, as the market adoption moves towards more of these on-demand, Web services, and Web-based approaches, some de-facto standard or configuration becomes more prevalent and ultimately dominates the market.

Do you think we are at that phase now? Is the whole SaaS, on-demand, and hosted-services delivery market going to be looking for that de-facto standards, and BEA, Microsoft, IBM, Red Hat, open source, and mixed-breed configurations will all be vying for that.

Shimmin: Yes, and I think they will all be legitimate too. All that’s really going to matter is that they are able to speak the common language of SOA for interfaces into the company’s line of business applications internally.

Gardner: It’s all price/performance, isn’t it?

Shimmin: Exactly right.

Gardner: One thing that open source might have an advantage with is paying for maintenance and support, and is therefore on an incremental payment basis, and not a large capital-expenditure approach.

Shimmin: That’s both for the software and for management of the software. The vendors who are really going to make it big here are the one’s where you aren’t just buying the software. You remember how IBM had to plug into the back of the servers where they would monitor and manage your systems as the vendors, who are able to do that, as part of their subscription, I think they are going to have a foot up on the others.

Kobielus: I agree with what you said, Dana. As more and more software vendors come out with multiple packaging concurrently of the same technology or the same functionality, they're going to have the repackaged software, the SaaS, the hardware-appliance version, the software-appliance version, and possibly the open-source version of the exact same solutions. There is always going to be like one initial version, usually the licensed and packaged software version, that will essentially set the standard for all the other alternate packaging of that same technology.

Let me get back to Business Objects. I don’t want to keep harping on them, but they are a perfect case in point. Business Objects this year rolled out an SMB-focused packaging of their core BI technology called the Crystal Decisions line. Now, they've got three versions of that: standard edition, professional edition, and premium edition. Business Objects this week said that next year, they're going to release, in conjunction with VMWare, a virtual-software appliance version of Crystal Decisions’ standard edition with more or less the same functionality.

Business Objects a couple of months ago announced that they're also going to be implementing a hardware appliance version of all their products with various partners. They're going to have SaaS versions of all their solutions at some point in the future.

So, what I am getting at is that they are going to have all these different ways of slicing and dicing the same technology. In the beginning of commercial television, its shows were just filmed radio shows, until the programmers of this new medium figured out, "We can actually package stuff differently, and should, for this new medium, because that allows us to do different things."

Software appliances are almost like filmed radio right now. Over time, though, they will be morphed and will have their own unique set of features entirely distinct from the packaged software ancestors from which they descended.

Gardner: That relates to a recent BriefingsDirect SOA Insights Edition that we did with Jim Ricotta, general manager of appliances at IBM. He certainly led us to believe that this is going to be an aggressive direction for IBM, probably more on the hardware/software combined appliance and not so much on the software-only appliance, but certainly these configurations are percolating up in the market.

What’s interesting to me is that for these service provider type of clients, the systems' integrators are not really involved. They've been pushed to the side, and it’s the vendors themselves that are creating these configurations, which the architects within these organizations will then adopt. Does anybody want to react to the systems' integrator role in this market?

Baer: Theoretically, in having SaaS, putting it on an appliance, or however you want to take out some of the pain of all this, the idea is to bypass the system integrator. On the other hand, the fact is that maintenance still tends to be very labor-intensive. Fortunately, with remote diagnostics, you don’t have to have that labor on the site, and so a lot of that could go towards the maintenance part.

The chances are, even if you are small or mid-sized business, you may start consuming services from a number of different SaaS vendors. In other words, you are unlikely to put all of your eggs into the Salesforce basket. You might have some Salesforce. You might also have some Business Objects. Somebody's got to make the tool work. So, I see roles for SIs here, if they can be agile and adaptive.

Gardner: So, we'll wait to see some standards configuration and perhaps some leadership by certain vendors for this infrastructure on a price/performance basis. Once that settles out within the service provider sector, then that will give them, ready entree into enterprises. It’s about the right time for enterprises to start moving meaningfully towards service infrastructures.

Baer: The one part I would disagree with is on the standards. There may be standards, but as we have had Web-services standards, even though you have WSI, which is supposed to be a test bed for interoperability, you still don’t have true interoperability. What they do is get you in the ball park.

Gardner: I was referring more to industry de-facto standards, not necessarily technical or enforced standards.

Kobielus: Right, but my sense is that there are always going to be some final assembly and fasteners that you're going to have put together that are just not going to be completely covered by standards. That’s an area for SIs. It sounds like its ideal for them. They will make the stuff work.

Shimmin: You know what, guys, the threat to the SIs isn’t so much from standards. It’s from the business process management (BPM) front, and the modeling and empowering the company to take control of the application it's building from the business process out.

Gardner: In a sense, the infrastructure is coming off the wire, and therefore it’s value added. It’s all about the process integration.

Shimmin: That’s right. Most of the money these guys make is on long-term engagements, where they send folks in to talk about how the process is going to be built into the software.

Kobielus: It’s all about best practices, those templates in the business content, all of that. And, before long, the final assemblers -- where all the money is going to be made in what’s now called the software industry -- the SIs themselves are going to be agile not only among different vendors, but within any one vendor's product architecture.

They will mash up the ultimate versions of the software -- open-source software, software virtual appliance, SaaS, whatever -- of any one vendor's product portfolio in creative ways in which it probably hasn’t even occurred to the vendors themselves. So, they are going to rely on the SI to be creative mix and mash-up developers.

Gardner: Wouldn’t that be what Project Genesis might be alluding to?

Shimmin: Exactly, but that’s not so much for the integrators but for the enterprises that are employing the integrators

Gardner: And then to your point, Brad, that, in a sense, diminishes the role and value of the integrators?

Shimmin: Right. It just lessens the amount of money they make. Their role, as Jim has said, is going to be important, because they're the ones who have to do the final assembly, It just means the upfront costs aren’t as great.

Kobielus: Actually, that brings up another point I wanted to make. Once again, I can definitely draw a parallel with what Business Objects and other BI vendors are doing. I wasn’t at the show; Brad was, but as I was reading through the Project Genesis press release, they talked about social networking and involving the various roles -- in other words, IT people, and also end users and subject matter experts in the development and mash up of SOA and Web 2.0 -oriented applications.

The direct parallel in the BI world, as well as master data management, data integration, and data quality, is that they are all enabling the database administrators (DBAs), subject matter experts, and data stewards, who are business users who have responsibility for particular datasets like customer data. They are providing them with the tools to collaborate over the lifecycle and management of master reference data, etc.

Every one of the BI vendors that’s worth anything is innovating in that exact way. They're bringing the end user, the people with domain expertise, into the ongoing development and administration of these environments through Web 2.0 AJAX -type tools and social networking-type services.

Gardner: Let’s move on to our next subject, which is somewhat related, now that we’ve gotten into this, and that is Capgemini, a large global SI, saying, "We’ll help Google applications and support them inside of enterprises." This is an interesting development.

So far, we’ve heard about Google being interested more in how small offices, individual users, and productivity users would use their SaaS. Now, we have this opportunity for enterprise play. Is this a movement towards more of a SOA and SaaS for more than just rudimentary desktop productivity applications?

Shimmin: Well, Dana, it’s two things. First, it’s a legitimization of this approach, and it ratifies that it's something that’s valid for the enterprise, which I think is terrific over the long-term, because we do need competition to the Microsoft desktop paradigm.

Second, it heralds a direction in which the SIs want to go, which is being able to be more nimble in how they deploy the software. They like to be in bed with the company they work with, but they don’t want to have to wake up there every time.

Gardner: Single-source problem, right.

Shimmin: Right.

Gardner: Does anybody else have an impression about Capgemini, and I guess others saying. "Yeah, we’ll help companies figure out how to take something like Google applications, widgets, and other SaaS and integrate so enterprises can best deliver value to their internal constituencies?"

Kobielus: It’s another approach to client virtualization. There is the client, in this case a full desktop productivity suite, now completely virtualized across the Web, but virtualized in a way that's enterprise friendly and enterprise safe,because you've got an experienced SI and outsourcer like Capgemini behind it. As Brad said, it definitely legitimizes Google in this area. It legitimizes this approach. As to whether anybody in any great numbers will adopt this service, I seriously doubt it.

The Microsoft hegemony holds tight on the desktop and will for the foreseeable future for reasons that many other analysts have commented on over the years. I thought it was an interesting story, but I don’t think it’s going to rock the industry any time soon.

Gardner: That does make a good sense, when we only factor how these applications would be delivered into an enterprise’s internal constituencies. What about some innovative business development activities, whereby a company that is public-facing or is interacting with its clients outside of the organization, might want to start integrating and mashing up such services as Google provides?

Is there an opportunity for an enterprise to become more of a service provider, but start borrowing services from Google at very low cost, because Google monetizes through advertising and then integrate those, mash those, into what they provide to their end users?

Kobielus: Yeah, that’s one potential role. More and more of the Web 2.0 paradigm is focused on user-generated content. So, we have the whole Wikipeida approach. There is a potential for somebody who wants, for example, to start up a commercially-driven or advertising-supported Wikipeida-type distributed reference book, to contract out to Google/Capgemini. They would provide that collaboration infrastructure with the authoring tools and the version control and all that to support such a venture, where it’s user-generated content, and not everybody just slapping up HTML or dot-doc content.

You want people posting rich content that has been prepared in accordance with templates that the proprietor of this online encyclopedia, or whatever, might dictate. So, you say, "Well the price for participating in our environment are the constraints. You use the tools that we provide, and, by the way, they are Google tools supported by Capgemini, or whatever.

Gardner: I am thinking a little differently, not tools, but actually having word processing, communications, instant messaging, presence, and mobile commerce functions and features embedded into high-value offerings. Enterprises and companies could deliver these out to end users, supported by Google’s infrastructure and services, and integrated through an organization like Capgemini.

Another way to look at this is that as Enterprise 2.0 activities kick in, once again, the SIs might be saying, "We don’t want to be left out of this. "We have to get into this mash-up business." What do you think, Tony?

Baer: Definitely, That’s definitely what I was thinking, when I was talking with Capgemini folks the other day. We always talk about how SaaS and SOA are supposed to simplify the job of systems integration. Well, as you put all of this SaaS and SOA out, you still need to put this together. Maybe new mechanisms and emerging and you want to outsource a lot of this stuff. Something like what Capgemini is doing with Google apps is an interesting stake in the ground and it makes perfect sense from that standpoint.

There's another thing I want to tie in. I talked before about the impact of Capgemini getting in there with Google apps. When I spoke with the Cap folks, they said un-categorically, “No, we do not expect to displace Microsoft Office, but there is still a large audience out there that’s not using Microsoft Office, where the economics have been pretty marginal at best to get people connected.”

I'm not even talking about the developing world, but people are in line operations who typically have not been connected by computers and don’t need this sort of thing. This is exactly the type of play they'ree talking about.

Gardner: The whole mobile end-point world, right?

Baer: Exactly.

Kobielus: The whole Net PC paradigm and the thin client Larry Ellison tried to get on board in the late 90s didn’t go anywhere in terms of the diskless workstation or the shop floor worker who didn't have their own PC, but would share a virtual PC with lots of other folks.

Gardner: But, what about a convergence device like an iPhone and all the other knock-offs that get closer to the iPhone function set?

Baer: A couple of things were playing at the time. One, Larry was trying to do it within his walled garden. Two, the technology and the bandwidth just weren't there. I'm looking at the next generation of iPods. Forget iPhones for a moment. It will trickle down to iPhones, as well, and we're talking about 160 GB iPods.

Gardner: The Browser connects through Wi-Fi.

Baer: That is as much as my desktop holds at this point. So, all this stuff is getting miniaturized. At that point, you do get a certain convergence. Of course, these guys are not going to be heavy word-processing users, but they might need to do some texting or very light spread sheeting, and that’s exactly what this is made for. So, no, it's not going to take away the 90 percent-plus market share of Office, but there is certainly a niche out there.

The question is will it be significantly monetized lucrative enough for an organization like Capgemini, because they have very high cost structures. Just as we were talking about before, is SaaS going to be a play that really will contribute significantly to BEA. A lot of this is the economics of how an organization can make money. One other point I want to plug in from before is that I think it makes sense for BEA to sell all these SaaS providers. However, I think the SaaS providers are going to squeeze BEA and say, "Look, we don’t want to pay those upfront cost, we also want to pay per subscriber based on the number of subscribers we sell."

Dana Gardener: What if there is meaningful advertising revenue thrown into the mix? What if Google, through it's interfaces and it's huge lead and dominance of small keyword-based ads juxtaposed to either search or application activities, can change that business model fairly dramatically by making a lot of this stuff free or even subsidized, with the quid pro quo being that a couple of text ads will show up?

Baer: It sounds like a nice idea. For very small companies, it will be fine, but I wonder, when you start getting to a larger company, are they going to want to have ads penetrate through their walled garden.

Gardner: Not to their employees -- to their customers. They can start delivering services out to their customers subsidized by Google, and the quid pro quo is that the end users get a service for very low cost themselves, but they get to look at a couple of contextual ads that actually might relate to what they're doing.

Kobielus: This gives me an opportunity to go back to the radio analogy. It reminds me, Dana, of the approach, where your PBX manufacturer provides music on the hold and gives your vendor the option of piping in some commercial radio station from your area. So, your customers are sitting on hold and listening to the local radio station. Up pops an advertisement or commercial for your competitor.

So, if I am an enterprise, and I'm contracting with Google/Capgemini for this SaaS-based productivity suite provided out to my customers and or my employees, suddenly, my customers are looking at my competitor’s ads. Obviously, if you want to go that route as an enterprise, then you want to say it to Google/Capgemini, "Okay, I’ll take your ads as part of the price, but block my competitors ads and particular companies and particular market places." I'm not talking about porn. If I'm Microsoft, I don’t want to see Oracle’s ads in my environment.

Gardner: It would be very easy for an algorithmic approach to reduce any of that overlap of ads, so that there is no conflict. But, the ads would relate to user activities. It's just like when you go on hold, some companies put in their ad, which says, "Now that we are still listening to elevator music, by the way, we have a discount on three quarters of our catalogue this month and go to www.our-company-best-deals and take advantage of that." So, there is a whole unexplored way of marketing lead generation advertising revenue being driven into a SaaS model. What do you think, Brad?

Shimmin: I agree that Google really shines at stuff like getting out ads. Who knows if that model is going to play out for the consumer or the enterprise, which is what interests me the most. Obviously, that’s not going to fly for them, but what does fly for me, technologically anyway, is the whole human workflow for the information worker.

That's why this is great. Having Capgemini validate what Google is doing is great, because if you're building an enterprise app with workflow that involves instant messaging or some other form of presence communication, or even a live spreadsheet, this is going to cut down on the cost tremendously. It opens up the ability to tie those services into the given out the application that they are building, because as we all know Microsoft is making a lot of strides in tying office apps into it's own .NET and BizTalk and SharePoint services.

Gardner: Like with these "Live" initiatives, right?

Shimmin: Right, but it's not for the J2EE community.

Gardner: The business model becomes more important than the object model.

Shimmin: Exactly.

Kobielus: I see a role for a Google apps or Capgemini-type SaaS service in the area called governance, risk and compliance (GRC). One of the functional components of a full-fledged GRC environment, something called enterprise feedback management (EFM), where the environment regularly surveys all the stakeholders and end users to determine to what extent a given business process is being performed in compliance with various regulations and standards and so forth.

When you're talking about GRC over a distributed supply chain, you have many organizations and personnel not all under single management and not all using the same desktop productivity tools. They may need some common survey feedback and notification mechanism, calendaring environments, and so forth, that they could all share within the context of enterprise feedback management. That’s where an outsourced SaaS service like the Google apps and services might come in quite handy, a common EFM bus and it could span firewalls.

Gardner: Well, we had a far-reaching discussion today, but one of the things I observed is that we are talking about companies that previously would have been odd bedfellows. We are talking about Google and Capgemini, BEA and Adobe, Microsoft and Oracle and yet they are all relating to themselves on a new level, which is around the services delivery, different business models, and more importantly different partnerships and relationships.

So, this is a new era. We talk about SOA a lot, but what’s going to drive this in many respects is this new business approach to delivery of services, monetization of services. and then the underlying infrastructure that supports that. Well, thanks everyone for joining us today. We have been talking about BEA World Conference. This is the week of September 10, 2007 and also the announcement by Capgemini of supporting Google applications this week. Any parting thoughts, let’s start with Tony.

Baer: Basically, I pretty much agree with what we were talking about regarding where we might end up going with the Capgemini and Google Apps. There's a lot of unexplored territory. You mentioned, Dana, about the potential of attaching either advertisements or third-party copy. Well, go into an elevator in a major office building. You are probably already seeing CNN on the screen there. So, think about it as being a precedent.

Gardner: A precursor to what you are going to get through your mobile device or your PC soon.

Baer: Exactly.

Gardner: Jim Kobielus, parting thoughts.

Kobielus: We're already living in ad-glutted world. I think people will gladly pay a premium for an ad-free environment. That’s one of the reasons I think people will still continue to stick with desktop tools like Microsoft Office. It's like, "Wow, good, an ad-free environment that I am paying for and I am going to keep it that way. I'll keep my head clutter down to an absolute minimum."

Gardner: Well, I wouldn’t be surprised if Microsoft starts embedding some ads into it's Live offerings too. Brad Shimmin, parting thoughts.

Shimmin: On the alliances front, any alliance is a good sign, no matter how odd, strange, or otherwise it may be. The fact that we have these strange alliances is a signal to both the maturity of this market that we all talk about all time and also the vibrancy of it. So, I am thrilled that we would have stuff to talk about.

Gardner: Innovative alliances, right?

Shimmin: Yup.

Gardner: Well, great, I want to thank our panel Tony Baer, principal at OnStrategies, Jim Kobielus, principal analyst at Current Analysis and Brad Shimmin, principal analyst at Current Analysis. This is Dana Gardner, principal analyst at Interarbor Solutions, and you have been listening to BriefingsDirect SOA Insights Edition Vol. 25. Thanks, and come back next time.

If any of our listeners are interested in learning more about BriefingsDirect B2B informational podcasts or to become a sponsor of this or other B2B podcasts, please fill free to contact Interarbor Solutions at 603-528-2435.

Transcript of BriefingsDirect SOA Insights Edition podcast, Vol. 25, on industry mergers and acquisitions. Copyright Interarbor Solutions, LLC, 2005-2007. All rights reserved.