What the Bubble Got Right     September 2004(This essay is derived from an invited talk at ICFP 2004.)I had a front row seat for the Internet Bubble,because I worked at Yahoo during 1998 and 1999. One day,when the stock was trading around $200, I sat down and calculatedwhat I thought the price should be. The answer I got was $12. I went tothe next cubicle and told my friend Trevor. "Twelve!" he said.He tried to sound indignant, but he didn't quite manage it. Heknew as well as I did that our valuation was crazy.Yahoo was a special case. It was not just our price to earningsratio that was bogus. Half our earnings were too. Not inthe Enron way, of course. The finance guys seemedscrupulous about reporting earnings. What made ourearnings bogus was that Yahoo was, in effect, the center ofa pyramid scheme. Investors looked at Yahoo's earningsand said to themselves, here is proof that Internet companies can makemoney. So they invested in newstartups that promised to be the next Yahoo. And as soon as these startupsgot the money, what did they do with it?Buy millions of dollars worth of advertising on Yahoo to promotetheir brand. Result: a capital investment in a startup thisquarter shows up as Yahoo earnings next quarter—stimulatinganother round of investments in startups.As in a pyramid scheme, what seemed to be the returns of this systemwere simply the latest round of investments in it.What made it not a pyramid scheme was that it was unintentional. At least, I think it was. The venture capital business is pretty incestuous,and there were presumably people in a position, if not to createthis situation, to realize what was happening and to milk it.A year later the game was up. Starting in January 2000, Yahoo'sstock price began to crash, ultimately losing 95% of itsvalue.Notice, though, that even with all the fat trimmed off its marketcap, Yahoo was still worth a lot. Even at the morning-aftervaluations of March and April 2001, the people at Yahoo had managedto create a company worth about $8 billion in just six years.The fact is, despite all the nonsense we heardduring the Bubble about the "new economy," there was acore of truth. You needthat to get a really big bubble: you need to have somethingsolid at the center, so that even smart people are sucked in.(Isaac Newton and Jonathan Swift both lost moneyin the South Sea Bubble of 1720.)Now the pendulum has swung the other way. Now anything thatbecame fashionable during the Bubble is ipso facto unfashionable.But that's a mistake—an even bigger mistake than believingwhat everyone was saying in 1999. Over the long term,what the Bubble got right will be more important than whatit got wrong.1. Retail VCAfter the excesses of the Bubble, it's nowconsidered dubious to take companies public before they have earnings.But there is nothing intrinsically wrong withthat idea. Taking a company public at an early stage is simplyretail VC: instead of going to venture capital firms for the last round offunding, you go to the public markets.By the end of the Bubble, companies going public with noearnings were being derided as "concept stocks," as if itwere inherently stupid to invest in them.But investing in concepts isn't stupid; it's what VCs do,and the best of them are far from stupid.The stock of a company that doesn't yet have earnings is worth something.It may take a while for the market to learnhow to value such companies, just as it had to learn tovalue common stocks in the early 20th century. But marketsare good at solving that kind of problem. I wouldn't besurprised if the market ultimately did a betterjob than VCs do now.Going public early will not be the right planfor every company.And it can of course bedisruptive—by distracting the management, or by making the earlyemployees suddenly rich. But just as the market will learnhow to value startups, startups will learn how to minimizethe damage of going public.2. The InternetThe Internet genuinely is a big deal. That was one reasoneven smart people were fooled by the Bubble. Obviously it was going to have a huge effect. Enough of an effect totriple the value of Nasdaq companies in two years? No, as itturned out. But it was hard to say for certain at the time. [1]The same thing happened during the Mississippi and South Sea Bubbles.What drove them was the invention of organized public finance(the South Sea Company, despite its name, was really a competitorof the Bank of England). And that did turn out to bea big deal, in the long run.Recognizing an important trend turns out to be easier than figuring out how to profit from it. The mistakeinvestors always seem to make is to take the trend too literally.Since the Internet was the big new thing, investors supposedthat the more Internettish the company, the better. Hencesuch parodies as Pets.Com.In fact most of the money to be made from big trends is madeindirectly. It was not the railroads themselves that made the most money during the railroad boom, but the companieson either side, like Carnegie's steelworks, which made the rails,and Standard Oil, which used railroads to get oil to the East Coast,where it could be shipped to Europe.I think the Internet will have great effects,and that what we've seen so far is nothing compared to what'scoming. But most of the winners will only indirectly beInternet companies; for every Google there will be tenJetBlues.3. ChoicesWhy will the Internet have great effects? The general argument is that new forms of communication always do. They happenrarely (till industrial times there were just speech, writing, and printing),but when they do, they always cause a big splash.The specific argument, or one of them, is the Internet gives us more choices. In the "old" economy,the high cost of presenting information to people meant theyhad only a narrow range of options to choose from. The tiny,expensive pipeline to consumers was tellingly named "the channel."Control the channel and youcould feed them what you wanted, on your terms. And itwas not just big corporations that dependedon this principle. So, in their way, didlabor unions, the traditional news media,and the art and literary establishments.Winning depended not on doing good work, but on gaining controlof some bottleneck.There are signs that this is changing.Google has over 82 million unique users a month andannual revenues of about three billion dollars. [2]And yet have you ever seena Google ad?Something is going on here.Admittedly, Google is an extreme case. It's very easy forpeople to switch to a new search engine. It costs littleeffort and no money to try a new one, and it's easy tosee if the results are better. And so Google doesn't haveto advertise. In a business like theirs, being the best isenough.The exciting thing about the Internet is that it'sshifting everything in that direction.The hard part, if you want to win by making the best stuff,is the beginning. Eventually everyonewill learn by word of mouth that you're the best,but how do you survive to that point? And it is in this crucialstage that the Internet has the most effect. First, theInternet lets anyone find you at almost zero cost.Second, it dramatically speeds up the rate at whichreputation spreads by word of mouth. Together these mean that in manyfields the rule will be: Build it, and they will come.Make something great and put it online.That is a big change from the recipe for winning in thepast century.4. YouthThe aspect of the Internet Bubble that the press seemed mosttaken with was the youth of some of the startup founders.This too is a trend that will last.There is a huge standard deviation among 26 year olds. Someare fit only for entry level jobs, but others areready to rule the world if they can find someone to handlethe paperwork for them.A 26 year old may not be very good at managing people ordealing with the SEC. Those require experience.But those are also commodities, which can be handed off tosome lieutenant. The most important quality in a CEO is hisvision for the company's future. What will they build next?And in that department, there are 26 year olds who cancompete with anyone.In 1970 a company president meant someone in his fifties, atleast. If he had technologists working for him, they were treated like a racing stable: prized, but not powerful. But as technology has grown more important, the power of nerdshas grown to reflect it. Now it's not enough for a CEO tohave someone smart he can ask about technical matters. Increasingly,he has to be that person himself.As always, business has clung to old forms. VCs still seemto want to install a legitimate-looking talking head as the CEO. But increasingly the founders ofthe company are the real powers, and the grey-headed maninstalled by the VCs more like amusic group's manager than a general.5. InformalityIn New York, the Bubble had dramatic consequences:suits went out of fashion. They made one seem old. So in1998 powerful New York types were suddenly wearingopen-necked shirts and khakis and oval wire-rimmed glasses,just like guys in Santa Clara.The pendulum has swung back a bit, driven in part by a panickedreaction by the clothing industry. But I'm betting on theopen-necked shirts. And this is not as frivolous a questionas it might seem. Clothes are important, as all nerds can sense,though they may not realize it consciously.If you're a nerd, you can understand how important clothes areby asking yourself how you'd feel about a companythat made you wear a suit and tie to work. The idea soundshorrible, doesn't it? In fact, horrible far out of proportionto the mere discomfort of wearing such clothes. A company thatmade programmers wear suits would have something deeply wrongwith it.And what would be wrong would be that how one presented oneselfcounted more than the quality of one's ideas. That'sthe problem with formality. Dressing up is not so much bad initself. The problem is the receptor it binds to: dressingup is inevitably a substitutefor good ideas. It is no coincidence that technicallyinept business types are known as "suits."Nerds don't just happen to dress informally. They do it tooconsistently. Consciously or not, they dress informally asa prophylactic measure against stupidity.6. NerdsClothing is only the most visible battleground in the waragainst formality. Nerds tend to eschew formality of any sort.They're not impressed by one's job title, for example,or any of the other appurtenances of authority.Indeed, that's practically the definition of a nerd. I foundmyself talking recently to someone from Hollywood who was planninga show about nerds. I thought it would be useful if Iexplained what a nerd was. What I came up with was: someone whodoesn't expend any effort on marketing himself.A nerd, in other words, is someone who concentrates on substance.So what's the connection between nerds and technology? Roughlythat you can't fool mother nature. In technical matters, youhave to get the right answers. If your software miscalculatesthe path of a space probe, you can't finesse your way out oftrouble by saying that your code is patriotic, or avant-garde,or any of the other dodges people use in nontechnicalfields.And as technology becomes increasingly important in theeconomy, nerd culture is rising with it. Nerds are alreadya lot cooler than they were when I was a kid. When I was incollege in the mid-1980s, "nerd" was still an insult. Peoplewho majored in computer science generally tried to conceal it.Now women ask me where they can meet nerds. (The answer thatsprings to mind is "Usenix," but that would be like drinkingfrom a firehose.)I have no illusions about why nerd culture is becomingmore accepted. It's not because people arerealizing that substance is more important than marketing.It's because the nerds are getting rich. But that is not goingto change.7. OptionsWhat makes the nerds rich, usually, is stock options. Now thereare moves afoot to make it harder for companies to grant options. To the extent there's some genuine accounting abuse going on, by all means correct it. But don't kill the golden goose. Equity is the fuel that drives technical innovation.Options are a good idea because (a) they're fair, and (b) theywork. Someone who goes to work for a company is (one hopes) adding to its value, and it's only fair to give them a shareof it. And as a purely practical measure, people work a lotharder when they have options. I've seen that first hand.The fact that a few crooks during the Bubble robbed theircompanies by granting themselves options doesn't mean optionsare a bad idea. During the railroad boom, some executivesenriched themselves by selling watered stock—by issuing moreshares than they said were outstanding. But that doesn't make common stock a bad idea. Crooks just use whatevermeans are available.If there is a problem with options, it's that they rewardslightly the wrong thing. Not surprisingly, people do what youpay them to. If you pay them by the hour, they'll work a lot ofhours. If you pay them by the volume of work done, they'llget a lot of work done (but only as you defined work).And if you pay them to raise thestock price, which is what options amount to, they'll raisethe stock price.But that's not quite what you want. What you want is toincrease the actual value of the company, not its market cap.Over time the two inevitably meet, but not always as quicklyas options vest. Which means options tempt employees, ifonly unconsciously, to "pump and dump"—to do thingsthat will make the company seem valuable.I found that when I was at Yahoo, I couldn't help thinking, "how will this sound to investors?" when I should have beenthinking "is this a good idea?"So maybe the standard option deal needs to be tweaked slightly.Maybe options should be replaced with something tied moredirectly to earnings. It's still early days.8. StartupsWhat made the options valuable, for the most part, isthat they were options on the stock of startups. Startups were not of course a creation of the Bubble, but theywere more visible during the Bubble than ever before.One thing most people did learn about for the first timeduring the Bubble was the startupcreated with the intention of selling it.Originally astartup meant a small company that hoped to grow into abig one. But increasingly startups are evolving into avehicle for developing technology on spec.As I wrote inHackers & Painters, employees seem to be mostproductive when they're paid in proportion to the wealththey generate. And the advantage of a startup—indeed, almost its raison d'etre—is that it offers somethingotherwise impossible to obtain: a way of measuring that.In many businesses, it just makes more sense for companiesto get technology by buying startups rather than developing it in house. You pay more, but there is less risk,and risk is what big companies don't want. It makes theguys developing the technology more accountable, because theyonly get paid if they build the winner. And you end up with better technology, created faster, because things aremade in the innovative atmosphere of startups instead of the bureaucratic atmosphere of big companies.Our startup, Viaweb, was built to be sold. We were openwith investors about that from the start. And we were careful to create something that could slot easily into alarger company. That is the pattern for the future.9. CaliforniaThe Bubble was a California phenomenon. When I showed upin Silicon Valley in 1998, I felt like an immigrant fromEastern Europe arriving in America in 1900. Everyonewas so cheerful and healthy and rich. It seemed a newand improved world.The press, ever eager to exaggerate small trends, now gives one the impression that Silicon Valley is a ghost town.Not at all. When I drive down 101 from the airport,I still feel a buzz of energy, as if there were a gianttransformer nearby. Real estate is still more expensivethan just about anywhere else in the country. The people still look healthy, and the weather is still fabulous.The future is there.(I say "there" because I moved back to the East Coast afterYahoo. I still wonder if this was a smart idea.)What makes the Bay Area superior is the attitude of thepeople. I notice that when I come home to Boston.The first thing I see when I walk out of the airline terminalis the fat, grumpy guy incharge of the taxi line. I brace myself for rudeness:remember, you're back on the East Coast now.The atmosphere varies from city to city, and fragileorganisms like startups are exceedingly sensitive to such variation.If it hadn't already been hijacked as a new euphemismfor liberal, the word to describe the atmosphere inthe Bay Area would be "progressive." People there are tryingto build the future.Boston has MIT and Harvard, but it also has a lot oftruculent, unionized employees like the police whorecently held the Democratic National Convention for ransom, and a lot of people trying to be Thurston Howell.Two sides of an obsolete coin.Silicon Valley may not be the next Paris or London, but itis at least the next Chicago. For the next fifty years, that's where new wealth will come from.10. ProductivityDuring the Bubble, optimistic analysts used to justify highprice to earnings ratios by saying that technology was going to increase productivity dramatically. They were wrong aboutthe specific companies, but not so wrong about the underlyingprinciple. I think one of the big trends we'll see in thecoming century is a huge increase in productivity.Or more precisely, a huge increase in variation inproductivity. Technology is a lever. It doesn't add; it multiplies. If the present range of productivity is 0 to 100, introducing a multiple of 10 increases the rangefrom 0 to 1000.One upshot of which is that the companies of the future maybe surprisingly small. I sometimes daydream about how bigyou could grow a company (in revenues) without ever havingmore than ten people. What would happen if you outsourcedeverything except product development? If you tried this experiment,I think you'd be surprised at how far you could get. As Fred Brooks pointed out, small groups areintrinsically more productive, because theinternal friction in a group grows as thesquare of its size.Till quite recently, running a major companymeant managing an army of workers. Our standards about howmany employees a company should have are still influenced byold patterns. Startups are perforce small, because they can'tafford to hire a lot of people. But I think it's a big mistake forcompanies to loosen their belts as revenues increase. Thequestion is not whether you can afford the extra salaries. Can you afford the loss in productivity that comes from makingthe company bigger?The prospect of technological leverage will of course raise thespecter of unemployment. I'm surprised people still worry aboutthis.After centuries of supposedly job-killing innovations,the number of jobs is within ten percent of the number of peoplewho want them. This can't be a coincidence. There must be somekind of balancing mechanism.What's NewWhen one looks over these trends, is there any overall theme?There does seem to be: that in the coming century, good ideaswill count for more. That 26year olds with good ideas will increasingly have an edge over 50year olds with powerful connections. That doing good work willmatter more than dressing up—or advertising, which is thesame thing for companies. That peoplewill be rewarded a bit more in proportion to the value of whatthey create.If so, this is good news indeed.Good ideas always tend to win eventually. The problem is,it can take a very long time.It took decades for relativity to be accepted, and thegreater part of a century to establish that central planning didn't work.So even a small increase in therate at which good ideas win would be a momentouschange—big enough, probably, to justify a name likethe "new economy."Notes[1] Actually it's hard to say now. As Jeremy Siegel pointsout, if the value of a stock is its future earnings, you can't tell if it was overvalued till you see what the earningsturn out to be. While certain famous Internet stocks werealmost certainly overvalued in 1999, it is still hard to say for surewhether, e.g., the Nasdaq index was.Siegel, Jeremy J. "What Is an Asset Price Bubble? AnOperational Definition." European Financial Management,9:1, 2003.[2] The number of users comes from a 6/03 Nielsenstudy quoted on Google's site. (You'd think they'd havesomething more recent.) The revenue estimate is based onrevenues of $1.35 billion for the first half of 2004, asreported in their IPO filing.Thanks to Chris Anderson, Trevor Blackwell, Sarah Harlin, JessicaLivingston, and Robert Morris for reading drafts of this. The Long TailRussian TranslationJapanese Translationcsell_env = 'mud';// Begin Y! Store Generated Code // Begin Y! Store Generated Code csell_page_data = {}; csell_page_rec_data = []; ts='TOK_STORE_ID';// Begin Y! Store Generated Codefunction csell_GLOBAL_INIT_TAG() { var csell_token_map = {}; csell_token_map['TOK_ITEM_ID_LIST'] = 'bubble'; csell_token_map['TOK_BEACON_TYPE'] = 'prod'; csell_token_map['TOK_RAND_KEY'] = 't'; csell_token_map['TOK_SPACEID'] = '2022276099'; csell_token_map['TOK_IS_ORDERABLE'] = '2'; csell_token_map['TOK_STORE_ID'] = 'paulgraham'; csell_token_map['TOK_URL'] = 'http://geo.yahoo.com'; csell_token_map['TOK_ORDER_HOST'] = 'order.store.yahoo.net'; c = csell_page_data; var t = csell_token_map; c['s'] = t['TOK_SPACEID']; c['url'] = t['TOK_URL']; c['si'] = t[ts]; c['ii'] = t['TOK_ITEM_ID_LIST']; c['bt'] = t['TOK_BEACON_TYPE']; c['rnd'] = t['TOK_RAND_KEY']; c['io'] = t['TOK_IS_ORDERABLE']; YStore.addItemUrl = 'http%s://'+t['TOK_ORDER_HOST']+'/ymix/MetaController.html?eventName.addEvent&cartDS.shoppingcart_ROW0_m_orderItemVector_ROW0_m_itemId=%s&cartDS.shoppingcart_ROW0_m_orderItemVector_ROW0_m_quantity=1&ysco_key_cs_item=1§ionId=ysco.cart&ysco_key_store_id='+t[ts]; } // Begin Y! Store Generated Codefunction csell_REC_VIEW_TAG() { var env = (typeof csell_env == 'string')?csell_env:'prod'; var p = csell_page_data; var a = '/sid='+p['si']+'/io='+p['io']+'/ii='+p['ii']+'/bt='+p['bt']+'-view'+'/en='+env; var r=Math.random(); YStore.CrossSellBeacon.renderBeaconWithRecData(p['url']+'/p/s='+p['s']+'/'+p['rnd']+'='+r+a); } // Begin Y! Store Generated Codevar csell_token_map = {}; csell_token_map['TOK_PAGE'] = 'p'; csell_token_map['TOK_WS_URL'] = ''; csell_token_map['TOK_SHOW_CS_RECS'] = 'false'; csell_token_map['TOK_CURR_SYM'] = '$'; var t = csell_token_map; csell_GLOBAL_INIT_TAG(); YStore.page = t['TOK_PAGE']; YStore.currencySymbol = t['TOK_CURR_SYM']; YStore.crossSellUrl = t['TOK_WS_URL']; YStore.showCSRecs = t['TOK_SHOW_CS_RECS']; |
|
