Few people are more passionate — or more insightful — about the Internet than IBM’s John Patrick. “Some people call me the strategist, some call me a visionary, and some call me chief dreamer,” he writes on his website. And in fact, Patrick’s strategies, dreams and visions have served IBM well — so well that they can be credited with having helped to rescue the company.
In the early 1990s, Patrick was one of the pivotal individuals who helped a stagnant Big Blue to reinvent itself by embracing the Internet and e-business as bedrocks of its business plan. Having enlisted the support of CEO Lou Gerstner, Patrick went about evangelizing to the whole company the importance of this emerging medium. In papers and speeches, he preached the need for IBM to revamp its internal communications, external marketing and business strategy. His wakeup call served as a vital stimulus for a company stuck in a morass of mounting losses, deepening divisions and falling stock value.
Even today, with IBM having wholeheartedly embraced the Internet (and reversed its flagging fortunes in the process), Patrick is still proselytizing. Recent passions include support for the open-standard Linux platform that IBM now uses in its mainframes; the creation of the “second-generation Internet” that will provide more bandwidth for a faster, higher-quality access; and preparing for a world of constant connectivity.
Patrick, who holds degrees in electrical engineering (from Lehigh University), management (University of South Florida) and law (LaSalle University), joined IBM straight out of the military in 1967. In his more than three decades at the company, Patrick has held multiple marketing, sales and management positions. Immediately prior to his current role, he had been the vice president of marketing for personal systems, a position in which he was responsible for creating the ThinkPad notebook brand. Earlier in his career he helped pioneer the IBM Credit Corp., one of the largest computer leasing companies in the world.
Outside of IBM, Patrick is a founding member of both the World Wide Web Consortium and the Global Internet Project, as well as a senior member of the Institute of Electrical and Electronics Engineers. For more on Patrick’s life and work, visit his extensive personal website, at johnpatrick.com.
This biography was written by Alex Low, a writer and researcher for The Next Big Thing.
- Business-related book: Crypto by Steven Levy
- Non-business-related book:The Bear and the Dragon by Tom Clancy
- Business-related website:Wall Street Journal Interactive
- Non-business-related website:Google
- Gadget:Kyocera SmartPhone 6035
‘My first computer was …” A TRS-80 Model I
“If I had to choose another, completely unrelated career, I would be…”:
– The owner and operator of a Harley-Davidson motorcycle shop
“How do get a Net attitude? It’s very simple. You talk to the kids and ask them what they’re going to be doing five years from now with the Internet.”
Jeffrey Rayport: They say — or at least it has been fashionable to say over the last couple of years — that large companies do not know how, or even if they think they need to, they don’t know how to reinvent themselves. IBM, obviously dominant in the mainframe era, saw PCs go by, arguably saw the initial phase of the Internet revolution go by, and then IBM reinvented itself. You played a part in that. How did you become the exception to the rule?
John Patrick: Well, I think any company can reinvent itself. It’s not really a matter of size. I would say it’s more a matter of attitude: How you think about what you can do or what you can’t do? And back in the early ’90s, early ’93 I guess it was, when Lou Gerstner came along, he had a very strong focus and a vision for the future of the company. And it really energized everybody. We got around the idea of the Internet in a very major way, and we saw it playing to the strengths of the company; we saw it as being something very real for business, which was quite a different view than most people in our industry had at the time. It was a very lonely spot, actually, to stand up and say, “The Internet is for business.” And then along came e-business as a theme to take that to the next level. And that really got the whole company around it. It became popular externally through our advertising, and it became a real galvanizing force for everybody in the company. And it still is. And we see it having quite a bit
more legs to it right into the next generation of the Internet and the next generation of e-business.
Rayport: Let me back you up to that time in the early ’90s, the beginning of the Gerstner era at IBM. One of the big ideas in the last two or three years has of course been the idea promulgated by Clayton Christensen of the Harvard Business School, the book The Innovator’s Dilemma, that essentially said that big companies tend to fall in love with the future function configurations that are relevant to their target markets and they chase those price points and chase those customers right up into oblivion and therefore miss whatever the threat is — the attacker, the new way, the new price point that delivers a similar set of benefits. You just said it was a lonely position to be in early in the ’90s. How did this culture, you, the leader like Lou Gerstner, how did you guys withstand the obvious gravitational pull of what the rest of the industry was saying was right to do?
Patrick: Well, I don’t think we avoided the gravitational pull of our customers on purpose. I mean, our customers spend a lot of money on information technology and we need to serve them and serve them really well. But that does not mean that you can’t also innovate. That’s the one difference I might see there –it’s not binary. It’s not like, well, we either serve our top 5,000 customers or we become a startup kind of a company. And in reality, size can work to your advantage if you have the right attitude about innovation.
Rayport: But you had to go to those big customers in effect and say at a time when they didn’t know how to spell Internet, let alone find a URL on the Web. You had to go to them and say, “Look, we are not going to deliver the same IT-based benefits in the same old way. We’re going to do it in a new way and we’re going to drag you kicking and screaming along,” as opposed to taking the path of least resistance with those accounts.
Patrick: We really said, “We’re going to do both. We’re going to enable you to fortify what you have, and we’re going to allow you to extend it through the Internet.” In the early days, that wasn’t necessarily perceived as being really with it, but a few years later it became to be quite profound what we had done.
Rayport: Meaning that in the early days, people would have said, “Trash the main frames, throw out the minis, move to a distributed convening environment.”
Patrick: Right. Throw away everything you have and start over. And what we said was, “Wait a minute, there’s some real assets here that we have and that our customers have that play to the real strengths of size, of the global nature of many of these companies. If we can find a way to Internet-enable that infrastructure and build it out to the Internet, we might have a more enduring approach than throw away everything you have, start out there where the Internet is and hopefully some day you’ll be able to link it into the core of resources of the company.”
Rayport: Well, it’s funny you said that. I remember three or four years ago running into a CEO in the Midwest who was thrilled with the realization — he wasn’t sure he was right but the hypothesis that he had at the time was, maybe a mainframe could be a server. And it was one of those funny moments where you said, “Of course it can be a server; it’s a computer.” But the idea was that mainframes were over, they were dead, they were dinosaurs, they could have no possible relevance to the Internet era.
Patrick: There was a time that people thought that, but it didn’t last very long, actually, for exactly the reason that we’re discussing — people began to realize that there are enormous databases, there are enormous transaction capabilities, there are enormous security infrastructures and networking infrastructures. And as the Internet reaches out from those enterprises and allows a whole new world of transactions, those numbers of transactions begin to skyrocket. And the importance of that back end becomes more and more critical. And each day that goes by, that’s still becoming more and more obvious, in fact, that people need a hugely powerful infrastructure. And as we move into the next generation of the Internet, with pervasive computing and all sorts of wireless devices and always-on connectivity and very high expectations from very large numbers of people, all that drives up the number of transactions, drives up the amount of data, drives up the requirement for a very highly reliable, scaleable, manage
able infrastructure. So, it’s an old ugly sort of word, infrastructure, but it’s starting to sound pretty sexy because it’s going to be essential.
Rayport: And that stuff does come in handy if you’re running a large enterprise.
Patrick: It really does.
Jeffrey Rayport: You are an unusual individual in the Internet age in the sense that you’ve been with one company, IBM, for several decades, 30 years plus. Was there a moment in the early ’90s when you thought, wait a second, it’s no longer going to be business as usual, we’re going to have to do something radically different with this company to continue to become successful?
John Patrick: Well, I would broaden that, actually, to say, as I think about it there have been three occasions in my short 34 years that I would say were really major shifts. The first was in 1981 with the PC. And the second was in 1991 with TCP/IP, the language of the Internet. And the third was in 1999 with Linux. And all three of those had certain things in common. They were grassroots initiatives. They were not tops-down drives by our company or any other company. They were very much community-oriented, standards-oriented, grassroots-oriented, bottoms-up. Secondly, in all three cases we saw a lot of venture capital, and a lot of smart people leave universities and their jobs at companies and gravitate into those three areas. And lastly we saw, in all three cases, a lot of companies in the information technology industry that said, “Who needs it?” In 1981 with the PC, of course, we were very much behind the first really standards-oriented PC, but at our core certainly people at IBM said, “Well, it’s inter
esting, it may be good for games, not sure what else, but it’s not a real computer.”
Rayport: Right. And you could argue that that attitude and the set of decisions that flowed from that created Microsoft.
Patrick: You could argue that, I suppose. Right.
Rayport: I mean in the sense that the commercialization on a widespread basis, the disk operating system was handed off to Microsoft, which created MS-DOS, which created an empire.
Patrick: Yeah, you certainly could argue that. Very true. And Digital at the same time said, “Why would anybody ever want a computer in the home?”
Rayport: Right, the famous Ken Olsen quote.
Rayport: Who could not imagine there would be any consumer use for computing at all?
Patrick: Right. And then in ’91, TCP/IP, of course it had been around a long time before that, but right about 1991 is when I would say it began to be taken seriously by some, not by many. It emerged as very interesting and some would say powerful in the research and the academic community, but from a business point of view, once again all the IT vendors said, “TCP/IP, who needs it? It’s not reliable, it’s not this, it’s not that.” IBM said it, Digital said it, Novell said it, Microsoft said it. Basically all the IT vendors said it.
Rayport: That’s like the entire industry.
Rayport: But I mean in fairness to all the people who made that statement, TCP/IP had been around for the invention of the ARPANET starting in the late ’60s, 1967. So, all of sudden to wake up in 1993, 26 years after this innovation had existed, at least in the basic sciences/military/university communities and say, all of a sudden a 25-year-old, 26-year-old technology or a set of standards and protocols actually has relevance to business. That sounds insane.
Patrick: It did sound insane. I remember when Bob Moscowitz from Chrysler Corporation — he was an IT manager or executive there; I’m not sure exactly what his job was. But I was in Prague, Czechoslovakia, at the Internet Society in June of ’94 and he made a speech about how Chrysler was going to move all of their networking to TCP/IP. And even at the Internet Society a lot of people went, “I’m not sure about that, whether it can really stand up to the test.” Then in 1999 we saw Linux, which again had been around for 10 years or so prior to that. But in ’99 it started to look serious to some, not many.
Rayport: Now why? What was it about the open-source movement at that point that all of a sudden seemed critical for mainstream applications?
Patrick: Well, I think it was a focus on things are just too complicated in this industry and maybe there’s something to this idea of being able to have something that’s open sourced that can run on all platforms and it potentially could attract the resources to really grow up and become significant. Again you had some IT vendors say, “Linux? You’ve got to be kidding.” Microsoft said it. “We have Windows. Why would anybody want Linux?” Sun said, “We have Solaris, why would anybody want Windows?”
But we looked at quite differently. We learned a lot about the experience of the PC and from TCP/IP. We saw those two movies and we said, “We’re not going to watch that movie a third time. This is big.” We really believed that it would change the event and still do believe very much so that this will change the whole industry. So, we said, “Well, we can watch this happen, or we can help make it happen.” So we made a fairly bold move and decided to not only embrace it but to contribute significant amounts of people and technology into the Linux movement. In fact, this year we’re investing an additional $1 billion in Linux because we do believe that it can change the face of the industry and address this issue of “things are too hard.” It can address it because the skills are available: Every kid from every good college, they know about Linux. I always say that those who doubt it, just go to Amazon or a bookstore and see what you find about Linux, see how many books you’ll find and talk to the kids.
Rayport: You mean, this is by no means a background phenomenon.
Patrick: Exactly, and so it’s time. We’re seeing now major supercomputer installations being built on Linux. The NCSA just made a decision for a major project to study the theory of relativity. It’s going to be the largest academic supercomputer ever. It’s built on e-servers from our company, along with Red Hat Linux. No one would have predicted that, even a few years ago. And in the commercial environment, major banks are starting to put their toe in the water. They see a world where they have this large number of servers and every time their e-business capability grows a little bit, more servers and it’s more “sneakerware” — you know, people running around managing all of these servers. They say, maybe we actually could put Linux on one of these mainframes. We know how they scale and have virtual instances of Linux. So, it’s maybe a little to early to declare victory on this but these very large customers taking a lead from a lot of academics are starting to see the possibility of having multiple virtual Linux environments supporting what they already have in the way of middle ware, data bases and communications infrastructure and start to simplify things. So, there’s a lot of potential here.
Jeffrey Rayport: You have been lionized in a lot of departments recently, everything from the popular business press, Fast Company magazine to the management tone volume manifesto that Gary Hamel wrote called Leading the Revolution. You talked before about the criteria you would have for knowing that you had hit an inflection point in history, be it ’81 with the PC, ’93 Internet, now Linux a year ago. How do you make a revolution happen, especially in one of the world’s largest and most successful corporations?
John Patrick: Well, there’s not a simple answer, but I think there are ingredients to that. Most importantly, you have to have support from the top to want to do that. It comes back to this attitude factor, and it takes a real commitment to allow some wild ducks. There are many heroes at IBM that have been written about in some of these things that you talk about, but that’s just sort of just one person. There are many, many heroes at the company that have built skunk works and pilot projects. A fellow over in Germany decided to put Linux on the mainframe and he didn’t have permission to do it, he didn’t have the budget to do it. He just did it, and people today in the company, they laud heroes like that.
Rayport: But when you talk about attitude, is a Lou Gerstner happy when he finds out that some guy is operating on the fringe without permission, without a license and doing stuff that is potentially radical, and by the way, threatens a bunch of the software that your very own sales force is pushing down the industrial channel?
Patrick: There’s a pretty good track record since Lou came on board, that in fact wild ducks do things that turn out to be not so wild after all. And I think another key ingredient here is to have a skunk works. We have a number of them in the company. They’re sort of hidden away here and there and they’re not managed too closely. They don’t have the same performance plans that traditional development groups might have.
Rayport: You mean in terms of hurdle rates or ROI targets–
Rayport: —or even revenue expectations?
Rayport: That would otherwise kill a lot of the innovation–
Rayport: That’s an oft-told story.
Patrick: Yeah, I often say that, and the little skunk works that our group operates, that the most important and significant things we’ve done have been things we weren’t asked to do. There were no requirements for them. And in fact, that had we asked should we do one of these, we’d have probably found no support to do it. We just did it.
And I think another important ingredient is to really, beyond the campuses, be involved with the key universities. Our group, for example, has a very strong relationship with MIT and we have a number of students from there that come into our group. A few years ago we started something quite unique called Extreme Blue, which is a recruiting programs designed to go to the top 10 computer science schools and find the very best students, sophomores and juniors, bring them into a small group into the skunk works, put them in teams of three along with a mentor, hopefully a wild duck from some part of IBM, a distinguished engineer, or a IBM fellow, or somebody that has a great idea that’s sort of a little bit on the fringe and get this group of students. And it’s unbelievable what they can accomplish in a summer. They don’t know it’s supposed to be hard, they don’t know that you’re supposed to need three years to do something but they only have 12 weeks, so that’s what it takes. So you listen to these kids, you watc
h them, and you talk with them and you’re around the clock with them during the summer and it just changes your perspective. And in fact, our chairman loves to meet with these students. He spent a half a day with them last summer just talking to them about their ideas and them showing him what they’ve accomplished and it’s magical.
Rayport: That is so interesting, because I think a lot of us are beginning to realize that the current generations of really teenagers, maybe college students — it would be a little bit too early today, but in three or four years — this is the first generation of human beings who are brought up with an instinctive relationship to digital technology rather than the one you and I have and all of our colleagues, which may be very savvy and sophisticated, but nonetheless, it’s a learned relationship. We learned it as a second language.
Patrick: Yeah, precisely. I’m writing a book right now about this very thing and it’s called Net Attitude, and in there I discuss this, how do you get this attitude, an Internet attitude, where do you get it from? And I say, well, it’s very simple, you talk to the kids and ask them what they think about the Internet and ask them what they’re going to be doing five years from now with the Internet. And if what they say doesn’t make sense to you, talk to them some more. If you can’t find a kid, borrow one. And if you can’t find any kids, talk to the seniors. This is a very–
Rayport: You mean literally, 70, 80, octogenarians?
Patrick: 60, 70, 80. I would back it up a little bit. I would say 45 to 102, it’s in that range, it depends but–
Rayport: Why? What is it about the two ends of the age spectrum that are so valuable to you?
Patrick: Well, there’s two different sets of things going on. With the kids, of course, they grew up with the games and so to them it’s just natural, and that’s just the way they are. With the seniors, it’s a different set of things that have driven them to have an Internet attitude. People in the 45 to 70, for example, in that range, a lot of them have lost their jobs so they have a need to form new relationships. Some of them lose a spouse; some of them are having grandchildren for the first time; some of them are making major changes in their lives, their career, family, etc. And all of these kinds of changes create … there’s a pent-up demand, they need help. Where do they go for it?
Well, they’re going to the Internet. That’s where they’re going. MyFamily.com, for example, is a site, I think it’s the top 10 sites now. And people go there to find others that have the same kinds of challenges and problems and needs, and they come together. So a little matter of “Windows is hard to install” or “the modem doesn’t work” or “I can’t get my cable modem or DSL” — they’re motivated, they make it work. They either get somebody to fix it for them or they just persevere because they want to get connected.
Rayport: It’s so funny because I think of the children, of course, it’s everyone’s experience: You want to program your VCR? Forget a PC, go find a kid. You want to figure out how to use your Palm 7 and get it online on the wireless network? Find a kid. But the issue with seniors is, it strikes me as potentially more poignant, which is the parallel universe made of bits starts to become very valuable when the analog or physical universe is less accessible to you. But it’s because you lost a loved one, or you can’t travel the way you used to, or you can’t do extreme sports because you’re 72 years old, not 22 years old. All of a sudden the ability to overcome distance and physical reality through online connections becomes very powerful.
Patrick: Extremely powerful, and there’s another difference as well. In addition to that, seniors are going to be much less likely to accept anything less than a great e-business solution out there. Today, e-business, for the most part, is not meeting people’s expectations. Christmas time, a study showed that 43 percent of people that went to a website specifically to buy something found it, but couldn’t buy it because they could not complete the transaction.
Rayport: You mean they found the shopping cart technology too complex?
Patrick: Either the shopping cart broke, or they couldn’t understand it, or they just could not navigate, or the server went down, or for whatever reason they could not complete the transaction. Oh, and also you find these cases where “click here for more information” and as likely as not, up may come a page that says, “Call us 9 to 5, Monday to Friday during our normal business hours.” You still find a lot of sites out there where you “click here to buy” and it says, “Print out this form and fax it to us.”
Rayport: It is so funny, I mean, you’re reminding me of the early days that we have laughed at for the last two or three years when a respected retailer like J.Crew would put a site up in ’96, ’97 on the Web, and when you wanted to look at merchandise, it said, “If you don’t have a catalog, get one.” This is not exactly a service-oriented relationship.
Patrick: And then you come back to the seniors. When the kids see these lame sites, they’ll just get mad and go away and say, “I’ll never go there.” But the seniors are going to say, “Why? Why can’t you? Why do I have to fax you a form? Why do I have to get in a car and drive to the bank to get my signature notarized? Why do I have to fill out a stock power form to transfer one share of stock to my granddaughter?”
And they’re not going to accept answers like, “Well, that’s because that’s the way we do it” or “That’s our business practice or policy.” So, I think it’s going to be very interesting over the next few years as we move into the next generation of the Internet, that seniors will have a real impact and they’re going march on Wall Street, or they’re going to go to stockholder meetings, or they’re going to demand answers for, why can’t this e-business meet my complete end-to-end needs? It’s going to be a very positive influence.
Rayport: Well, it’s funny, because in a way you’re talking about wisdom that we should have known a long time ago for a very simple reason: AOL. AOL had various step function increases in its user base over time. One of them was military users and military cities. But SeniorNet and senior participation, for exactly the reasons we’re talking about, was a phenomenon we saw in the late ’80s, early ’90s with the commercial online services.
Patrick full interview
Jeffrey Rayport: You clearly — if anyone has Net attitude, it sounds like you do. How did you manage to stay at one company instead of doing what would seem to be the obvious thing, which is leave IBM behind, move to San Jose, become an entrepreneur?
John Patrick: Well, I don’t want to move to California for thing, but the real reason is that IBM has kept me challenged. I’ve been very fortunate to be involved in new things throughout my career, and in particular in the last six or seven years, it’s been sort of nonstop Internet and it’s just been a wild ride. And the support from the top has been fantastic. Lou Gerstner got the Internet, like, instantly. When we first showed him a webpage back in late ’93, early ’94, I’ll never forget it, we showed him this page we had marked up of IBM.com. We weren’t the first big company on the Web, but we were one of the first to — we had a lot of firsts actually with regard to the Web. We showed that first page to Lou, he looked at it and said, “Where’s the buy button?”
Now this is 1993, early ’94. There was no Amazon, there was no SSL, there was no e-business, there wasn’t even e-commerce. And here’s a guy that looks at this, because he saw it obviously as a communications potential, and it linked obviously in his mind to transactions. Nobody else was thinking that way. So, that kind of top-down support made it a lot easier for many of us who were really a virtual team of Internet people. And we began to discover each other and we found that there were a lot of people around the company that had ideas, technologies, and so it’s hasn’t been a dull minute for the last seven years.
Rayport: We all know that the public hysteria, the fever pitch, especially on Wall Street, around Internet issues that came after that moment, starting perhaps with the Netscape IPO or the Amazon IPO in spring of ’97. That burst — whether it’s a bubble or not, it’s gone. Has that changed for you the climate in which you do this kind of radical thinking or promulgate innovation within a big corporate environment?
Patrick: Not at all. There is some e-backlash, so to speak. But you really have to put it in perspective. On average, there’s about 5,000 bankruptcies a month in America, just in America. For the last 20 years it’s been between 50,000 and 60,000 bankruptcies per year. So, last year we had 221 dot-com failures. And they all made the headlines. One of the first to go under was Boo.com, a retailer in the UK. And I vividly recall a Wall Street Journal story that said Boo.com bankrupt, maybe the Internet is in trouble. And of course, it has nothing to do with the Internet.
Rayport: Right. The real story was maybe, a couple of 20-somethings burned their way through $127 million and did it with Concorde flights and servings of champagne and caviar. It had nothing to do with technology.
Patrick: No, nothing.
Rayport: Or e-commerce.
Patrick: No. It has to do with the basics. And the basics are: What is your market? You segment your market; you look at how big are those segments that you’re intending to attack. What’s your cost structure? What prices do you set? What’s your distribution strategy? How are you going to deliver great customer service? Those are the things that determine whether you’re going to be successful or not. It’s a given that all companies will be on the Internet, all companies will be in e-business. If they’re not, that’s like not having a fax machine.
Rayport: But some of this it is the French thing plus ça change — the more it changes, the more it stays the same. And you’re saying, “Look, it’s not about the Internet — it is a tool for business. It’s about a bunch of business fundamentals that never went away.”
Patrick: Right. And the fundamentals are now getting a lot more focus, and the Internet is changing the way you think about the fundamentals. The Internet changes the way you think about your distribution channel; it changes the way you think about your agents in an insurance industry or your brokers in a financial services industry, etc. I think the good news here is, frankly, that the, let’s call them the existing companies of the world, they get it now. And they’ve looked at the experience of these dot-coms, and some of them maybe are breathing a sigh of relief that there are some dot-com failures, and they say, “Well, OK, now I don’t have to worry anymore.” But most of them, I think, realize that we’ve got to hurry up here and leverage the infrastructure that we have, both the IT infrastructure as well our physical and people-oriented and relationship-oriented infrastructures. And, in fact, we can do a lot better than what they tried to do because we have the brand, we have the feet on the street, we have all that infrastructure.
Rayport: You just talked about the reaction that some boards, not the smart boards of directors, but some boards have had it, which is great. There’s a meltdown … checks a box on e-business, we’ve spent like .002 percent of revenues this year, we’re fine. Some of the data about the consulting industry related to the space, the Internet professional firms, would suggest that in fact a lot of companies have put the brakes on with a sense of relief, which is, we don’t need this kind of capital intensity anymore to respond to the e-business threat. You obviously see this through the IBM global services consulting arm. How real is that? And how many companies have in fact taken a false assurance out of the crash of the Net sector or the e-commerce sector recently?
Patrick: Well, I’m sure some have, but I don’t think very many. Certainly, the customers that we deal with — the households names of the major institutions around the world — we don’t see any slowdown going on in our consulting business. To the contrary, it’s quite strong. Our e-business services business is probably the fastest-growing part of the company. It’s 50, 60, 70, huge growth rates.
Rayport: And even now in the midst of what some people call the nuclear winter of the e-business era.
Patrick: Yeah. What we’re reading about in terms of what might on the surface look like a pullback in e-business services are the boutique services. The boutique services focused on “how do I build a web page?” or “how do I do this one little aspect of things?” The real growth and the real spending is going to occur not on the fringe, but in the integration. How do I make my frequent-stayer point system as a hotel chain relate to my reservations system? If I’m an airline, how do I make my gate scheduling system meet up with my flight arrival system?
Rayport: You mean, and now we’re not talking about the interface anymore, we’re back into the world of middleware, large systems?
Patrick: We’re talking about middleware. We’re talking about, right, exactly, the holy grail of e-business is integration. How do I make the systems — these subsystems that were built on different platforms in different decades — how do I make them talk to each other? How do I utilize things like message queuing, code name for like e-mail between systems to be able to exchange messages, not between client and server but between server and server, to be able to stitch things together? The old approach to that was business process reengineering. The way we make these systems work together is we reengineer them. Well, that’s a long process, maybe 18 months or longer. Not to say that business process reengineering is something you shouldn’t do, but there’s a much faster path, which is to link these systems together through messaging. And so, these kinds of subjects are in great demand, both the middleware itself and, in particular, the services for how do we do this and how do we do it fast.
Jeffrey Rayport: John, I hear you essentially talking along two streams. One is that there are a bunch of business fundamentals that never change and haven’t gone away. In fact they’re probably more true today than they ever were, more essential for management. But the other one, which is, there actually is something new under the sun and it has to do with Net attitude and connectivity and messaging and ability to do a bunch of stuff we could never do before. What’s reality there?
John Patrick: Well, the reality is that the way in which IT applications are constructed is about to be revolutionized. It’s going to go through the biggest change we’ve seen in decades as a result of a new set of standards emerging on the horizon called Web Services. This is not like IBM Global Services, or Microsoft Services or any other vendor. It’s an industry approach to utilize new standards which allow for applications to be created which can be residing in a — think of it like a yellow pages in the sky. And let’s say, for example, a company might be an expert at creating a sales tax calculation routine. And they’re the world’s experts in sales taxes, they know every one of the 85,000 taxing jurisdictions and how they work and what’s taxable and what isn’t. And so, they create this application that calculates correct sales taxes. And they pushed it or publish it into this directory in the sky using a new standard called WSDL. It’s a lot of alphabet soup here, but they put it in this yellow pages. Meanwhile, another company wants to build a new e-business application which has a dependence on calculating sales tax. And well, in the old days they might have either written that themselves, or acquire some monolithic e-business or e-commerce application program that includes this sales tax routine. But in the new generation, they may discover through, there’s a standard that allows you to discover this yellow pages and you find that somebody’s already done this and, in fact, they offer it as a service.
Rayport: So, that unlike a yellow pages which is actually at the end of the day just a listing, this is actually a set of utilities.
Patrick: It’s actually, right, it’s a utility–
Rayport: It could be one from the yellow pages interface as the entire solution.
Patrick: As the entire, this particular
Rayport: I mean the plug-in.
Patrick: The plug-in. You think — right, exactly — you think of it like a little Lego that does this particular function. And there’s a standard called SOAP — our industry is so great at coming with these user-friendly, easy-to-understand-and-remember terms — but SOAP is a Simple Object Application Protocol. What it means is that it allows you to call upon that utility that’s in the yellow pages and actually execute it within your e-business. Now, what this means is two things. For application development, it changes everything because now with this new standard, people will create things, put them in the yellow pages, and it’s platform neutral and it’s language neutral. This is something we’ve never had before. The user of that application, they don’t care if it was written in Java or COBOL or, they don’t care, Fortran, anything.
Secondly, from the user point of view — the end user, the consumer, or the purchasing agent in an industrial sense — they’re going to see websites that do a lot more and have a lot more function, a lot more end-to-end capability. To them it’ll just a be a webpage, “click here to do something.” But when they click here, that’s going to set off a series of interactions server to server where all this functionality comes together and brings the result to the person. To them it’s just one webpage, but behind the scenes we’re going to see an explosion of communications and transactions and storage. It’s going to quite change the way IT works today. It’s going to disaggregate the monolithic nature of IT that we have today.
Rayport: So, you’re really talking about a world in which enterprise IT infrastructure or architecture becomes maybe not a patchwork quilt, but something equivalent, say, to what a microcircuit chip looks like where you’ve got a bunch of things plugged in and they’ve got to all be made by the same company, or I should say, a motherboard, something like that, where literally, the functionality of a given company’s infrastructure or a supply chain, and industry supply chain infrastructure is made up of hundreds, maybe thousands, of operating programs and capabilities that make up a complete singular solution.
Patrick: Exactly. Yeah, it’s radical.
Rayport: How close are we to that?
Patrick: Well, it’s moving very fast because it’s so compelling. The first part of this was establishing standards, and there are four standards that make this possible. The first and most important is XML. And XML, as we all know by now, perhaps, is a standard which gives the web context. It provides a way to put little tags on the back of a webpage, as a way to think of it, that describe not what the page looks like, but what it contains, what it means, gives it context. And using this capability of tagging information, it’s carried to another level to create a standard for this directory in the sky, it’s called UDDI.
Rayport: So, we’ve moved from the hypertext markup languages to XML (extensible markup language), which now points to this higher level of interoperability that you’re describing — now we get another acronym.
Patrick: Right, right. And then the third piece of this is this capability — I’m embarrassed to use all this alphabet soup, but it’s unavoidable — is the ability to take an application that exists today that was written in whatever language, maybe it’s mainframe COBOL, it might be PC Fortran, it might be whatever. And to be able to wrap it in such a way, to put a wrapper around it to make it neutral. So nobody cares any longer whether it’s Java or COBOL, PC or mainframe, here or there, it doesn’t matter. And then lastly, is this thing called SOAP, the ability to call upon this routine. So, it’s the ability to publish it into the directory first, then to be able to find it in the directory, and then to be able to what we call bind it, to be able to include it in application software.
Now, the status of this is that this concept of Web services, the key standards have been led by Ariba, IBM and Microsoft, along with 37 other companies. So, there’s a large number of companies behind this and there’s agreement that these standards are going to remain neutral. And this is breakthrough. We’ve never had before an approach that had this degree of neutrality. There have been many attempts, but the attempts either didn’t include Microsoft, or they were Java only, or they were just the UNIX guys, but this is a very broad group of companies. IBM, Microsoft, Ariba and 37 other companies — it’s a very broad mix. And so, it’s out there and people are beginning to experiment. And I think we’re going to see, this year, some very serious moves. We certainly intend to. We’ve adopted these new standards in all of our middleware so that people will be able to just build out from what they have.
And this is particularly exciting because companies that already exist, they have a lot of software. They have billions of dollars of investment over decades and application software that does things. Without rewriting it they will be able to take this application functionality and put this wrapper around it that makes it neutral and put it into this yellow pages, either public yellow pages or they can create their own. And this will change the way in which people collaborate. It will allow for new levels of make/buy decisions. It will allow for a lot more functionality to be shared and to be included and everything gets better.
Rayport: It’s interesting because it, the way you’re describing this, you’re essentially talking about a way to build operating systems at the highest level for business — widespread, large-scale business enterprise. It sounds a lot like the open source move and the arguments for the Linux approach that says that hundreds, thousands, tens of thousands of programs all over the world will collaborate.
Patrick: It’s related. This idea of Web services is not necessarily open source. There will, I’m certain.
Rayport: You mean in the sense that many of those objects inside the wrappers are actually black box objects one way or another.
Patrick: Yeah. There may be a lot of, exactly, there may be a lot of proprietary content underneath the covers, but people won’t care because what they’ll be interested in is, “What does that piece of code do? What are the inputs to it, what are the outputs to it?” For example, inputs are the taxing jurisdiction and the amount of the purchase. The output is the amount of the sales tax. You have two inputs and one output. Another one might have one input and five outputs — there’s a lot of possibilities here. And that’s what people will care about, and since it’s neutral, they really won’t care what language it’s written in.
Jeffrey Rayport: So, five years from now will businesspeople be talking about the Internet and the Web? Or will they be talking about something entirely different?
John Patrick: Well, I think what’s going to happen there is, we’ll begin to think of the Internet and the Web as different things. Today, most people think they’re the same. I get a new PC, I plug it in the wall, I plug it in the phone jack and I’m on the Internet. When they say, “I’m on the Internet” they mean, “I’m on the Web.” Well, what’s going to change is that first of all, you won’t anymore think about logging on to the Internet, you’ll just be on the Internet. Just like you don’t log on to the power grid to use your toaster, you won’t log on to the Internet to do a lot of different things.
Rayport: So, for those lucky college students who have T3 connections through their dorm rooms or the small percentage of consumers who use cable modem or DSL lines, the always-on phenomenon will become the experience of the Net.
Patrick: Yeah, it’s very profound. And it’s not such a small number, it’s growing very rapidly. There’s about 6 million in the US right now. There’s 6 million in Korea. In Singapore they have fiber optic capability to the curb of every neighborhood. Taiwan has DSL capability to every residence in the country. Not that they all have it, but they all could have it. So, it’s growing quite rapidly and the implication of always-on is significant. It changes the experience, for sure, but it also changes the way you think about the Internet.
Since my house is always on, yeah, I can surf the Web while I’m in the house, but I can also connect to the house when I’m not in the house. So, I can be riding to work on the train, for example, and I realize I forgot to put the garage doors down. I grab my phone, I connect over the Internet to my home, I push, make a selection, door down, I push the button and down goes the door. Now that had nothing to do with the Web, but it has a lot to do with the Internet.
And so, we will begin to think of the Internet as a communications infrastructure that allows us to communicate, that allows us to monitor our father or mother’s pacemaker. It allows us to receive real-time streaming weather data from the roof top of our house. To see what’s the weather over at the vacation home. Or is that garage door down? Did I close the blinds? Or in an industrial sense, what’s going on with the city water supply or that oil pipeline or other kinds of real-time industrial data that is streaming. So the Internet becomes the communications mechanism. Instant messaging, by the way, is a utilization of the Internet that doesn’t use the Web.
Rayport: What you’re saying, John, is in effect, the Web is today, in most of our minds, the dominant interface for tapping into or accessing the connectivity of the Internet.
Rayport: You’re describing a world where, whether it’s the instant messaging screen or it’s the toaster or it’s the garage door, it’s the cellphone, or any number of other devices, all of these become interfaces, many of which have nothing to do with browser software that we associate with the Web.
Patrick: Yeah, nothing to do with the Web. I go even further to say that I believe that the Web browser will recede to where it started. Web browsers are great for browsing. No big surprise to me, I just saw in one of the major papers in the last day or so a study that said doing taxes on the Web isn’t as good as doing taxes with software on your PC. Of course not. The PC — and when I say the PC, that’s either with Windows or a Linux desktop or it might be a non-PC device, but whatever our sort of environment is that’s local to us — it has a look and feel that we become accustomed to and it’s not dependent on the network. At points in time we use the network to update information in both directions, but when we want to interact and do things, pay our taxes, pay our bills, do various forms of banking, any kind of transaction-oriented things, a browser doesn’t really feel right to me for that purpose. I want to use an application that’s built for the desktop of my choice and also has the communications capabilities over the Internet that a Web browser has.
Rayport: Meaning that it reaches when it needs to reach out.
Jeffrey Rayport: You’re saying that, and it sounds paradoxical to me, that in the presence, on the eve of a world that is ubiquitously connected, always-on, and maybe even wirelessly always-on, we’re talking about a renewed focus on local applications.
John Patrick: Absolutely. Absolutely.
Rayport: Does that make sense?
Patrick: It does to me.
Patrick: Because first of all, the always-on is going to be there, but not for everybody and not all the time. And secondly, it’s this look-and-feel aspect. Let’s take e-mail. So you’re in your inbox and you’re using Microsoft Outlook Express or Exchange or Lotus Notes or Eudora.
Rayport: Like any number of the e-mail clients that live in our laptops or PCs.
Patrick: Any of them, any of them. We all have them on our laptop or on our PC. And so, you’re going through your inbox and you want to delete a message. You hit delete, you go to next one. You hit the enter button, boom, there’s a message, you forward it. We get a rhythm, right, we know how it feels, we know how it reacts. Now go to Hotmail, or not to pick on them particularly, but any Web-based–
Rayport: In other words, any Web-based e-mail.
Patrick: Any Web-based e-mail. OK, click delete, whoops, well, nothing happened. And I see that I’m waiting, what am I waiting for? Well, I’m waiting for the network. Well, I thought the network was fast. Well, it is fast but it’s not always fast for everybody all the time, everywhere. Now, someday it will be, but we’re not there yet. And it will be a considerable period of time when we will have an asynchronous set of bandwidth environments where it will not always be fast, or it will be slow in one direction, fast in the other direction. And so, for things like deleting e-mails or working with your e-mail, Web-based approach is not necessarily the best answer. And I see it all the time, you read about e-businesses that say we’re Web-enabled, it’s like we’ve done everything. Well, OK, but can I still pay my bills on the airplane? And this is going to become, I think, a new focus over the next year or so.
Rayport: Is it fair to look at some of the wireless devices people are using in large numbers to access Net connectivity? IMode from DoCoMo in Japan being an amazing example, now at 17-plus million users. It’s happened in a space of less than 18 months. Is that a vision of the future?
Patrick: Oh, very definitely. In fact, our vision of the next generation of the Internet encompasses seven characteristics. And we’ve talked about some of these already, but fast, of course. But then always-on, everywhere, natural–
Rayport: Meaning wearable, miniature, part of me?
Patrick: It can be, for some people. This natural part means language translation, it means asking a question in a language that I prefer, not necessarily the one the website prefers, and having it get to the expert for that question, who answers it in whatever their language is, but I get it back in my language.
Rayport: So, a sense of whatever is natural language–
Patrick: Natural to me, right. Natural is having digital radio, it’s having all of my music collection in my pocket on a microdrive or other methods of storage. So then, natural, then there’s intelligence, and this has to do with what we discussed with Web services. XML brings intelligence, not just a publishing medium but putting structure and context into the Web. And then No. 6 is making it easier, that’s where Linux comes in and we talked about that quite a bit. And No. 7 is trust, which has to do with digital IDs and authentication, which is a very important subject.
But back specifically on your question about wireless. Wireless, without a doubt, everyday it becomes more important, but it’s different in different parts of the world. In Asia, as you pointed out, it becomes more of a social phenomenon with the kids not only using DoCoMo to talk to each other and send e-mails, but to send their pictures of themselves to their friends. Little black-and-white pictures on narrowband capabilities is not third-generation wireless, it’s today’s wireless, it’s slow. But it’s fast to send a little picture and that’s what the kids are into.
Now, in Europe, it’s evolving differently. In Europe the phones all have a little chip in it, a little SYM chip that stores the electronic ID of the phone and the user. That chip can also store a digital ID. And so, they’re very well set up to be able to use the phone as an e-commerce enablement. To be able to walk up to a vending machine and push a few buttons and get the soda and have it charged through a micropayment or directly to your credit card. Then in the US, of course, we have sort of a hodgepodge of unrelated capabilities, but there’s hope here. This glass is half full. There are standards evolving, namely, XHTML, which replaces today’s mark up language with an XML-based markup language that’s device independent. And this is really crucial because people that create content today for the most part are creating pages for Web browsers on PCs. That’s probably where 95 percent of the target is.
Rayport: And given your vision of the future, that is one small piece–
Patrick: Small piece, less than half.
Rayport: — of that universe of interfaces that we will be relying on to access that content.
Patrick: Yeah, it will be maybe 30 percent, maybe percent. And the other, by the way, that’s not because PCs are declining. PCs are not declining. I can’t imagine giving up my PC.
Rayport: But that just means that the total pie is dramatically expanding.
Patrick: The pie is growing. It’s the growth of lots of new things.
Jeffrey Rayport: So John, what is the next big thing?
John Patrick: Well, there are many, but I guess the next generation of the Internet, those seven things. There’s a lot of meat under all those. If you said what’s the most significant, I would probably say Web services is the most profound because it changes not just for a user, but changes for all users the amount of function. It is what I would say is most likely to get us to the point where we have end-to-end capabilities. Most websites today, they focus on this one little sliver, click here to buy. And they’re driving to get the goods out of the warehouse into your hands. When you talk about end-to-end applications you start to think about, well, what’s the optimum packaging. I spend 10 minutes buying something on the Web, I spend 20 minutes unpacking it, going to the recycling center, cleaning the white polywhatever peanuts off of my clothes and off my front yard and the wind blows them all over the place. Who’s thinking about the whole range of things? From the moment I get the idea I need something to when I’m completely satisfied, that’s where the bar is going to be set. And Web services is going to be the way to bring a lot of new functionality to e-businesses without having them be required to write all that application software. They’ll be able to basically e-source those additional functions, pull it out of that yellow pages in the sky and build a very robust capability to meet our expectations.
And this is overarching thing going on there is that the Internet is transferring power from institutions to people. And the people have a very high expectation. They want to know why can’t I do this or that or the other thing. And they’re going to start asking those questions more frequently, and they’re going to look for e-businesses that can meet their needs, and they’re going to be very discerning about who they do business with. So this is a great opportunity to leverage the brand that companies have, tie it together with this functionality, tie it together with trust through digital IDs and make people really happy, and most importantly save their time. As we move forward the thing we’re going to be shortest of is our time. So, any e-business that saves me a lot of time, they’re my friend. And today we all spend a lot of time in lines. So it’s a bright future, I’m excited about it.