What’s the Story with Internet Stocks?


Internet Stock Madness and Internet Infrastructure


Ben Goertzel

June 10, 2000




What’s the story with Internet stocks?   Has the US stock market just been taking itself on a completely meaningless rollercoaster ride?  Things were booming till fall ‘98, then everything went sour for a while, and then things soared unimaginably, until just a few months ago, when the bottom seemed to fall out.  For a while, sexy Net start-up firms had been seeing their valuations double every three months on the basis of no profits, minimal revenues and infinite promise.  Now these shops are back to their valuations from nine months before, and more than a few have already gone broke due to the new, tighter capital climate.   Over the next 6 months, unless there’s a huge unexpected market boom, we’re likely to see a series of high-profile Internet bankruptcies.   But what’s going on?   Is it just the case of a speculative bubble bursting, of investors coming back down to earth and recognizing that they’d been duped into overvaluing a bunch of worthless hype?   A high-tech international silicon-powered tulip bulb craze?


Actually, the situation is much more interesting than that.  To understand what’s really been going on – and what’s likely to happen in the near future – one has to think fairly hard about the fundamental nature of the Internet … what it’s going to mean to the economy of the future, and what investors are able to know or guess about it today.  One comes up with a whole host of new concepts -- sticky customers, customers as co-developers, intelligence, distributed processing and profiling infrastructure.  As is often the case, what at first looks like insanity and chaos turns out on careful study to be intricately structured.  The Internet is a new world, and it’s not structured in a way that our standard cognitive toolkit, shaped by the old world, is configured to understand.  Part of the revolution that’s happening is going to have to be a revolution of our minds.


Clearly there’s been some hype-based overvaluation going on: the Net was talked about, so prices went up, so it was talked about more, etc.   But there’s also something deeper.  A technological and sociocultural revolution is afoot, and the financial markets, in their own peculiar way, have gotten wind of this fact.   People sense, intuitively, that the Net is the start of something huge and amazing, and they also realize that whatever company owns a part of this golden glimmering future is going to be amazingly rich.  They’re gambling that the Internet start-up they invest in today is going to own a part of the future.   And there’s real truth here: some of the puny little start-ups of today will be the behemoths of tomorrow.   The rest will be acquired or disappear.


To see to the heart of the situation, we need to take a step back, and look at the whole situation, economically, culturally, psychologically and of course technologically. 


First of all, from a large-scale financial perspective, the Internet stock frenzy is a consequence of the overall migration of capital from fixed-income securities to equities that we’ve seen over the last decade or so, which happened for very rational reasons: It became obvious even to 80-year-old grandmothers that the stock market was outperforming bonds and other more conservative financial instruments.  People decided to buy stocks, so the stock markets filled up with money, and there was an undersupply of stocks.   This in itself was enough to drive up prices somewhat, producing the basis for the Net stock frenzy.


Then there’s the notion of a productivity boom.  Every 50 years or so, it seems, technology creates big jumps in wealth.  Higher productivity leads to higher earnings for companies, and ultimately also to higher earnings for labor.   The historical record shows in the past, huge productivity-enhancing technological advances – factories, cars, highways, electricity, railroads and so forth – have led to huge restructurings of the economy, turning winners into losers and vice versa.  It’s quite rational to expect that the computer revolution will do the same.  Either Internet firms will benefit directly from beating existing companies, or new profitable industries will emerge, allowing new wealth to prosper alongside the old.   The point is, something real has changed; it’s not just a speculative boom.  The influx of money into equities occurred at just the right time to fund the firms pioneering the post-industrial revolution.


But where exactly is this revolution going?  As the great physicist Niels Bohr said, “It is difficult to predict, especially the future.”  How can anyone tell which ones of the zillion start-ups out there are really going to prosper in the new economy?   Everyone has their own guess.   And many investors, knowing they don’t know enough to make really good guesses, wind up jumping on bandwagons, picking stocks that others seem to think will ride the wave into the future.  Add onto this plain old speculation – people picking stocks they think others are going to pick, regardless of what they think about the company behind the stock – and you have what we’ve seen over the last couple years: a market that acts totally psychotically crazy, but overall keeps heading in the upward direction, as the net-centric future gradually creeps in on us, and more and more of the economy begins to rely on these highly valued technology firms.  It’s very hard to tell where fundamental optimism about the future of the Net begins, and  speculative silicon-tulip-bulb stock frenzy leaves off.  But what makes the current situation so interesting is the mixture of the two.


People intuit that something huge is happening, the early stages of a revolution, and their intuition is right on.  The spreadsheets of all these sexy start-ups may not look so hot, but on the other hand, everyone knows that the world doesn’t run on analytics alone.  Often intuition senses subtle patterns that the analytical mind misses.  And if the intuition’s a bit random sometimes, so what?  If you’re not sure where the target is, use a shotgun and you’re likely to hit something….


Amazon.com’s investors are well aware that there’s no way amazon.com could ever make enough money selling books to justify their tens of billions valuation, not even if everyone in the world started reading ten times more than they do.   But it doesn’t matter.  Amazon is staking out a part of the future.  They’re branching out into all kinds of online commerce, with varying degrees of success in different domains.  Music, toys, travel.  Whatever.  No one knows exactly how they’ll be getting their piece of the pie five years from now, but as the pie expands at its furtive pace, it’s a sure thing they’ll be at the forefront of the crowd struggling to slice it up and grab the slices of the new parts that don’t even exist yet.  


How could Autonomy, an obscure company with 100 employees making fairly run-of-the-mill text retrieval software, be worth $9 billion earlier this year (now down to $4 billion)?  No one really believes Autonomy can sell enough of their current product line to justify this valuation.  Part of the hype about Autonomy really is foolishness.   In a recent article glowing with enthusiasm, Wired magazine touted Autonomy’s statistics-based text analysis engine as if it were a major scientific breakthrough, when in fact.systems based on similar principles have been around for decades, and Autonomy’s engine underperforms old standards like INQUERY on standard tests.  But then, maybe this doesn’t matter.  With their oodles of cash, Autonomy could hire a team of scientists to replace their engine with something smarter.   The point is: Everyone knows information will rule the new economy, and whomever understands how to make and market tools for managing Internet information, is going to own a chunk of the golden prize, one way or another, even if right now we can’t see exactly how.   Autonomy understands something about how to manage information in the Internet world – their core technology may be standard stuff, but their marketing staff is out there every day figuring out how information retrieval technology can help Internet businesses grow, and help old-fashioned businesses become Internet businesses. 


Essentially, investors are giving companies like Autonomy and Amazon permission to use their money to explore uncharted regions of economic space.  Investors realize they can’t figure out what the right approach is going to be.  They trust the management of successful Net companies, who have staked out even a moderate amount of territory, to do their exploration for them.


People whose business it is to predict stock prices have traditionally approached the problem in two ways: fundamentally and technically.  Fundamental analysis is based on carefully studying a company: its business model, its management team, and so forth.   Technical analysis is based on tracking a company’s stock price and recognizing patterns in the ups and downs induced by speculation and collective psychology.   On a fairly crude level, the Internet stock mania can be understood in terms of these two categories.   It’s based on fundamentals in the sense that people are investing in companies they think are going to own the future.  These aren’t the ordinary fundamentals, because they’re future-oriented, not present-oriented; but they’re fundamentals nonetheless.  And the maxed-out Internet stock mania that rose to a fever pitch in early 2000, then dropped like a barrel going over Niagara falls, was also based on technical analysis in a very simple way: People saw the stocks going up, and assumed that this particular numerical pattern was going to continue.


Technical and fundamental thinking, and speculation, have been around a long time  What’s new and different in the Internet case is two things: the speculations are happening every second instead of every few hours, because of the advent of electronic trading; and the “fundamental” evaluations of companies involve future potentials rather than present business, with all the prediction error that this implies.  So the added volatility is no surprise.


The market right now is actually a relatively rational one.  Internet stocks are still valued pretty well.  They may be half what they were six months ago, but still, for companies with no profits and only slight revenues, even these valuations are pretty damn good.  Among investors, there seems to be more of a focus on fundamentals, on trying to assess whether a given firm really has a snowball’s chance in hell of becoming part of the net-centric future, as opposed to simply assuming that anything with .com in its name is guaranteed to be worth billions a few years down the line.


The worry, however, is that things may swing too far in the opposite direction.  Sure, things were a bit too crazy for a while, and firms with absolutely no fundamental value either now or in the long run were getting ridiculously high valuations.  But one hopes the market doesn’t overcorrect itself, and punish firms that really do have a decent chance of owning part of the future.  It’s fine if we continue to see a high rate of buzzword turnover – different “Internet company flavors of the month” flying by.  What we don’t want to see is the capital markets becoming overly short-term in their orientation. 


It seems reasonably likely that we’ll see a series of high-profile Internet bankruptcies toward the end of the year, and that this may trigger a backlash against even high-quality Internet stocks.   In reaction, companies will adjust both their behavior and their pitches.  Some behavioral change will be healthy and some will create missed opportunities as firms and investors forego long term investments.   But we have to hope that CEO’s and investors are too clever to go too far in this direction, and that they recognize that the companies that are dying are those that never had much of a chance in the first place.  The hope is that this particular dip in the road heading up toward the future will serve to make the market more intelligent, to make them evaluate Internet firms more carefully in terms of their real long-term potential.   The coming tech revolution is real.  It’s not easy to tell which current firms have a chance to be part of it, in a lasting way.  But it’s not impossible either.


Estimating Valuation, Internet-Style


So what’s the secret, then?  How can we tell which of the numerous Internet firms have a real chance of owning a chunk of the future?


It’s clear that current profits aren’t the secret.  Profits are important once an industry’s pecking order is known;  but prior to that, profits should be reinvested to create market share.   This used to be obscure economics lore, but now it’s the new common sense.  Now everyone takes for granted that, in emerging markets, one is not investing in a company based on its current profits, but rather based on what one thinks the company is going to do in a future that’s radically different from the present.  


But just because short-term profitability isn’t relevant, doesn’t mean that valuations should be totally pulled out of thin air.   There are systematic approaches.  For instance, Yobie Benjamin, the Director of Global Strategy at Ernst & Young, has proposed the formula


Net Future Expectations = Discounted Cash Flow + Net Future Opportunities


Net Future Opportunities = Innovation + Execution + Flexibility


This is reasonable enough.  All these factors are important.  People like new ideas, because the future of the Net is going to be new and exciting.  Obviously, execution is important, because it’s hard to take even a great idea and make it practical or popular.  Because no one knows what the future will bring, flexibility is key.  


Implicit in this formula is that it’s OK for firms in a new industry like the Net to have large fixed expenses for things like R&D.   This follows from the expectation that the size of the Internet market is rapidly increasing.  If expenses are fixed and revenue goes up by 5, due to the expansion of the market, then a solid company will turn a profit.


But Benjamin’s formula leaves some important things out.   The way I think about these issues is best captured by a modification to his equation, like this:


Net Future Expectations = Discounted Cash Flow + Net Future Opportunities


Net Future Opportunities = Business Qualities + Product Qualities


Business Qualities = Execution + Flexibility         


Product Qualities  =


+ Infrastructure Potential

                        +  Option Value of Current Customers

                        +  Intrinsic Switching Cost for Customers


Of course, you could expand a formula like this forever.   But this seems to capture the factors that have been crucial in the Internet market so far, and that are likely to play a big role in the immediate future. 


Infrastructure, first of all, is  a crucial idea that needs to be rethought again and again as the economy transforms itself.  Internet infrastructure is anything that’s a necessary condition for survival on the Net.  Any firm that provides a good or service that every Net company is going to take for granted two or three years down the line, if they don’t already, is a reasonably good bet to own a chunk of the world after this phase of the revolution is over.  Being infrastructure is one brilliant way to get sticky customers.


But not everyone can be infrastructure.  If you can’t force customers to use you by being infrastructure, you have to resort to other means to make your customers stick.  Customer loyalty is always important, but it’s even more important when one’s trying to guess whether a firm is going to be immensely valuable a few years from now.  This is the “land grab” idea underlying a lot of early Internet marketing.  Grab up the customers now while the grabbing’s good – each customer now has a humongous “option value”, far beyond the immediate revenue one may get from them.  One is grabbing the customer for the indefinite future, and in the future Internet-mediated transactions may consume all their income.  Of course, the presupposition here is that the land is really being grabbed in a lasting way – that the customers obtained now are going to stick.


The question of how non-infrastructure firms can attract sticky customers leads one into the wild, weird world of Internet marketing, where customers are actively engaged partners rather than merely targets of advertisements and purchasers of goods and services.   In order to make a customer stick, one wants to make the cost of switching high, in one way or another.  If one can’t make the cost of switching high in an objective sense, one can make it high psychologically, by entangling one’s product in the customer’s life.   One way to do this is to engage the customer in the process of creating a good or service in the first place, which companies like GeoCities have excelled at.  Another way is to make your product a habitual part of the customer’s daily routine, which is the rationale behind many personal service sites such as PayMyBills and Kozmo.  Finally, one can, as in the case of E-Bay, allow customers to participate in the valuation of the goods they are purchasing and feel as if they’re equal partners in determining how much they pay, making them want you to stick around as the other partner who is enabling this.  In all these says, instead of of being Internet infrastructure, a firm becomes part of their customers’ “personal infrastructure.”



Varieties of Infrastructure


We don’t know what the Net’s going to be like in 2 years, let alone 5 or 10.  But we can tell some basic things about it.  We know there are going to be network cables and routers.  We know there are going to be computers and operating systems and communication protocols.  These things that have to be there, that every Net company is going to use, one way or another – these are Internet infrastructure.  Firms that provide necessary infrastructure are likely to be in it for the long haul.  This is a small subset of the total pool of Internet firms, but it’s an important one.   After all, there’s no stickier customer than the one who has to use your product because it’s a part of the basic infrastructure underlying all their transactions.


How many kinds of infrastructure are there?  The first categorization to keep in mind is hardware versus software.  Hardware infrastructure in the Internet worlds means two things: computers and networks.  Whatever particular direction the Net economy takes, we’re going to have more and more computers and more and more networks; that’s a no-brainer.    This underlies the success of Intel and its new rival chipmakers, and also Cisco Systems’ lofty valuation.  Cisco owns a big chunk of the network hardware market, a market that’s been expanding steadily, even as the nature of the companies using the network hardware has shifted from one subsector to another.  Business-to-consumer, business-to-business, whatever – they all need routers. 


Software infrastructure is equally crucial.   Microsoft proves that.  Their hold on the OS market allowed them to capture the software application market.  Their shady business tactics, as revealed in the recent antitrust suit, are most remarkable due to their apparent unnecessariness.  Microsoft was extremely intelligent in their business strategy: they realized that whomever owns the infrastructure has not only a hell of a good business in infrastructure, since everyone is their customer, but also a shot at owning the applications built on top of the infrastructure, due to their ubiquitous brand name and their understanding of the world in which applications operate.  They surely would have succeeded even without the various dishonest practices that have been uncovered.


Operating systems are not the only kind of software infrastructure, though.  Network software is at least as important.  Cisco owns a lot of this as well, right now, on the lower level: software for routing messages.   The Java language, with its powerful and easy-to-use networking libraries, is a crucial part of Internet software infrastructure, which is why Microsoft eventually stopped fighting Java and started working with it (Microsoft’s Java Virtual Machines are now the best in the world).   Akamai’s clever system for rapidly serving up high-volume Websites anywhere in the world is a beautiful software infrastructure innovation. 


Inktomi is a good example here.  They started out as a search company, but their real leap in value occurred when the market realized their technology for scaling their search application was more valuable than their search technology itself.  It was an infrastructure product, a product anyone could use to allow their software to scale to deal with huge volumes of Internet data.


These are standard types of infrastructure, already widely acknowledged, and companies providing this kind of infrastructure are generally already generously valued.  The insight that being infrastructure provides sticky customers is already prices into their valuations, to a great extent.  But that doesn’t mean the notion of infrastructure isn’t useful.   Given the novelty of what’s happening all around us, it’s almost a sure thing that there are new kinds of infrastructure that haven’t even been thought of yet. 


There are new kinds of software infrastructure that are just beginning to emerge.   One thing we’ll see over the next five years is massively distributed processing infrastructure.   There are a lot of PC’s in the world, and most of them usually aren’t doing anything useful.  This is a huge unexploited computational resource which before long is sure to be exploited on a commercial basis.   Projects like SETI@home and distributed.net have gotten their first, providing free-download screensaver programs that do computations – each computer in the distributed network doing a small part of the total computation, and a central server collating the results.  These projects are carrying out computations with a humanistic focus: analyzing data for indications of extraterrestrial life, or factoring large numbers searching for primes.  But the same approach could be used for all sorts of other computations, including data mining and other things of practical value.  There’s an interesting tie-in here with recent distributed file sharing systems like Gnutella, which allows individual computers to share MP3 files in a distributed “peer-to-peer” way, without any central server.   Exactly what form the distributed processing infrastructure will take 2, 5 or 10 years from now is far from certain.  Perhaps home computer users will get paid for leaving their computers on at night, so their CPU’s can be used by companies that rent out millions of CPU cycles to other firms needing to run complex data mining jobs.  Perhaps a smart version of Windows will someday analyze the data on your hard drive to improve its personalized performance, talking to other nearby computers to get new insights on how to analyze your data best.   There are zillions of options.  One way or another, distributed processing will be another ingredient of the software infrastructure of the future.  



Intelligence Infrastructure


One new market sector that is likely to play a big role starting in a year or two, increasing perhaps to a dominant role in half a decade, is intelligence infrastructure.  Of course I have to admit a bias here: my own company, Intelligenesis, aims to provide intelligence infrastructure for the new economy.  I’ve devoted the last three years of my life to the proposition that intelligence infrastructure is going to transform the Net as we know it.


Every firm living on the Internet is going to require artificial intelligence in one way or another.  And the more intelligent your competitors are, the more intelligent you’ll have to be.   Media sites require intelligent categorization, intelligent search, intelligent information filtering, intelligent personalization.  Advertisers require intelligent targeting of advertisements.  As the online economy becomes a hypereconomy with intelligent agents bartering and auctioning for each micro-transaction, software intelligence becomes necessary for getting the right price.  Financial transactions require intelligent prediction based on a full range of numerical and textual information. 


Right now intelligence is provided to Internet software applications in a haphazard and slipshod way, embedded in various applications or supplied by specialized AI widgets that carry out very narrow, limited functions.  In the future intelligence will be a service that one turns on like electricity or the telephone – or Internet connectivity, or a computer operating system.  There will be various intelligence services that can be requested from an intelligence server, to solve the AI needs of any product or service.  The Webmind system that we’re developing at Intelligenesis is intended to serve as the backbone of an intelligence service like this.


What other kinds of infrastructure will emerge as the Net economy develops?  Answer this question, and you’ll have an important clue as to which current firms are going to have sticky customers.  Firms with potential to provide new kinds of Internet infrastructure are better than average bets to survive to own a chunk of the future.   



Sticky Customers


Infrastructure firms need to put their efforts toward being sure they’re really adopted as infrastructure, by the rest of the industry.  But not everyone can be infrastructure.  Non-infrastructure firms need to focus on getting customers – and not just any customers, but sticky customers.  Customers that are going to stick with a firm through the transition from the current economy to the net-centric economy of the future.  An initial core of sticky customers will grow and grow as the Net economy grows up. 


The curse of the typical successful business-to-consumer e-commerce firm is to have many customers but few sticky ones.  In other words, zero switching cost, economically and psychologically.  Customers obtained through high-priced advertising campaigns, customers easily lured away by others’ similar campaigns.  Valuations for consumer retail dot-coms plunged before anything else did early this year, partly because investors realized that these firms were spending huge advertising dollars just to obtain and retain customers, without any real strategy for working their way into customers’ hearts in a lasting way. 


Business-to-business e-commerce has more potential to get sticky customers, because businesses have much higher switching costs than consumers.  They need to find companies that deliver on time with high quality and good service and adapt to their changing needs.   Each time they switch they have to do a lot more research than a consumer does, because the cost of a bad switch is much higher for them.  But the problem with this market is, there’s not room for a heck of a lot of different players.  A given niche of the B2B market, say automobile parts, is not likely to support a whole bunch of different intermediaries.  It seems probable that each niche is going to be dominated by one or at most two e-commerce firms.   At this stage there’s still a lot of competition for each niche and it’s not so easy to pick the winner.


How can an Internet firm make customers stick with them over the long haul?   The traditional answer is brand loyalty.  People need to be proud of the brand they use, owe a debt to a brand, think it enhances their image as cool, intelligent, rich, or whatever.  But the new generation is notoriously fickle, and the lightning speed of the Net economy encourages this in all participants regardless of age.  In the Net world, brand loyalty doesn’t come so easily through the tools of traditional marketing.  Instead the notions of community and process loom large.   A company  must involve the user in an emotional relationship, either with other customers in a way mediated by its business, or with its business directly.  


This is why, in the Internet domain, giving products and services away free is such a valuable strategy.   If one can give something away, but create an emotional relationship with the customer in the process, one may be gaining a great deal.  As the Book of Proverbs says, “One man gives freely yet gains even more.   Another withholds unduly, but comes to poverty.” 


One can give away early versions of products as part of a land grab methodology -- establishing an early install base, getting customers familiar with one’s products and creating a practical and psychological switching cost, so customers will want to pay for fully-featured product versions later on.  This is a very common practice in the Internet industry now – betas are free, 1.0 versions cost money.  Or, one can give away simple products and charge for complex ones, the simple ones creating a market for the complex ones.  Adobe, with its free Acrobat Reader, has established its PDF format for high-quality printable Internet documents as a standard.   One can give away consumer products and charge for corporate products -- it worked for Netscape, at least to some extent; the firm didn’t realize all its dreams, but certainly did better than the run-of-the-mill start-up.  Or one can give stuff away as a matter of principle and make money off services and add-ons: this is the Linux approach, which has gotten Linux a long way so far, and the story of Linux is just beginning.  There are dozens more examples – the Net is no longer dominated by free stuff, like it was in the 70’s and 80’s, but giving away goods and services to gain relationships has proved itself a lastingly valuable strategy.


The different aspects of fundamental value in the Internet space interconnect in various subtle ways.  For example, the practice of giving away software for free ties in with one of Benjamin’s core business values, flexibility.   Consider the Linux community, a temporary, self-managed community of individuals working toward a common task.  The not-for-profit nature of the enterprise naturally leads to a tremendous flexibility, in which different players participate in the system as their time permits, coming in and out of the dis-organization freely.  This could be considered a model for a new kind of business organization and a new kind of economy.  But of course, flexibility taken to this extreme has its own problems associated with it.  The Linux community is immensely creative on a technical level, but is weak at sensing customer needs: the engineers building Linux are themselves customers, but atypical customers, and are hence more interested in serving their own needs than the needs of a broader customer group.



Net-savvy  Marketing


Even if a firm is lucky enough to provide a service that everyone needs, they still need to show people that they need it.  And chances are, it’s not possible to provide the service in exactly the right way without participation from the customers themselves in defining the service.  The heart of Internet marketing is the creation of sticky customers through the involvement of customers in processes.  In the new economy, you can’t think of customers as purchasers of products and recipients of advertisements.  You have to think of them as partners, you have to weave your product into their lives, and weave them into the process of creating your products.


The basic ideas of marketing are no different in the Internet world from anywhere else.   Marketing still comes down to presenting a story, telling it well, endowing it with flash, sizzle and glamour, and getting the message out there with a marketing campaign that will get people to listen to your story in the first place.   


But there are some really significant differences.  In the Internet world, companies cannot make themselves valuable just through spending money on advertising.  The dismal failure of so many heavily advertised dot coms this year demonstrates this viscerally.   There’s a whole different psychology required.


Marketing’s traditional connections to customers are no longer sufficient in a real-time world.  Focus groups, market research, consumer surveys, and other traditional tools for probing the consumer’s wants and needs have always been limited, but their limitations are far more severe in an Internet context.  By the time a systematic market survey has been done, the market may have changed completely!  Furthermore, many of the customer segments one is concerned with don’t even exist, so it’s impossible to survey them – the problem is to create them.   Internet marketers have found that more continuous connections with customers can provide information that focus groups and surveys cannot.  Thus we have companies that observe a customer’s behavior on the Web, and allow Websites to modify their response to that customer in real-time.  The beginning of this is dynamically targeted advertising, but this is only the beginning.


In the ultimate extension, Internet  marketing isn’t just about marketing, it’s about the whole product lifecycle -- about involving the customer as a partner in development and production.   This will be a tough transition for traditional companies to make, because most firms today have focused their processes on improving time to market and, by inclination and culture, see the customer as an end target rather than a partner.   Instead of time to market, what firms need to focus on in the Internet economy is “time to acceptance.”   The customer needs to be invited to participate in creating the product, in which case acceptance arises even before product delivery.   This creates sticky customers almost inevitably, because the customer feels themselves as part of the product.  They’re sticking to themselves!  AOL has been masterful at this, at gradually modifying their product, which initially was quite poor, to meet customer needs.   They listened to their customers, not via annual surveys, but via daily e-mail and chat feedback.  They responded on Internet time. 


Who knows how far this will go?  Targeted marketing may turn upside-down.   We take for granted now that companies can gather information about customers and use or sell it as they wish.  But  soon enough, customers may take ownership of their own preference and transaction information, and use this to their own advantage.   An individual computer user can now get much more comprehensive and accurate profiles of their own commercial activities than any individual vendor can.  This is a source of value which customers cannot currently exploit, but new firms may arise which allow customers to exploit it.    “Infomediary” firms will seize the opportunity to act as custodians, agents and brokers of customer information, marketing it to businesses on consumers behalf while protecting their privacy at the same time.   This is another kind of Internet infrastructure, as new and exciting as intelligence infrastructure.  Profiling infrastructure takes customer information and gives it to the customer, and gives the customer the ability to sell it to vendors as he wishes.  It deserves the name infrastructure because it’s a service that virtually every product or service on the Net is going to want to make use of.  It doesn’t exist yet.  In five years it will.  The firms that get there first will probably own a hefty chunk of this particular portion of the future.


Another way to look at this is: To get sticky customers, you have to stake a claim to a particularly important part of your customer’s mind    Firms can make themselves valuable through staking out turf in the collective consciousness, through grabbing onto the mass psychological intuition that is at the root of the Internet stock phenomenon, and evolving this intuition together with their customers.  They can make themselves valuable through marketing that engages their customers with them in the process of creating the new net-centric world – that makes them part of the psychic infrastructure of 21’st century culture as it lurches toward its next phase of being.



AOL has done a great job of this – not only, as mentioned above, by responding to their customers continually and flexibly; but by providing a service that defines a substantial part of their clients’ lives.  AOL is an Internet service provider, yes; it’s a news and entertainment source; but to a large extent, today, it’s chat.   AOL’s proprietary chat rooms are a whole network of subcultures.  In a fundamental sense these subcultures aren’t owned by anyone – they’re part of the collective culture – but in practice, to participate you have to log onto America Online.  Through listening to its customers, AOL has worked itself into a position where it’s an integral part of many of its customers’ lives.


Another illustration is the current round of ads for the online trading firms like E-trade and Datek and Ameritrade.  These ads offer you freedom, excitement and wonder.  Now, thanks to the splendor of the Net, you can do what was previously the province of the Masters of the Universe only, the Wall Street elite!   For those who like to invest, there is a real service being offered here – trades at one tenth the transaction cost that brokers charged a few years ago.  But what is being sold is not just cheap trades – what’s being sold is a vision, and not just a vision of stock trading, a vision of the wondrous future of the Net, where untold powers and privileges are yours for the plucking.  And the more you trade on E-trade, the more you get locked into the E-trade community, the internal E-trade messageboards and so forth.  You become a sticky customer, a stuck customer.  Quite possibly you will stick with e-trade into the Fabulous Net-centric Future….   Because not only does E-trade have a good business model and sticky customers – they got these sticky customers because they get it, marketing-wise.  They know how to get customers in the new economy, by interacting with them, fundamentally relating with them, sucking them into a world; not just studying them dispassionately and flooding them with advertisements. 


E-trade isn’t infrastructure.  They’re not providing something every software product or service on the Net has to use, or will ever have to use.  They can’t get sticky customers through simple ubiquity, so they have to do it through culture and process, through responding flexibly to customers’ needs and providing services that burrow deeply into customers’ lives.  They seem to be succeeding just fine.



Riding the Wave of Transformation


So what’s the story with Internet stocks?   Is there hype here?   Is there speculation?   Is there craziness?   Of course there is!   Life wouldn’t be any fun without it.   But unlike the tulip bulb craze, which was pure speculation, or the biotech bubble, which was based on future expectations uncorrelated with short-term developments, the Internet stock craze has truly sound fundamentals underlying all the wackiness.   The Net really is transforming the way we create wealth, and more generally the way we live.   Just log on, you can live it every day.


In spite of the chaos of fluctuating Internet company valuations, there are definite factors that make some Internet companies more valuable than others, in the long term.  Sticky customers and infrastructure are important, as are bread-and-butter new economy values like innovation, flexibility and execution.  These fundamental values are sometimes hard to pick out among all the hype, but they’re there.   Many of the firms that grab sticky customers or define Internet infrastructure today will have great wealth tomorrow.  Firms that understand the new realities of marketing on a gut level will have a much better chance of success than those that don’t. 


Of course, these factors don’t tell you everything.   It is difficult to predict, especially the future.  But the uncertainty of any particular firm’s future shouldn’t prevent us from seeing what is almost certain: the Net isn’t going away.   It’s here to stay, and it’s going to pervade everything, and whomever contributes significant value to it, is going to be rewarded more and more as this transformation happens.