Parallax : Calculating Technology's Future: The Emerging Application Operating System

About four years ago I was sitting in the front conference room at Fidelity Ventures in downtown Boston. I had a few of my friends from Fidelity in the room along with one of the founders of BEA. We were hypothesizing about the future of IT and I proferred an opinion that packaged applications would become the new hub of IT. It just made sense to me - packaged applications are so critical and run so many processes they just can't be treated like a sidecar application. This idea went over like a lead balloon.

Well, I put my money where my mouth was and started Newmerix, a company specifically design to help people manage packaged applications. Now, 4 years later i'm thinking that was a pretty good bet. The recent moves of Oracle, SAP, et al are pretty clear that the next phase of packaged applications is not about applications but about IT application infrastructure. This blarticle explains a bit more...

The Emerging Application Operating System

Author's Preamble:
There is a saying here in Colorado: “If you don’t like the weather, just wait a minute”. Apparently technology punditry (insert inflated ego here) is sort of similar. I wrote this blarticle about 3 weeks before Oracle broke open its second piggy bank and ponied up for Siebel. I wanted to keep the original article intact because, as you’ll see, some of what I was predicting for Siebel has come true. It is still undetermined how Siebel’s recent application platform introductions will shift under their shinny new Oracle moniker, but if nothing else, it gives me something more to write about.

Anyone who has gone through media training knows the “sound bite trick”. It’s a carefully planned but nonchalantly delivered little snippet that, if nothing at all else gets quoted, you declare victory when you see it in print. It’s similar to a trick used by the old door to door salespeople trying to sell knives. If you can’t get someone to buy the 10 piece steak set, at least get them to spring for the paring knife that can cut through a titanium rock. But alas, the sound bite trick doesn’t always work; sometimes you can get the press to say what you want, and sometimes you can’t. And here’s a good example.

For probably a good 3 months before, during, and after the Oracle/PeopleSoft acquisition, Newmerix was bombarded with press interviews asking us what we thought about the merger of these two giants. I was determined to get placement off a little sound bite we conjured up one day: “It’s the 10 Billion dollar battle over IT”. After a few shots across the bow with no success, I became resolute that no matter what I was asked, I’d find an answer that included this sound bite until I saw it in print. It went something like this:

Press: “Mr. Robertson, what’s your take on the Oracle/PeopleSoft merger?”
Me: “It’s the 10 Billion dollar battle over IT”

Okay – not much of a stretch, but when we got not hits on this one, I started to get desperate,

Press: “Mr. Robertson, how long has Newmerix been in business?”
Me: “Well before the 10 Billion dollar battle over IT started”

You get the point. But come on, what’s not to love about that sound bite? It’s sexy. It’s provocative. It’s a big idea. It sizzles like a cracked egg on a hot sidewalk. And apparently it’s completely uninteresting to anyone in the press. So rather than picking up my toys and going home on this one, I am going to stay convicted that the people “want to know”. I’m taking matters (and the liberty of this blog) into my own hands and I’m going to print it myself. So here it is:

“It’s the 10 Billion dollar battle over IT”

Done and done.

So why was I so adamant about this sound bite? Well the truth is that there really is a battle a’brewing on the IT front and Oracle just took of the gloves. This move was the culmination of a long story that had been building up for many years in the packaged application market. You see, way back in the beginning it was all very détente. PeopleSoft did human resources applications (hence the “People” part of PeopleSoft), Oracle did financials (e.g. sold more Oracle databases), SAP organized your suppliers and inventory, and Siebel just sat around smoking a menthol being the cool kid who invented the CRM market. That worked great for a while until some analyst crashed the party and call the whole thing ERP. Now all of a sudden everyone found themselves in the same big ERP market (okay, CRM is not ERP, I admit it). Trust me, many things happen for good “market” reasons in the software industry (like changing architectures for new technologies, companies building things customers actually ask for, etc..), but many more things happen in the software industry because doing it can put you in the top right Gartner Magic Quadrant. Seriously (and to Gartner’s credit), its like leaving an unwrapped snickers bar on a picnic table surrounded by ants. We all know what’s going to happen. And thus the ERP arms race began.

PeopleSoft got into financials, Oracle got into manufacturing, SAP got into human resources and Siebel.. well.. kept smoking their menthols for the time being. This proliferation went on for a number of years, spurred on by the Y2K phenomenon where it was cheaper to buy a whole new SAP system than hire a few COBOL programmers to fix your mainframe software. But finally, as much as Siebel and CRM tried to stay Switzerland in this whole thing, eventually they (and CRM) got sucked in as well. The problem though is that after a few more years the conversation sort of ended up like this “My financials packed it better than your financials package.” “No my financials packaged is better than your financials package”. “Okay, well my CRM is better than your CRM”. “No, my CRM is better than your CRM”. Well this gets kind of boring and relatively unstrategic pretty quickly for both the customer and the company. Tit for tat is not a growth strategy (look for this hot new bestseller any day now). And this boring story is where we were clearly wedged by the end of 2003.

Then, out of nowhere, SAP changed the game. To be fair, their first move didn’t change the game as much as it nudged it in a new direction. And in fact, most people didn’t even really pay much attention to their first move. It sort of seemed weird and out of place. The first thing they did was say “we’re building a web server and an app server”. To the casual observer this seemed like a very odd thing to do. “We have good web servers!” the pundits proclaimed. “We have successful application server companies (and hey, even Oracle has one)”, the analysts ranted. It’s was almost as if SAP said “we’re going to build our own database”. (Ooops, they did that too). So you get why people might have skipped past that announcement and flipped right to last nights American Idol winners. But SAP was doing something very very strategic in the background with this manuever. Apparently someone at SAP had been reading the Microsoft playbook one day and said “hold on, if they build some of their apps on SAP, why couldn’t they build ALL of their apps on SAP. We just need to give them a platform”. You can only imagine the ensuing conversation included key phrases like “world dominance” and “expansion strategy”. So they put their heads together and they came up with NetWeaver.

Apparently NetWeaver was marketed with the same strategy as .NET when it first came out (perhaps they really did have the Microsoft playbook). What I mean by that can be summed up by one word: Obtuse. The NetWeaver launch was the packaged application industry’s equivalent to the Onceler’s “You need a Thneed” strategy in the Lorax. Apparently everyone needed one and apparently NetWeaver could do anything you needed it to. But if you looked closely, right there buried in the text of every NetWeaver print ad was the golden ticket: TCO.

The argument is actually a pretty solid one. You build all of these applications, most of them now as J2EE apps on top of an application server, and you manage your SAP apps as well. Why not just run everything on one application platform. Less vendors, less integration, easier management, world peace (well okay, maybe not less integration). It sort of makes really logical sense. Plus, with a hefty head start, SAP would have something new to talk to the market about. And since this announcement, SAP has done a nice job of going even further. They have defined two key pieces of the application platform strategy that are intimately linked: xApps and Composite apps.

Intrigued by all of this (and with a healthy corporate travel budget in hand), I traipsed out to SAPPHIRE this year to try and figure out what these things really were. Unfortunately, 3 days later I was still confused. Then I came across a very good book on NetWeaver by Dan Woods. I actually remembered Dan from his days at TheStreet.com. When I was building my last company, Service Metrics, Dan was one of our first customers. Dan is one of those excellent CTOs who can balance technology knowledge and market understanding very well. He may not remember this (or me), but my sales guy, intent on getting TheStreet.com to be one of our early customers was going to get the deal done “no matter what”. When we went to visit Dan at his office in New York, Steve actually blocked the door to Dan’s office, spread eagle goalie style, and said “you’re not leaving until you sign this contract”. Well, Dan’s a good sport (or not very quick on his feet) and he ended up signing the contract (we eventually did let him out of his office). So here’s the innovation of xApps and composite apps that I learned from Dan.

Well basically a composite app means that you can mix modules delivered from various vendors (SAP to start with but other vendors are being heavily recruited) with custom modules you write yourself. If you love SAP’s project management module, but hate their requirements module, just glue their stuff with your stuff and off you go. They do all the plumbing, all the analytics, all the UI, security, etc.. An xApp is just a fancy word for “SAP module that can be used to build a composite app”. Why three days and 7500 people at SAPPHIRE could not articulate these simple concepts is baffling to me.

Okay – so why do I call it an application operating system. The magic is not in the standard 3 tiers (web server, app server, database). That’s old hat now (although it’s an old hat that SAP makes themselves now). The key is in the plumbing. What NetWeaver has done well is establish a basic set of features that you need to build an enterprise application nowadays and has provided all that plumbing through easy to use coding tools. Application integration, portal services, security services, consolidated data management, BPEL process designers and management, workflow management, analytics. It’s all in there and all pretty much drag and drop to get it to work through WebDynPro (their authoring environment) and other development tools (e.g. Eclipse). Just join the SAP Developer Network and you’ll see how much stuff is there. This sounds a lot to me like what Microsoft provides in their operating systems. A set of basic services through the OS (networking, security, OLE, COM, etc..) all with slick coding tools on top to help you pull it all together. Some analysts have called it the Applicstructure, others the Business Process Platform. I’m sticking with Application Operating System.

So back to the “10 Billion dollar battle over IT”. The folks at Oracle clearly saw this coming. They had a lot of the pieces (Oracle AS, Forms, JDeveloper, the database, EBS, Identity Management tools, BPEL Manager, etc..) and they needed a response. The problem though is that they didn’t have a big enough installed customer base of EBS users to convert to their own Application Operating System. Well with PeopleSoft and JDE’s combined 11,000 customers plus their own 10000 or so EBS customers (the numbers are hard to pin down), it gave them a pretty good answer to SAP’s 25,000 strong customer base. It clearly wasn’t about the apps (what company really wants to own 3 different financials packaged applications if you count EBS, PeopleSoft and JDE). This move was about owning the future of IT. It was about owning the operating system that all applications were going to be built on in the next 10 years. And I have to say it was a choice move on their part (yes, I’ve been watching 80’s movies). Oracle Fusion is their response to NetWeaver. If you look at their building block markitectures you will see a striking number of similarities. Not surprising given what an application operating system has to provide.

So why should you care? Its simple – you are going to use this and you are going to have to figure out how to manage it too. The platform and plumbing is great, but the vendors have historically given their users very limited help in managing their tools. How do you roll changes into this operating system? How do you patch? How do you manage different modules all changing at the same time? This will not be an easy task. Fortunately for Newmerix, we already do a pretty good job of this. One of the happy byproducts of this new direction is that more apps are now going to live on top of packaged application architectures. Since Newmerix built our whole Automate! management suite to take advantage of this architecture, we’re well positioned to help across NetWeaver, Fusion or Nexus. Trust me, we’ll be there when you need us.

Oh yeah, I forgot to mention Nexus. Nexus is Siebel’s soon to be response to Fusion and NetWeaver. I say “soon to be” because it hasn’t really dawned on Siebel yet which battle they should be fighting. All the early reports on Nexus say it’s a strategy to go after the 70% of companies that build their own CRM solutions. Okay – that make sense if the world is only CRM and lollipops. But sooner or later you have to come to grips with the fact that the market has moved on (and no one really likes menthols anyway). The good thing for Siebel is they have all the pieces. They have a nicely architected package application infrastructure. Its customizable (one of the key requirements of CRM installations) and very flexible. There is no reason why they couldn’t add a dash of J2EE integration, a pinch of analytics, pick up a portal player, and be right there riding shotgun with Oracle and the gang. Keep an eye on this, Siebel may be down on the matt at the seven count right now but I have a feeling they are going to get up soon and fight another round.

Author’s PostScript:
Okay – so Siebel got bought. That one I missed (I was actually predicting Microsoft myself). Relating to the acquisition in a recent C|Net article, I found one of best backhanded quotes I have ever seen from our friend Mr. Gates about Larry’s recent conquests. It goes something like this “Larry said the software industry would consolidate. After spending 10 Billion dollars he was completely right”. Very nice Mr. Gates, very nice! But back to Siebel for a second. In all of this morass, they finally did get the composite application religion. Nexus is now a full fledged composite application platform capable of integrating your own custom J2EE apps with Siebel. Hey, they were not first to market (or second for that matter), but who’s counting. Now just one more lingering question: If you’re Larry Elison and you wake up one day with four, count em four application operating systems, what the heck do you do? So much more to come I am sure…