Irish Information Society Commission

Andrew S. Grove
September 20, 1999

DR. ANDREW S. GROVE: Good morning. My aim today is to share with you some thoughts about what is going to happen to all of us as we get embroiled in the world of commerce based on the Internet. And this is not an option; in my opinion, it is a mandatory progress for all of us simply because the economic drivers toward the Internet world -- Internet commerce -- are so powerful.

Let me start by setting forth a few thoughts about why that is so. And I will start by giving you a few comments about what I call e-Business Economics. My thesis for this subject has been, for some period of time, a fairly basic observation -- in spite of all the attention the so-called Internet companies get these days, it's a transitory period of time. In a few years' time, there will be no Internet companies -- there will just be companies -- and all companies that are going to operate in the economics of a few years, in the future, are going to be Internet companies. But that is going to be the mandatory way for companies to operate in the global network of commerce.

This is easy to say, and I would like to share with you just a couple of data charts giving some commendation for the statement. The first of them -- I apologize it's US-based data; I don't have a data collection like a commerce department for the world. What it shows is the top line - the yellow line - is the US gross domestic product, and it doesn't grow like it grows in Ireland. It only grows about 3% a year rate, but on a semi-log chart like you're seeing here, 3% annual growth -- you can't even see it. Below that in blue you see the chart showing the portion of GDP that, according to various estimates including the commerce department and some private consultants, represents the portion of the GDP that is taking place on the Internet -- Internet Economy. When you look at 99 corresponding to a hundred billion dollars, it is two divisions below the yellow line that corresponds to 1% of the GDP that's estimated to take place in the United States on the Internet. But notice how much faster the blue line grows and how rapidly the purchase, 10% of the US GDP by the year 2003; it doesn't take much of an imagination to extrapolate with your eyes those curves where the percentage in the latter part of the new century will be even closer.

Now this is statistical data -- aggregate data. Let me zoom in a little bit closer and give you some examples where I really do know the numbers that have to do with a portion of Intel's business that is conducted on the Internet. In 1997, the percentage was zero. We started a project in 1997 moving on commercial transactions on the Internet. By 1998, we reached 20% of our total. In 1999, we're likely to see 40% of the total, or 1,214 billion dollars. If I go back in the previous chart, I notice that in 1999, for the total US, the number is one hundred billion dollars. If 40%, or 12 to 14 billion dollars, that Intel alone does on e-Commerce, represents about 12 to 14% of the US e-Commerce, the GDP seems awfully high. And since I'm quite sure of this number, it suggests that a previous number is probably understated, if anything else.

Now why is it that the blue curve progresses so rapidly? Why is it that the Intel chart moves as rapidly as it does? E-Commerce and e-Business are attractive to both buyers and sellers. For buyers, it turns the market for goods into a very, very efficient marketplace, as efficient as the public stock markets. You're able to apply the same principles as people have applied to financial securities to the purchase transactions dealing with all kinds of goods. The efficiency to sellers is an internal efficiency. It is not so much that sellers particularly like to operate in efficient markets, because pricing in efficient markets tends to be lower than the pockets that can be maintained in inefficient markets. But the internal efficiencies you gain by moving your enterprise to an electronic basis are enormous. And that gives us the starting point for discussing in a kind of pseudo mathematical term why e-Commerce is so valuable.

I want to take a step backwards and go back into very basic fundamental securities theory that says that basically the value of all enterprises is the value of a cumulative discounted cashflow from today to eternity. And if you bear with me for a simple mathematical shorthand for that, the value of any company is the intergrowing in terms of same, the summation in terms of time, or the discounted cashflow of the business from here to eternity. This is true today, it was true yesterday, and it's going to be true tomorrow, no matter how the commerce is conducted. E-Commerce doesn't change it, but e-Commerce does change that discounted cashflow that we are summing up year by year going to the future, and it introduces an uplift to both -- to two components of this kind of cashflow. I call those an external component and an internal component, and both of those improve.

So let me kind of give you a graphical illustration of how this might work. If I take an old fashioned business I call the bricks business and measure its value and discounted cashflow potential, it has two components - two ways that we can improve. We can reach more customers and move the top of this block further up, dare I say reach, and we can make an internal business more efficient and move the right side of the line further to the right and we can increase its efficiency. So existing businesses constantly work on both of those; they increase their operational efficiency by moving -- by improvement on top of improvement -- that line to the right. And they constantly go after more customers by marketing the products, by selling their product, by advertising their products and reaching more and more customers. This is how ordinary bricks business operate.

In the last five years, we have witnessed a whole lot of companies that were born on the Internet and fundamentally grew into existence on the basis of the possibility provided by the opportunity. I call those the "clicks" businesses. And the clicks businesses are very powerful and have a powerful benefit out of the Internet. They can reach companies all over the world because they operate on the Internet. And in a few minutes we're going to talk to Fran Rooney from Baltimore Technologies, a perfect example of a clicks business and a company that uses the potentials of the Internet to reach customers very effectively all over the world through electronic networks that they operate. But clicks businesses have internal efficiencies also and inevitably, as they grow, operation -- internal operational considerations -- come in and that effort is going to be focused on moving their right boundary toward more efficient internal operations. And if you want a live example of that, even though most Internet commercial companies that traffic in goods were using basic infrastructure, today there are 60 to 100 million square feet of warehouse spaces being built or plan to be built by US business, e-Commerce companies. A more close to home example is that two major warehouses, almost totally a million square feet, are being built by Amazon.com outside London alone. All of these are done to move the line to the right to improve the internal efficiency of Internet businesses.

But the future is not in bricks businesses and the future is not in clicks business, but the future is in like a -- if I can steal a phrase from the co-CEO of Schwab -- clicks and mortar business that combine the possibility of the electronic infrastructure for reach, and the internal efficiency of the existing business such that this area under this curve, this block, can be improved in both the lateral direction, the efficiency direction, and in the vertical direction using the Internet for customer services. The power of the combination of these two methods of operation is so tremendous, the companies that do not avail themselves of either one dimension or another will be at a very major substantial structural disadvantage, a competitive disadvantage, versus companies that do. And Vivian was kind enough to refer to my book. My book "Only the Paranoid Survive" is about a phenomenon called a strategic inflection point. It's bending a curve where, depending on the action a company takes, you either proceed upward into new levels of operational excellence or downward towards marginalization. I submit that all businesses, whether they are bricks origin or clicks origin, are today at a point of choice such as strategic inflection point and, depending on their embrace of the two elements that we've talked about, they'll either write new competitive strategies or they'll be marginalized, bringing upon the general thesis that I started this presentation with that all companies that are going to be around will be doing both.

Let me give an illustration with a brief video of the company, WOLF Garten, specialists in garden products. Here they are talking about how they are using the power of the Internet to achieve, in every case, the strategic inflection point.

(Cue WOLF Garten Video)

DR. ANDREW S. GROVE: That completes this part that I was going to talk about -- the rationale -- why it is such a powerful force and why it's inevitabe for all of us in business to embrace in our way the power of the Internet.

I would like to shift gears now and talk a little bit about how companies might go about implementing e-Business, and I'd like to break the phases of implementation into three phases. One of them is building the infrastructure for Internet - for e-Commerce. The second one is modifying the non-electronic processes to take advantage of the infrastructure that you build. And the third and most advanced phase is something that the gentleman from WOLF Garten, Mr. Wolf, referred to in this video; it is beginning to exploit the data and the feedback which come in from your customer base to actually modify the products and services that you offer.

Let us start by talking a little bit about Phase 1, and I'm going to use the example, which is Intel's example, where, in a two-year period of time, we built an infrastructure that's capable of being the base for almost half of our total commerce with our existing customers that I mentioned earlier. The progression that we faced, that we used here, is first we took advantage of the implementation of an all electronic relatively modern infrastructure for our back office software. That's been going on since 1992. We started that without regard to the Internet because in 1992, frankly, not even Intel was particularly conscious of the possibilities that it would bring in just a few short years. But as it became presence -- as the Internet and its incredible reach, cost-effective reach, became relevant to us, parallel with the back-up solution of the modest software -- we started experimenting with putting marketing messages on our web presence. Then, in 1997, we decided to bring those two together and marry our web presence with our back office implementation in a unified fashion. The key thing that we've adopted here is you have to be very mindful of who your customers are -- who you're trying to aim all this effort at -- and we made a conscious effort to use our approach to improve our internal efficiency with our existing products and our existing customers. Correspondingly we set out to marry the electronic capability presence of our customers or help them establish such electronic capability that they could become part and parcel of the Internet economy web, meaning the word in a slightly different phrase. We have expanded this to tens of thousands of customers, all of whom are capable of using the same unified infrastructure. And obviously we will not be able to describe the details of it, but I'd like to refer you to a case study that's available, of course, on the web, at www.intel.com/business/.

Phase 2 involves, as I mentioned earlier, modifying the actual non-business processes -- not electronic business processes -- that existed before the implementation of that Phase 1 infrastructure, modifying them in such a way that they can take advantage of the infrastructure that is in place. And these range over a gamut of approaches, from the very basic automating of the tactical operations like simplifying the order-end process using electronic order entry forms without rekeying and without telephone help, that go directly to the customer database. That's the obvious one and it is absolutely -- I'm not minimizing it - it's a very important one, implementing this -- that led Intel to the average time that it takes to enter and order from 15 minutes before web implementation to three minutes after web implementation, and eliminates the kind of -- the more equivalent or actual literal equivalent of -- having our customers on hold while the customer representative consults different computer databases to answer the availability and give answers regarding the availability of certain product while the customers stay on hold. All of this can be eliminated so it provides not just efficiency to ourselves intramurally, but provides efficiency to the customers that we're dealing with in this fashion.

But I want to tell you an anecdote, which is a less obvious one. It's not a transaction-oriented one. Intel provides a great deal of confidential documentation to our customers. We basically educate our customers how to use our products and on how to use them in combination with other products. Some of them involve fairly confidential information, and controlling this information so that it only goes to our customers and not to our competitive customers is a major task. We contain this kind of information by distributing 142,000 confidential documents each year. Our sales engineers deliver this information to the customers and have the customers sign a receipt for it, and we keep a database of this. It took a very modest 56 sales engineer days per year to distribute these documents, and that's the number I was given. I think that's probably a low estimate. By putting these documents on the web and on the suitable security control, allowing the customers to use their security clearance, if you wish, to download those documents, those 56 days became eight days per year and the sales engineers were then free to do what they were hired to do, which is selling, instead of doing document control. Again, this is an internal efficiency.

In a similar fashion, we gave a very valuable efficiency to our customers because, particularly in areas away from major metropolitan areas where we have local sales coverage, these customers had to wait until the sales engineer would visit them. For instance, Asian countries where we don't have sales offices everywhere there's customer locations -- sometimes they'd have to wait for weeks before the sales engineer could get there with the latest information or updated information. These customers are getting the information statement. To achieve these types of savings, whether it's in transactions or whether it is less obvious activities, requires the merging of a variety of internal disciplines ranging from the sales accommodations to the information systems organizations and each of them bring -- have to bring -- their particular expertise and relate to the expertise and needs of the other party. It is not a simple process. It's a difficult process, time-consuming process, but this is where the pay-off comes.

Very few companies operate what we call Phase 3. That's where we're all heading. It's where we gain competitive advantage out of the customer data that we have in our databases and turn it around and use it to offer product offerings -- service offerings -- on the basis of our customers by their actions, not necessarily by their comments, indicate what they want. Just to give you two examples of companies that have pioneered activities along this line -- one is Dell, which has gained a well-known and well-chronicled competitive advantage by using customer information on the web or other direct means to formulate their manufacturing plants and, therefore, is able to operate at substantially lower interprocessing costs than its competitors. Amazon.com is using purchaser information through various on-line experiments to assess customer interest in new offerings. This is how they decide to add new lines of business to their website and web offerings by having tested their customers' interest in advance. These are just very early implementations, but picture that the same way when web pages and web marketing have spread through companies small and large all over the world. If Phase 3 reaches all companies all over the world, what amount of internal efficiency is going to give way to the area of the curve under the click and mortar company direct angle that I showed you earlier?

So, to summarize, the three phases that we all have to go through in that order is: building infrastructure, modifying the existing business processes to deal with the data value of the infrastructure, and then heading increasingly towards exploiting the data for better implementation. And with that, I would like to ask Fran Rooney, CEO, Baltimore Technologies, to come up here and give us a few comments about how he sees the Internet as a provider of additional discounted cashflow for his company. Fran, would you join us, please.

How did Baltimore start in the first place?

MR. FRAN ROONEY: Just over three years ago, we saw a huge opportunity for developing in each business relating to the whole area of Internet economy. It was quite clear at that stage that the rate of technological stage was going to be so big, it was going to fundamentally change our lives and how we looked at business. For years what we saw was that the economy of the world had relied on an infrastructure of trust that would no longer apply in the Internet economy. For example, things like birth certificates, passports, I.D. cards and signatures would no longer apply in the new world. Those rules didn't apply and the new infrastructure was needed. So we set about building a set of products that would provide governments and organizations in this infrastructure where they could conduct e-Business in the Internet Economy in a secure way and based on an infrastructural trust.

DR. ANDREW S. GROVE: Infrastructure that would replace the physical means of identification and different means of building trust.

MR. FRAN ROONEY: So instead of providing, for example, the written signature, we had, for example, digital signatures.

DR. ANDREW S. GROVE: I understand you had a spectacular demonstration on that recently.

MR. FRAN ROONEY: This time last year, we had the American President, Bill Clinton, and the Taoiseach used our technology to digitally sign and electrically communicate. It was the first one signed by Heads of State.

DR. ANDREW S. GROVE: Heads of State are a limited market.

(Laughter)

MR. FRAN ROONEY: ...we had the President of the US, the Taoiseach, there's maybe one or two more.

DR. ANDREW S. GROVE: Who are your other customers and what do you do for them?

MR. FRAN ROONEY: Our customers use our technology in different ways. Some of our customers use our products to make money; they use our products to reach out to customers, to provide a more efficient way of reaching their customers, so they can actually make money, get to the market quick, sell the products in an easier way. Other customers are using our products to save money, using it in a secure way, making their business more efficient, being able to bring their employees together in an efficient way, and it's a principle that we've adopted in Baltimore using our own philosophy of making our operation more efficient.

DR. ANDREW S. GROVE: Could you give us some examples of your customers?

MR. FRAN ROONEY: One very good example is the Australian tax office. They're not saving taxes, but making it more efficient for their taxpayers to file their taxes, their tax returns, on-line. So essentially they're providing the website and the taxpayer can log on and fill out the tax return on line. So no need for paper anymore, no need for queuing up at offices and getting all the interfaces. It's all done on line in a very efficient way. And another example here in Ireland is Bank of Ireland for their banking on line solution.

DR. ANDREW S. GROVE: What's interesting about it is here you are, an Ireland-based company, supplying the Australian tax authorities with software. Does the Internet come into your reach?

MR. FRAN ROONEY: Absolutely. We have almost four hundred customers now in over 40 countries, but we're located in 20 different offices around the world and to keep in touch, and for all of our customers and partners and employees to keep in touch, we have a secure website. People log on, and, through the Internet, our own internal Internet, they can keep in touch with what's happening. One of the things in our business, we started in 1996 with six people, we have almost 500 people now and our market cap is 700 million dollars now in just a few short years. But one of the things that was key to us when we started - like I said, we started with six, sometime early '97 we had about 15 people, but to be successful in our business we had to go big. You can't be small on the Internet, you've got to be big. So one of the things we did very quickly was to build our own website -- very impressive website. We started to advertise, get it out there, let people know it was around. And I remember one time in early '97, we went to a conference, a show and exhibition in San Francisco, with 15 people in the company, and a lot of the attendees were aware of us through our website. We had a very impressive stand and we had ten people in the stand - two-thirds of the company. But a lot of people were going around saying yeah, that's Baltimore, they're big, they're big, they're in Ireland, ten people in the stand. What the Internet allowed us to do was be big, give the impression we were big before we were. And to reach out to potential customers that we couldn't even have dreamed about reaching at that stage. So I think fundamentally we provide products to help customers build their own businesses, but we've adopted the philosophy ourselves to use the Internet to be bigger, be more efficient and run our business and bring everybody together in a very secure way that provides that infrastructure of trust.

DR. ANDREW S. GROVE: Super, Fran. Thank you very much.

Fran told us about Baltimore's role, and I would like to end up by talking a little about what Intel's role in building this infrastructure, and in general, is.

And I have to go back a little bit in history because, as a result of the Internet, the same way as companies are born, existing companies have to change their self-perception, their charter, their own mission, and so did Intel. Intel's mission prior to the recognition of the Internet as being the single environmental factor that shapes our destiny today -- was to be the building block supplier to the new computer industry, meaning the personal computer industry. Intel's mission today -- Intel's role in e-Business -- is to be the building block supplier to the Internet economy. Subtle difference in words, but I would like to give you some demonstration of how these differences manifest themselves in our physical vision of ourselves. What I'm going to do first is we're going to go back six years and show you a 30-second television commercial that we used to introduce the Pentium® Processor in 1993. The purpose of the television commercial, of course, is to paint a picture of who we are, what our product is, but basically to communicate in visual terms our mission to viewers who are getting Guinness only watching with one eye. It's a quick demonstration.

(Queue Intel Pentium® Advert Video)

DR. ANDREW S. GROVE: Now what is the equivalent of that today? I cannot do it in a 30-second commercial, that is not yet, but what I would like to do is -- all of you presumably have hit that return key and sent off a signal to the Internet expecting to do something, download a piece of music through MP3 or whatever, maybe more interesting than that even, but I really wonder how many of you thought about what happens when you hit that return key and those of you that thought about it, how many of you can visualize, even in a sketchy fashion, what happens when the signal goes over through this mystical cloud that's the Internet? And that's what I would like to do. I would like to take you on a journey of what happens - a pseudo journey example of what happens with the bit stream when you hit that return key. You're going to go out to somebody that's called your local Internet service provider, you're going to go on the Internet backbone. You end up finally, let's say, with the MP3 example, at the data center that stores all the music that's available on MP3 and sends that music back to you. That's the process that we go through, through a number of steps.

And the first of those is what happens when you start with your PC: You send a request to the website, the PC takes your request and splits it into digital packets -- those little green things are digital packets that leave your house through the local telephone company -- wires provided by your local telephone company reach your Internet service provider. What happens inside the Internet service provider? So let's go inside. We blow that building open so we can see inside. What happens there is first a web server interprets your request, the bits move on from there. But the bits have a designation to go to the MP3 site in their own language. They have to be renamed in the same fashion that the post office renames a letter -- uses a letter in the US postal, you have zip codes, you have an Irish equivalent of that. They get relabelled, a traffic server figures out the optimal route of many routes, and the router uses the instructions of the traffic management server and sends the packet on its way.

Now the data leaves the router and heads out to the high-capacity Internet backbone, an example of which is what was mentioned by the Taoiseach earlier, goes to an access point, the bits get juggled around and routed again, broken up into different streams because the bits sometimes go in several different ways and go ultimately to the Internet Data Center where our song or whatever context is located, go into the application server, which is a bit computer containing a lot of microprocessors that process this request, does the serves, looks up in the database and finds the blue bits, which is what we're after, that's our song. The blue bits get prepared for shipment back to the computer. They're ready to leave the Internet service provider and now we're ready to see them come home, and they will retrace the same route that we sent the requester bits, the green bits, and they're heading through the access point to the ISP and from the ISP to our house, in a total elapsed time is less than a half a second.

Now, I'm going to do just one more thing and ask for your forbearance on this: I'm going to run through this sequence, without talking. You'll see it run a little bit faster, kind of moving a step closer to real-time lapse, again keeping in mind that the whole thing takes a half a second, 480 milliseconds. You can't see it that way! I'm going to slow it down eightfold, which is still a bit faster than what you saw. This is what happens once more in this process.

(Queue DVD Simulation)

DR. ANDREW S. GROVE: I'm sure you all knew this, but it is really a fairly complicated process and it involves using all kinds of silicon chips in preparing, processing, communicating, shipping that information. And Intel's role is to be the building block supplier for as much of that process that you have seen as is possible. In the same way that the TV commercial described how we intend to supply silicon to the innards of the PC, our current role is to, as indicated in this animation, supply silicon to everywhere, the bits traversing from the original access computer out to the Internet cloud and back need to be processed. We operate it in a variety of ways. Of course the part that we are best known for is in the access device or the client computer where we are fortunate enough to have a significant presence in terms of our microprocessors coming from a previous mission. And in the last several years, we have figured out that the number of assets devices will sky rocket as more and more companies and more and more individuals get on the Internet and correspondingly the price has to come down, and we have segmented our product line to allow servicing the low-cost access devices as well as higher cost higher performance computers.

You have seen, however, that most of the action, once the bit stream leaves the computer, is in servers of a different kind and we have, in the process of segmenting our microprocessing products, developed a specialized line of microprocessors that are meant to operate very effectively and very high performance in different types of servers domain, web name application and all the types you have seen in this animation. And lastly, since networking and communication control processing is as prominent in this process as the actual data processing, we have accelerated our efforts to take the same kind of logic technology as we use in building microprocessors to build network processors and communication processors whose job is to be the traffic cop in shipping around all those bits, blue and green, on the Internet. We have undertaken a variety of acquisitions to accelerate the process of us becoming experts in this. We have bought several companies in the last year, all of them in one form or another related to communications. And we intend to continue with the boast of internal growth and acquisitions to make sure we can fulfill the mission statement as I indicated here.

More recently, we also realized that as so much of the action is going to be Internet, there's going to be a mandate of providing all those servers like you currently have at the ISPs. They will not be in a position to build their own data centers themselves. So we decided to enter into providing some of our own technology and some purchase technology in the form of a service organization called Internet Data Services, and these are the real servers that you see in a more figurative way in the animation -- they provide the muscle of these Internet Data Services. But again, as commerce becomes mission critical, it is no longer enough to provide raw server power. A second generation of servers offerings are going to be needed where the servers are monitored and managed in much the same way as large companies manage their data services operation through command centers in a real-time fashion with emergency back-ups, virus protection and other kinds of management services, so that our offering is trying to intercept the movement or manage services and provide the brain as well.

A major development in our company in the last several years has been the development of a brand new microprocessor, 64-bit microprocessor, the new generation, that is particularly aimed at operating with large databases, and we think it has the potential of becoming the engine for the Phase 3 class of e-Commerce that I talked about earlier. The first variation of this, a microprocessor came out of the labs a few weeks ago, is what it looks like, by the way. You can relate it a little bit to the graphics you saw earlier. The advantage of this is that because it operates with data on 64-bit chunks, large amounts of data can be contained in database and retrieved out of database, providing very rapid retrieval of customer information. It has a very powerful engine, accelerating encryption for Fran Rooney's type of security applications, as well as providing visualization, which is going to be mandatory as multidimensional databases are going to be the rule.

We are fortunate enough to be very early in the game of deploying servers in that worldwide server cloud that I used earlier. Just to give you some numerics; the server census, as best we can make out, for '99 is the bar, that low bar that I show under '99. As increased Internet -- the population of people and companies on the Internet -- grows by the year 2005, that is going to increase some more. As increasingly diverse services will be added to that, there is going to be further growth to that. The increasing sophistication of services can be calculated to require larger number of servers again, and lastly the critical dependence of corporations for the transactions on these servers will require certain amount of slack capacity and that will bring it up more. When you put it all together, 95-96 percent of our estimated server capacity, in the year 2005, is yet to be deployed, so this is a green field, field of opportunity for a lot of us, including Intel, and we hope to take advantage of new developments, particularly with the IA-64 generation processors.

So I'd like to sum up by reminding you, the destination for all of us, those of us that supply products for the infrastructure and those of us, sometimes the same ones, sometimes different ones, who use them to improve the reach and efficiency of businesses, is in the very early stages, but very critical stages. It is now that the decisions about what kind of future company we are each going to be are being made. And what I hope to do here for you is lay out in a fairly simplistic, and I apologize, over-simplistic way, the phases that we each have to go through one step after the other. Phase 1, which is the infrastructure, you can basically buy, that's available commercially - the bits and pieces of it and the integration services are available commercially. And Phase 3, when you get to it, will also be purchasable on the outside; companies provide database management software and necessary additions to that. Very important point, however, is Phase 2 is the critical phase where you adapt your internal operations to the infrastructure base put into Phase 1. You cannot purchase it, nobody can purchase it -- it is company-unique and every company has to exert the effort and put forward the effort to examine their operations and modify them in the understanding competition of the Phase 1 infrastructure. You can hire consultants and you can do all types of things, but ultimately that work remains yours and that's where the competitive differentiation in the future is going to come. Thank you for bearing with me and I will be very happy to answer questions.

* Other names and brands may be claimed as the property of others.