IBM CEO Lou Gerstner recently briefed securities analysts on IBM’s e-business strategy and performance.
He introduced his talk by mentioning the company’s portfolio restructuring, and the impact of that work on its growth.
“Over the past 12 months, we’ve made a series of strategic decisions that have affected short-term revenue, but will yield – and are already yielding – a more profitable portfolio. You saw that play out in our first-quarter results. ”
“Let me quickly review those actions. We sold the IBM Global Network to AT&T – a very smart decision, which looks smarter every day. We got out of a very capital-intensive business, got a very good price, and maintained access to a world-class network.
Gerstner said that the deal they had struck with Cisco was paying off, both in terms of OEM technology and their joint services business. They had aggressively reduced their commitment to Drams and were getting a greater percentage of semiconductor revenue from high-margin logic products.
He spoke of the massive restructuring of the company’s PC business, how they had been caught in the market shift from indirect to direct sales, and how they would slash the cost structure to make the business profitable.
“Now, there are lots of ways to play ‘what if’ with our revenue numbers. By one calculation, the restructuring plus HDDs cost us about a billion dollars in first-quarter revenue. Let me assure you that we intend to produce much better results as we move into the second half. But what’s most important from my perspective is that our fundamentals remain very sound – from our strategy, to the core building blocks of our portfolio.
I’m more than ever convinced that we are on the right track. That confidence starts with the whole e-business movement. This isn’t like some of the prior transitions that our industry has gone through, when IBM was forced to play catch-up. We were first on this one. And everything we’ve said about this revolution has come to pass: that this was going to be about business, not browsing; that it was going to be about institutional change, not individuals chatting or playing games; that the biggest issues our customers were going to face were about strategy and process transformation, not technology; and that the e-business world would represent a massive shift from PCs to enterprise servers.
All these trajectories are playing out and they play to our strengths. I know you want me to walk through IBM’s individual businesses. Before I do that, I think it’s important to provide you with a brief assessment of what we see happening in the industry today and how those major trends drive our strategy and our investments.
You all know that in prior sessions with you I’ve made some comments that were viewed at the time as somewhat controversial.
One was that the PC era was over.
And last year, I mentioned that the world was captivated by the dot-coms and suggested that some of them were fireflies before the storm – that the real storm was going to come when the world’s established enterprises came to the net.
Look, e-marketplaces are going to be important – at least, some of them are.
There will be lots of them. These e-marketplaces are going to broker a lot of transactions. And we’re going to build a lot of them with our partners, Ariba and i2.
But as usual, there’s a lot of confusion and a lot of people who seem to think B2B commerce is one and the same with e-marketplaces and e-exchanges.
And they’re not. E-markets are not synonymous with online business-to-business commerce.
There’s a big difference between making a sexy announcement and exploiting the net to drive real cost out of your supply chain.
You do not have to spin off your supply chain in order to get the benefits of e-B2B.
In fact, it’s just the opposite: you could be cutting off an important part of your infrastructure genome.
More and more of our customers realize that e-business isn’t about creating one dot; it’s about connecting all the dots that are important in an enterprise.
And I’ll say more about this subject of integration shortly. Before I do, I want to spend a minute on the first of two important trends that we see affecting our large enterprise and medium-sized customers today.
The first is a whole new infrastructure underpinning e-business. I talked about this last year – that e-business is shifting the workload from PCs into the computing and communications infrastructure. The demands and requirements already being placed on the infrastructure are enormous, and they’re only beginning. They will increase exponentially.
Now, a lot of people are talking about the importance of computing infrastructure today.
But this isn’t about rediscovering the old back-office.
There’s a very significant difference. The new e-business infrastructure will be open, not proprietary. It will be heterogeneous and multi-platform. It will incorporate, over time, a lot more ‘intelligence’, and it will be capable of supporting and managing a whole range of devices; not just PCs.
“Let me just touch on a few highlights, starting with servers in the context of an e-business infrastructure.
Historically, servers have been multi-purpose platforms – basically defined by proprietary operating systems – that ran applications accessed by terminals or PCs.
That’s a very simple world.
IBM created it.
Today, we see that that configuration is by no means adequate for serious e-business infrastructure.
Servers are not generic.
One size does not fit all—not in capability, not in price, not in customer, nor in application.
I know some of our competitors would like the world to believe that their one product line can address every need. That makes for nice soundbites, but it doesn’t reflect what’s going on with customers.
What is happening? Four or five years ago, most web sites offered only static content.
Users came and did simple queries.
They looked up news; they looked up the weather.
They checked a savings account balance; they looked up flight information.
This is the world of web servers.
It’s huge and important to customers, but it’s relatively straightforward in terms of what needs to be done: Multiple copies of static information – data or content of some form such as online catalogs, replicated on many servers, accessed by PCs running browsers.
This is where Unix is strong – as loads increase, you just roll in another server and keep going.
However, as you know, since 1995 we’ve said that the net’s not about just looking up static information; the net’s about serious business transactions of all kinds.
And that’s where the second server category emerges—the data sharing and transaction processing server.
You cross into this category when the user moves money from savings to checking, books the flight, trades stock, buys something from that catalog.
Databases are updated.
High volumes of transactions are executed.
And all of this has to be done in a secure, highly available, highly scalable environment.
This is classic mission-critical-can’t-go-down computing.
This is the world of 390 and with the kind of processing we’re driving today into our high-end Unix processors.
Demand for this class of server is going to grow as customers’ e-business applications become more transaction-driven.
And the third category is really new—it’s what we call appliance servers.
Some people call them “edge” servers.
They’re highly specialized servers; they’re dedicated to running certain applications and in most cases one application—managing caching functions, managing wireless devices, providing security/directory services.
These applications are just now emerging and they’re best supported by smaller servers running NT and Linux.
The point I want to make is that there is no ‘generic server’ and there’s no ‘generic server configuration’ for the e-business infrastructure.
The requirements vary depending on size of customer, the nature of the applications and where the customer is in terms of becoming an e-business.
And this is what we mean when we say that the infrastructure is becoming more and more intelligent.
The systems themselves are going to have to perform very sophisticated tasks automatically as e-business applications become more demanding and more pervasive.
For example, if I’m accessing the web from a data-enabled cell phone, the infrastructure should detect that and automatically convert the web site contents to a format that can be displayed on a very small screen.
If my native language is Spanish, the infrastructure should detect that in my user profile and do translation on the fly.
And as net extends beyond end-user devices and connects devices embedded in cars, appliances, packages and bio-medical implants, the back-end systems will have to sense, track, respond and manage billions—perhaps a trillion—devices all over the world.
This is orders of magnitude beyond what anybody’s systems can do today.
So as we see e-business becoming an indispensable part of the world, there are clear implications for server technology, as well as middleware.
And it tees up several meaty technical challenges for the research community of our industry.
And our researchers are hard at work on all of them.
Let me just mention two other aspects of e-business infrastructure.
Storage: We expect 75 per cent of all new hardware dollars will go to storage systems, with the bulk, about 70 per cent of that, in networked storage: storage area networks and network-attached storage.
And then, services.
I’m not talking here about traditional labor-intensive services here.
I’m talking about a lot of back-office infrastructure moving out – essentially outsourced – to a third-party service provider.
We see the beginning of this in our hosting and network caching.
You may know that we already have the largest hosting business in the world, and it will continue to grow in the kinds of services that we will deliver to customers.
I’ll talk about our services business in a moment.
So, e-business infrastructure is growing in importance.
It is creating huge opportunities for our industry, and we will go after those opportunities in two ways.
First, providing critical products and technologies in servers, storage and middleware to enterprises that will build and manage their own infrastructures; and to service providers like telcos and ASPs.
And secondly, we will continue to be a service provider ourselves, and again, you see that in our expanding hosting business.
Now, the second trend I want to talk about from an industry perspective is that the Net enables and at the same time demands totally new levels of integration.
This integration – this is a lot like what happened in desktop software, although much more complicated.
First there were spreadsheets and then word processors and calendars; and then because the customer needed it, they were integrated into suites of “productivity apps.” The same thing is now happening with enterprise applications.
Any enterprise that wants the efficiencies and cost savings of e-business has to integrate business processes and the enterprise applications that support them.
Think about a simple order taken over a Web site.
It sets off a chain reaction within the enterprise and across the trading partner network – through the inventory systems, pricing, fulfillment, manufacturing – across the whole supply chain.
The question is, are all the processes and applications to support the immediate, seamless, frictionless movement of critical business information and transactions in place? And not just processes and transactions within enterprises, but also the B2B linkages among them.
That’s what we mean by integration.
Every leading company is coming up against that wall right now as they try to exploit e-business.
How does any enterprise do it? Well, for a while, the answer was largely services.
Lots of people parachuting in to a customer’s business, doing lots of programming, lots of wiring together of all the different pieces.
Systems integration will continue to be an important function and a growth area.
But in the world of e-business, more than anything else the integration process is moving to an area called ‘middleware’. The once prosaic, unpopular middleware layer of software – databases, messaging, groupware, transactions and systems management – for all intents and purposes, is the layer software developers are writing to today.
Not the machine-level code; not the operating system.
So with those two trends as the backdrop – infrastructure and integration – let’s look at our individual businesses.
Let’s start with software.
As most of you know, we’ve developed a highly focused software strategy over the past few years.
Let me take just a minute to remind you that we manage our software business as two businesses.
The first is host or operating system software, which is now and always will be a continually shrinking percentage of our revenue.
We’re not fretting about this because, frankly, operating systems no longer hold the strategic importance they once had in our industry.
In a world of open standards, which is where the world is going, the operating system platforms – ours or anyone else’s in the open world – are not going to be control points anymore.
They won’t be where the growth is, and they won’t even command the margins they once did.
The second and more important software business that we run is the one I just talked about— – middleware.
This is where we have been investing for five years and investing heavily.
And frankly, we got a little lucky here.
When we started building up our middleware technical competency, our development, our marketing muscle, and we opened up our products to support non-IBM platforms, we had no idea how important middleware was going to be to the e-business world.
Middleware is what makes e-business work.
It accounts for 75 per cent of what our customers spend on software.
As I said to you earlier, it is the critical layer to which all the new applications are being written.
And we hold the number 1 or 2 share position in every major middleware category: Database, Web serving, messaging and integration, collaboration, transaction processing and systems management.
So I think you know our software business and I think you know where we are taking it.
You see we’re highly focused and we’re investing – both in the middleware products themselves and in the software sales and marketing support which we really didn’t invest in sufficiently in prior years.
We’ve seen how this kind of focus has paid off in revenue and share – for example, in the success of our database business in the last few years.
There is no software company in the world that is better positioned in multiple segments of the e-business world.
Let’s go to servers.
Earlier I described the server opportunity within the context of this e-business infrastructure.
Again, I said we see three categories of servers, based on the e-business applications the customer is investing in.
Let me talk about how we’re managing our server business to capture those opportunities.
In Web servers, Unix has captured a significant percentage of this market.
You know about the success of the S80, our high-end Unix product.
Later this week – in fact, Thursday – we will be announcing midrange servers that incorporate some of the S80’s high-end technologies.
As you know, the midrange below the S80 is the sweet spot of the Unix market and where Sun has had an open field.
Watch these new products; they incorporate a version of the S80’s proven SMP technology, as well as our high-performance copper technology in the microprocessors.
These technologies have enabled the S80 to win every single performance benchmark against Sun.
Now, in NT and Linux, which account for more than half of the Web server market, you know that Netfinity, our NT server, has taken share in each of the last two years against its competitors, and we intend to build on that momentum.
So, in Web servers, we’ve really been playing with half a hand: a strong offering in NT and Linux.
But we’ve been soft where it’s hurt us the most, which is midrange Unix—a hole that we’re going to plug this week.
Next, the emerging class of servers we call appliance servers.
Now, this is where NT and Linux will be strong, and as I said before, we feel very good about Netfinity’s offerings and all the work we are doing in Linux.
Let’s conclude with the data and transaction servers.
These are the mission-critical engines that can’t go down, and that can and must handle high transaction volumes with high security – classic 390 challenges, and increasingly, for our high-end Unix products as we infuse them with 390-level capabilities.
In the midrange, the AS/400 provides this kind of scalability and capability.
I’ve just mentioned the S80, which outsells Sun’s high-end offering 3 to 1 in this data-intensive, transaction-intensive environment.
In addition to pushing those technologies into the midrange, we’re also strengthening our Unix offering with technologies from the Sequent acquisition – our NUMA technology.
This will give the RS/6000 a kind of ‘parallel sysplex’ capability that is needed for high-end transaction processing and data sharing.
But again, the strongest data and transaction server in the industry continues to be 390.
Let me mention a few things we’re doing to continue to strengthen its capabilities in this category.
First, integrating Linux to make it easier for our customers to bring applications into the 390 environment.
This is very important to our customers who want to bring their applications closer to a shared database.
Secondly, customers have been pushing us to scale the 390 even further.
I want to tell you, I can’t count the number of customers who call me directly saying, you are the only hope we’ve got to scale our e-business servers to the level we need.
You will see in Freeway, our next-generation 390 – which we will announce later this year – the ability to do very, very sophisticated load-balancing to handle huge e-business workloads with unpredictable peaks.
Don’t take it from me, let me tell you what one industry consultant said very recently about 390 performance in e-business.
This is from the Meta Group: “By 2003 System/390 geographically dispersed parallel sysplex will have a design point of 100 percent up-time at the application level.
Current Unix platforms are down 35 hours a year.
Windows, down 88 hours a year.” This is what our customers are hearing from their consultants.
Let me turn to storage.
You know that this is an important market to us.
Many of our customers are growing their storage requirements by 50 percent every year, and that’s been accelerated by e-business.
You know we’re in the game with Shark.
It is making headway against EMC and we are committed to strengthening Shark going forward.
Shark is an architecture; it’s not a product.
It’s a modern storage area network-designed architecture.
We are ramping up our efforts in storage area and in network-attached storage—both critical elements of the e-business infrastructure.
But the fact is we don’t think that any vendor is going to make it with a “pure storage” play.
The strategy must be open and encompass servers, software, storage, and services.
Let me put in perspective an important transition that is taking place in the IT services industry today.
I learned back in my consulting days 25 years ago – services go through product cycles just like any physical product.
And we are right now in the middle of a huge transition in information technology services.
On the one hand, most of the activity and spending in large enterprises over the last few years was dominated by services built around ERP implementations, Y2K and customized systems integration.
That’s where the big guys made their money; that’s where the revenue growth was.
Over the last six months, a massive shift got underway.
Obviously, Y2K went away as a revenue source.
But much more fundamentally, our customers around the world are consumed by e-business.
No CEO worth his or her salt makes any important public statement that doesn’t say they are a leader in the e-business economy.
Go read the annual reports.
Go read the reports that CEOs are making at annual meetings.
So a very significant amount of the resources that heretofore were devoted to the more traditional services are shifting to e-business.
Now, that shift is driving extraordinarily explosive demand for e-business services that didn’t even exist 18 months ago.
And that’s why you’re seeing a lot of these small e-business consulting firms hit the radar screen.
Some of them are enjoying a great run right now in the niche they operate in, which is mainly a lot of work on front end, interface and web design.
But what’s happening is, as we move into real e-business, our customers start to get very serious about integration.
They become very serious about, “Okay, what do I do with this pretty front end and this nice e-business strategy? How do I drive it into my business processes and business applications?” Here the services requirement turns to much deeper levels of experience, expertise and capability.
I’ll make exactly the same point about web hosting.
There are companies making a good living right now in this space, or at least, they’re generating a lot of revenue, including IBM.
I’ve already told you our hosting business is the largest in the industry, and will probably double this year in size.
But again, let’s talk about capability.
You know, we started out in the hosting business with the most complex, high-volume sites imaginable: the International Olympic Committee, and a lot of commercial enterprises like the New York Stock Exchange.
That’s a different ball game than basic co-location services – roll in your server, loaded with your application, and we’ll mind it for you.
That’s a nice offering as far as it goes.
But customers who start simple are bound to increase their appetites – at least the ones who are growing.
They’re going to evolve to more complex implementations.
They’re going to want to do more transactions.
They want to buy more and more of their processing, their storage, their communications, their appliances, as services delivered over the Net.
And if you think back to the earlier discussion we had on the kinds of servers best suited to specific tasks – simple hosting is classic Unix.
But as customers move up the food chain to more complex transactions, they’re also going to drive up their server requirements.
If we can do the complex, we can do the simple.
I’m not sure the converse is true.
I should also point out that while this transition is going on – from traditional services to e-business services – it has caused a lot of slowdown in those traditional service categories.
It has done nothing however – it has not changed the outlook for traditional IT outsourcing.
Sure, the Y2K pause stopped everybody from thinking about outsourcing as they rounded the year.
But outsourcing still involves a totally separate set of decisions, and we see no change in our customers’ interest levels or their commitment to outsourcing.
In fact, it’s heating up, especially in Europe and Asia.
You may have seen, we signed a $100 million deal with Texaco yesterday.
And you’ll see a lot of announcements over the second quarter and the remainder of the year.
I told you two years ago that I believed services is the best part of the I/T industry.
Since I told you that, it has become not only the biggest part of the industry, but the fastest-growing part.
Everybody in the industry is scrambling to add services capability: Cisco-KPMG; Microsoft-Andersen; HP with Matrix One; CAP Gemini and Ernst & Young.
And I can understand the motivation behind these alliances.
They see the same opportunity we see.
But they pale in comparison with the capabilities of our $30 billion Global Services business.
I have zero – zero – long-term concern about the slowdown in services over the last six months.
It’s a blip in a seven-year track record of performance.
Let me just spend a moment on our business that involves the sales of component technology, primarily to other high-tech companies and customers in the computer and communications industry.
We’ve talked about the transition underway in our semiconductor business.
We’re getting out of this some-time high-revenue growth but zero-profit DRAM business, and instead growing in the high-end logic business.
We are now number one.
From a standing start, we are number one in the ASICs business worldwide.
This rebalancing – Drams coming down; ASICs going forward and other infrastructure products is going to drive substantial revenue and gross profit performance as we go through the rest of the year and into next year.
I’ve only touched on the whole explosion in wireless and other pervasive computing devices.
This is a huge opportunity for us in the area of building specialized semiconductors – low-heat, low-power consumption – which is what all or many of our new discoveries have been about.
These new technologies are required to build these new devices to fortify the world’s communications infrastructure.
Yesterday Alcatel and Nortel Networks each announced that they’ll be using our chips to drive advanced telecommunications networks.
You’ll hear similar announcements very soon from several other networking gear makers.
Let me close with this.
The e-business world we predicted is here.
It’s just beginning, and it drives demand for everything that’s core to our portfolio.
The question you have often is: How do we go to market with that portfolio? How do we tap into this opportunity? And, how do we win our share? Well, one way is what we just talked about – services, starting with our e-business strategy and leading to implementations in our customers that require hardware, software and more services.
And the other way we’re going to market represents a major departure from what IBM has done in the past, and is new since I last talked to you.
It’s a strategy decision we made last summer.
In this frenzy over e-business – where speed is essential – customers are building fewer and fewer applications in house.
Instead, they’re turning to companies – software companies – that have packaged leadership offerings today.
Recently, the killer app on the net has probably been business-to-consumer e-commerce.
But we all know that customers are increasingly aware of the fact that to get the real benefits and efficiencies of the net, integration is critical—all the processes and the applications that support them.
So customers are now shifting their focus to a whole new set of e-business applications, like customer relationship management, business intelligence, knowledge management and supply chain.
We made a decision last summer that we’re not going to build those applications.
What we decided is that we’re going to work with the leading software developers – segment by segment – in a very special way.
We’re going to give them access to things they desperately need, like the middleware that integrates their applications, and the technical and development infrastructure that we have in our company.
Access to our research treasure trove, and very importantly, access to our customer relationships.
In return, these software partners commit to specific revenue levels and to install a defined percentage of all new application licenses on our hardware and middleware supported by our services.
These are true partnerships.
They’re non-exclusive, but they’re all structured to begin with the commitment that we will lead with each other’s products – not just offer them as one of many options.
So far we’ve signed 12 partnerships with many more to come: with Siebel in customer relationship management; i-2, IMI, Aspect Development and EXE in supply chain; Ariba in procurement; Retek in retail distribution; S1 and TIG in banking and insurance.
The three-year value of just those first 12 agreements is projected at $5 billion in incremental revenue.
And most of that comes directly out of the pockets of our competitors.
Let me use the relationship with Siebel to illustrate how this works.
Siebel will lead with all our middleware products.
In fact, it will design its system, it will build its system, to our Application Framework, our middleware, instead of the competitive products they were using before.
Siebel has committed that they will install 20 percent of all new licenses this year on IBM platforms; 50 percent in 2002.
That’s up from 9 percent last year.
Now, some of you know Siebel very well.
You have spreadsheets and models, and you have a good sense of what kind of revenue they expect to generate this year, next year.
So if you just apply the standard industry ratios, which say that for every dollar spent on the application there’s $1 spent on a client; $1 spent on glass-house infrastructure; $1 spent on middleware; and $3 to $7 on services, then you get a pretty good idea of why we like the way this strategy plays out for us.
We’re not going to get 100 percent of that spending, but there’s no reason we can’t get the lion’s share – which is largely revenue that Andersen, Sun and Oracle and others have been collecting up until now.
So, let me try to summarize what’s been going on in the company since I last saw you a year ago.
A year ago, we had just come off a 14-percent revenue growth quarter.
We were on our way to 15-percent second quarter revenue growth.
Our business was hot.
We had just completed the AT&T Global Network sale, which generated an enormous gain for us.
So we made a decision last June.
We said, we had a great opportunity to restructure our portfolio, which will cost us revenue in the short term but will improve our fundamental strength of the company.
We were sitting high and ready to do that.
We got out of Drams.
We got out of network hardware.
We got out of a number of businesses.
And every one of those decisions has paid off very well.
Then came September.
And we got hit across the face with a wet fish when our customers stopped and shut down their whole business.
We should have seen it.
We predicted it; everybody predicted it.
But nobody expected it to be so concentrated.
So, the first thing that’s been going on since I last saw you is this enormous restructuring that we were able to do last year in spite of Y2K, still grow our revenue 7.4 percent, still meet adequate profitability, and still grow our stock price 17 percent.
The second transition I told you about is Y2K.
It’s over; it hurt us.
And the third one that’s underway is the services transition.
Again, we didn’t quite see that all of these huge traditional ERP systems integration contracts that drive our services business were going to be put on the back burner for Y2K, and then put on the back burner again as people began to focus on e-business.
That transition is still with us, as is the transition of our portfolio restructuring, because we will not be out of DRAM until the end of the year.
And some of the decisions we’ve made on the PC restructuring will still be with us through the course of this year.
Having said that, having described those transitions over the last three years, we still feel very good about this year.
We feel very good about our strategy.
We feel very good about that the two transitions we went through were worth the pain.
They were worth doing, because we are now ready to exploit this e-business opportunity that we have been investing very heavily in over the last few years and we continue to invest in,” Gerstner concluded.