strategy and economics of information

COMPETING IN THE INFORMATION ECONOMY

SEPTEMBER-OCTOBER 1997

STRATEGY AND THENEW ECONOMICSOF INFORMATION

by Philip B. Evans and Thomas S. Wurster

A fundamental shift in the economics of information is underway-a shift that is less about any specific new technology thanabout the fact that a new behavior is reaching critical mass. Mil-lions of people at home and at work are communicating elec-tronically using universal, open standards. This explosion in con-nectivity is the latest-and, for business strategists, the mostimportant -wave in the information revolution.

Over the past decade, managers have focused on adapting theiroperating processes to new information technologies. Dramatic asthose operating changes have been, a more profound transforma-tion of the business landscape lies ahead. Executives-and not justthose in high-tech or information companies-will be forced to re-think the strategic fundamentals of their businesses. Over thenext decade, the new economics of information will precipitatechanges in the structure of entire industries and in the ways com-panies compete.

Early signs of this change are not hard to find. Consider the re-cent near-demise of Encyclopedia Britannica, one of the strongestand best-known brand names in the world. Since 1990, sales ofBritannica's multivolume sets have plummeted by more than50%. CD-ROMs came from nowhere and devastated the printedencyclopedia business as we traditionally understand it.

How was that possible? The Encyclopaedia Britannica sells forsomewhere in the region of $1,500 to $2,200. An encyclopedia onCD-ROM, such as Microsoft Encarta, sells for around $50. Andmany people get Encarta for free because it comes with their per-

Pbilip B. Evans is a senior vice president of the Boston Consulting Groupin Boston, Massachusetts. Thomas S. Wurster is a vice president of theBoston Consulting Group in its Los Angeles office. They are the globalcoleaders of BCG's Media &) Convergence Practice, which provides con-sulting services to media companies and to a wide variety of otherclients focused on the convergence of media, information, telecommuni-cations, and computing.

DRAWINGS BY SALLY COMPORT n


sonal computers or CD-ROM drives. The cost ofproducing a set of encyclopedias -printing, binding,and physical distribution-is about $200 to $300.The cost of producing a CD-ROM is about $1.50.This is a spectacular, if small, example of the wayinformation technologies and new competition candisrupt the conventional value proposition of an es-tablished business.

Imagine what the people at Britannica thoughtwas happening. The editors probably viewed CD-ROMs as nothing more than electronic versions ofinferior products. Encarta's content is licensed fromthe Funk & Wagnalls encyclopedia, which was his-torically sold in supermarkets. Microsoft merelyspruced up that content with public-domain illus-trations and movie clips. The way Britannica's edi-tors must have seen it, Encarta was not an encyclo-pedia at all. It was a toy.

Judging from their initial inaction, Bhtannica'sexecutives failed to understand what their cus-tomers were really buying. Parents had been buyingBritannica less for its intellectual content than outof a desire to do the right thing for their children.Today when parents want to "do the right thing,"they buy their kids a computer.

The computer, then, is Bhtannica's real competi-tor. And along with the computer come a dozenCD-ROMs, one of which happens to be-as far asthe customer is concerned-a more-or-less perfectsubstitute for the Bhtannica.

When the threat became obvious, Britannica didcreate a CD-ROM version-but to avoid undercut-ting the sales force, the company included it freewith the printed version and charged $ 1,000 to any-one buying the CD-ROM by itself. Revenues con-tinued to decline. The best salespeople left. And

The way Britannica's editorsmust have seen it, Encartcan encyclopedia, it was a toy.

Britarmica's owner, a trust controlled by the Uni-versity of Chicago, finally sold out. Under newmanagement, the company is now trying to rebuildthe business around the Internet.

Britannica's downfall is more than a parableabout the dangers of complacency. It demonstrateshow quickly and drastically the new economics ofinformation can change the rules of competition,allowing new players and substitute products torender obsolete such traditional sources of compet-

itive advantage as a sales force, a supreme brand,and even the world's best content.

When managers hear this story, many respond,"Interesting, but it has nothing to do with my busi-ness. Britannica is in an information business.Thank goodness I'm not." They feel less secure,however, when they learn that the largest chunk ofBritannica's cost structure was not the editorialcontent-which constituted only about 5% ofcosts-but the direct saies force. Britannica's vulner-ability was due mainly to its dependence on the eco-nomics of a different kind of information: theeconomics of intensive personal selling. Many busi-nesses fit that description, among them automo-biles, insurance, real estate, and travel.

Every Business Is anInformation Business

In many industries not widely considered informa-tion businesses, information actually represents alarge percentage of the cost structure. Ahout one-third of the cost of health care in the United States-some $300 billion -is the cost of capturing, storing,and processing such information as patients'records, physicians' notes, test results, and insur-ance claims.

More fundamentally, information is the glue thatholds together the structure of all businesses. Acompany's value chain consists of all the activitiesit performs to design, produce, market, deliver, andsupport its product. The value chains of companiesthat supply and buy from one another collectivelymake up an industry's value chain, its particularconfiguration of competitors, suppliers, distribu-tion channels, and customers.'

When we think about a valuechain, we tend to visualize a linearflow of physical activities. But thevalue chain also includes all the in-

t formation that flows within a com-pany and between a company andits suppliers, its distributors, and itsexisting or potential customers. Sup-plier relationships, brand identity,

process coordination, customer loyalty, employeeloyalty, and switching costs all depend on variouskinds of information.

When managers talk about the value of customerrelationships, for example, what they really meanis the proprietary information that they have abouttheir customers and that their customers haveabout the company and its products. Brands, afterall, are nothing but the information-real or imag-ined, intellectual or emotional-that consumers

72 HARVARD BUSINESS REVIEW September-October 1997

NEW ECONOMICS OF INFORMATION

have in their heads about a product. Andthe tools used to build brands-advertis-ing, promotion, and even shelf space-are themselves information or ways ofdelivering information.

Similarly, information defines suppli-er relationships. Having a relationshipmeans that two companies have estab-lished certain channels of communi-cation built around personal acquain-tance, mutual understanding, sharedstandards, electronic data interchange(EDI) systems, or synchronized produc-tion systems.

In any buyer-seller relationship, in-formation can determine the relativebargaining power of the players. Autodealers, for example, know the bestlocal prices for a given model. Cus-tomers - unless they invest a lot of timeshopping around-generally do not.Much of the dealer's margin dependson that asymmetry of information.

Not only does information define andconstrain the relationship among thevarious players in a value chain, but in many busi-nesses it also forms the basis for competitive ad-vantage-even when the cost of that information istrivial and the product or service is thoroughlyphysical. To cite some of the best-known examples,American Airlines for a long time used its controlof the SABRE reservation system to achieve higherlevels of capacity utilization than its competitors.Wal-Mart has exploited its EDI links with suppliersto increase its inventory turns dramatically. AndNike has masterfully employed advertising, en-dorsements, and the microsegmentation of its mar-ket to transform sneakers into high-priced fashiongoods. All three companies compete as much on in-formation as they do on their physical product.

In many ways, then, information and the mecha-nisms for delivering it stabilize corporate and in-dustry structures and underlie competitive advan-tage. But the informational components of valueare so deeply embedded in the physical value chainthat, in some cases, we are just beginning to ac-knowledge their separate existence.

When information is carried by things-by asalesperson or by a piece of direct mail, for exam-ple - it goes where the things go and no further.It is constrained to follow the linear flow of thephysical value chain. But once everyone is connect-ed electronically, information can travel by itself.The traditional link between the flow of product-related information and the flow of the product it-

self, between the economics of information andthe economics of things, can be broken. What istruly revolutionary about the explosion in connec-tivity is the possibility it offers to unbundle infor-mation from its physical carrier.

The Trade-Off BetweenRichness and ReachLet's back up for a minute to consider why this issuch a revolutionary proposition. To the extentthat information is embedded in physical modes ofdelivery, its economics are governed by a basic law:the trade-off between richness and reach. Reachsimply means the number of people, at home or atwork, exchanging information. Richness is definedby three aspects of the information itself. The firstis bandwidth, or the amount of information thatcan be moved from sender to receiver in a giventime. Stock quotes are narrowband; a film is broad-band. The second aspect is the degree to which theinformation can be customized. For example, anadvertisement on television is far less customizedthan a personal sales pitch but reaches far morepeople. The third aspect is interactivity. Dialogueis possible for a small group, but to reach millionsof people the message must be a monologue.

In general, the communication of rich informa-tion has required proximity and dedicated charmelswhose costs or physical constraints have limited

HARVARD BUSINESS REVIEW September-October 1997 73

COMPETING IN

The Traditional Economics of Information

Richness(bandwidth,customization,interactivity)

traditionaitrade-off

Reach (connectivity)

the size of the audience to which the informationcould he sent. Conversely, the communication ofinformation to a large audience has required com-promises in bandwidth, customization, and inter-activity. (See the graph "The Traditional Econom-ics of Information.") This pervasive trade-off hasshaped how companies communicate, collaborate,and conduct transactions internally and with cus-tomers, suppliers, and distrihutors.

'•' A company's marketing mix, for example, is de-termined hy apportioning resources according tothis trade-off. A company can embed its message inan advertisement, a piece of customized directmail, or a personal sales pitch-alternatives increas-ing in richness but diminishing in reach.

When companies conduct husiness with one an-other, the numher of parties they deal with is in-versely proportional to the richness of the informa-tion they need to exchange; Citibank can tradecurrencies with hundreds of other banks eachminute hecause the data exchange requires littlerichness; conversely, Wal-Mart has narrowed itsreach hy moving to fewer and larger long-term sup-plier contracts to allow a richer coordination ofmarketing and logistical systems.

Within a corporation, traditional concepts ofspan of control and hierarchical reporting are predi-cated on the helief that communication cannot herich and broad simultaneously. Johs are structuredto channel rich communication among a few peo-ple standing in a hierarchical relationship to oneanother (upward or downward), and hroader com-munication is effected through the indirect routesof the organizational pyramid. Indeed, there is anentire economic theory (pioneered by Ronald H.Coase and Oliver E. Williamson^) suggesting thatthe boundaries of the corporation are set by the eco-nomics of exchanging information: organizationsenable the exchange of rich information amonga narrow, internal group; markets enahle the ex-change of thinner information among a larger,external group. The point at which one mode he-

comes less cost-effective than the other determinesthe boundaries of the corporation.

The trade-off hetween richness and reach, then,not only governs the old economics of informationbut also is fundamental to a whole set of premisesabout how the husiness world works. And it is pre-cisely this trade-off that is now being hlown up.

The rapid emergence of universal technical stan-dards for communication, allowing everyhody tocommunicate with everybody else at essentiallyzero cost, is a sea change. And it is as much theagreement on standards as the technology itselfthat is making this change possihle. It's easy to getlost in the technical jargon, but the important prin-ciple here is that the same technical standards un-derlie all the so-called Net technologies: the Inter-net^ which connects everyone; extianets, whichconnect companies to one another; and intranets,which connect individuals within companies.

Those emerging open standards and the explo-sion in the number of people and organizations con-nected by networks are freeing information fromthe channels that have heen required to exchangeit, making those channels unnecessary or uneco-nomical. Although the standards may not be idealfor any individual application, users are findingthat they are good enough for most purposes today.And they are improving exponentially. Over time,organizations and individuals will be able to extendtheir reach hy many orders of magnitude, oftenwith a negligible sacrifice of richness.

Where once a sales force, a system of branches, aprinting press, a chain of stores, or a delivery fleetserved as formidable barriers to entry because theytook years and heavy investment to huild, in thisnew world, they could suddenly become expensiveliabilities. New competitors on the Internet will beable to come from nowhere to steal customers.Similarly, the replacement of expensive, propri-etary, legacy systems with inexpensive, open ex-tranets will make it easier and cheaper for compa-nies to, for example, hid for supply contracts, join avirtual factory, or form a competing supply chain.

Inside large corporations, the emergence of uni-versal, open standards for exchanging informationover intranets fosters cross-functional teams andaccelerates the demise of hierarchical structuresand their proprietary information systems. (See theinsert "The End of Channels and Hierarchies.")

The Deconstruction of the Value ChainThe changing economics of information threatento undermine established value chains in many sec-tors of the economy, requiring virtually every com-


STRATEGY AND THE NEW ECONOMICS OF INFORMATION

The End of Channels and HierarchiesIn today's world, rich content passes through media, which we call channels, that can reach only a limited audience.The existence ofchannels creates hierarchy, both of choice (people have to gather rich information in an order dictated by the structure of the chan-nels) and of power (some people have better access to rich information than others do). Hierarchy of choice is illustrated by the deci-sion tree along which consumers are compelled to do their shopping in the physical world:they must choose a street,then a shop,then a department, then a shelf, then a product.They cannot select in any other sequence.They can return to the street and searchalong a different path.of course, but only by expending time and effort.

PHierarchical Decision Tree

/"Hierarchy of power is illustrated by the traditional organization•' chart, in which senior executives have a wider span of knowl-( edgethandotheirsubordinates.

Hierarchical Organization

Hierarchy enables richness but constrains choice and createsasymmetries in jnformation.The alternative to hierarchy is mar-tos, which are symmetrical and open to the extent that they areperfectBui traditional markets trade only in less rich information.

When the trade-off between richness and reach is elimi-nated, channels are no longer necessary: everyone communi-cates richly with everyone else on the basis of shared standards.This might be termed hyperarchy afler the hyperlinks of theWorld Wide Web.

Hyperarchy

The World Wide Web is a hyperarchy. So are a deconstructedvalue chain within a business and a deconstructed supplychain within an industry. So are intranets. So are structuresallowing fluid, team-based collaboration at work. So, too, is thepattern of amorphous and permeable corporate boundariescharacteristic of the companies in Silicon Valley. (So, too, inci-dentally,arethe architectures of object-oriented programmingin software and of packet switching in telecommunications.)

Hyperarchy challenges all hierarchies, whether of logic or ofpower, with the possibility (or the threat) of random access andinformation symmetry, it challenges all markets with the possi-bility that far richer information can be exchanged than that in-volved in trading products and certificates of ownership. Whenthe principles of hyperarchy are thoroughly understood, theywill provide a way to understand not only positioning strate-gies within businesses and industries but also more fundamen-tal questions of corporate organization and identity.

pany to rethink its strategy-not incrementally, butfundamentally. What will happen, for instance, tocategory killers such as Toys "R" Us and Home De-pot when a search engine on the Internet gives con-sumers more choice than any store? What will hethe point of having a supplier relationship withGeneral Electric when it posts its purchasing re-quirements on an Internet huUetin board and enter-

tains bids from anybody inclined to respond? Whatwill happen to health care providers and insurers ifa uniform electronic format for patient recordseliminates a major barrier that today discouragespatients from switching hospitals or doctors?

Consider the future of newspapers, which likemost businesses are built on a vertically integratedvalue chain. Journalists and advertisers supply



copy, editors lay it out, presses create the physicalproduct, and an elaborate distribution system deliv-ers it to readers each morning.

Newspaper companies exist as intermediaries be-tween the journalist and the reader because thereare enormous economies of scale in printing anddistribution. But when high-resolution electronictablets advance to the point where readers considerthem a viable alternative to newsprint, those tradi-tional economies of scale will become irrelevant.Editors-or even journalists - will be able to E-mailcontent directly to readers.

Freed from the necessity of subscribing to entirephysical newspapers, readers will be able to mixand match content from a virtually unlimited num-ber of sources. News could be downloaded dailyfrom different electronic-news services. Movie re-views, recipes, and travel features could come justas easily from magazine or book publishers. Starcolumnists, cartoonists, or the U.S. Weather Ser-vice could send their work directly to subscribers.Intermediaries-search engines, alert services, for-matting software, or editorial teams-could formatand package the content to meet readers' individualinterests. It does not follow that all readers willchoose to unbundle all the current content of thephysical newspaper, but the principal logic for thatbundle- the economics of printing- will be gone.

This transformation is probably inevitable butdistant. As newspaper executives correctly pointout, the broadsheet is still an extraordinarily cheapand user-friendly way to distribute information.Little electronic tablets are not going to replace itvery soon.

However, the timing of total deconstruction isnot really the issue. Pieces of the newspaper can be

Retail banks will not becomeobsolete, but their currentbusiness definition will.

unbundled today. Classified advertising is a naturalon-line product. Think how much easier it wouldbe to submit, pay for, update, search through, andrespond to classified ads electronically. Strippingaway classifieds, however, would remove 25% ofthe typical newspaper's revenues but less than 10%of its costs.

Newspaper companies have moved aggressivelyinto the electronic-classifieds business. They haveexploited their advantage as makers of the original

print marketplace to provide an integrated printand electronic offering that reaches the widest pop-ulation of buyers and sellers. This electronic offer-ing preserves the margins of 60% to 80% that news-papers need from the classifieds to cover their fixedprinting costs.

But as more and more people use the electronicmedium, companies focused on targeted segmentsof the electronic-classifieds market (operating on,say, 15% margins) will gain share. The greater theirshare, by definition, the more attractive they willbecome to buyers and sellers. Eventually, the news-papers will either lose business or (more likely) re-tain it by settling for much lower margins.

Either way, the subsidy that supports the fixedcosts of the print product will be gone. So newspa-pers will cut content or raise prices for readers andadvertisers, accelerating their defection. That, inturn, will create opportunities for another focusedcompetitor to pick off a different part of the valuechain. Thus the greatest vulnerability for newspa-pers is not the total substitution of a new businessmodel but a steady erosion through a sequence ofpartial substitutions that will make the currentbusiness model unsustainable.

Retail banking is ripe for a similar upheaval. Thecurrent business model depends on a vertically in-tegrated value chain through which multiple prod-ucts are originated, packaged, sold, and cross-soldthrough proprietary distrihution channels. Thehigh costs of distribution drive economies of uti-lization and scale and thus govern strategy in retailbanking as it works today.

Home electronic banking looks at first glancelike another, but cheaper, distribution channel.Many banks see it that way, hoping that its wide-

spread adoption might enable themto scale down their higher-cost phys-ical channels. Some banks are evenoffering proprietary software andelectronic transactions for free. Butsomething much deeper has hap-pened than the emergence of a newdistribution cbannel. Customersnow can access information and

make transactions in a variety of new ways.Some 10 million people in the United States reg-

ularly use personal-financial-management soft-ware such as Intuit's Quicken or Microsoft Moneyto manage their checkbooks and integrate theirpersonal financial affairs. Current versions of theseprograms can use modems to access electronicswitches operated by CheckFree or VISA Interac-tive, which in turn route instructions or queries tothe customers' banks. Such a system lets customers


.TRATEGY AND THE NEW ECONOMICS OF INFORMATION

pay bills, make transfers, receive elec-tronic statements, and seamlessly inte-grate account data into their personal fi-nancial plans. In addition, almost allfinancial institutions supply informa-tion at their Web sites, which anybodyon-line can access using a browser.

No single software program canachieve both richness and reach, yet.Quicken, Money, and proprietary banksoftware permit rich exchanges butonly with the customer's own bank.Web browsers do much less but reachthe entire universe of financial institu-tions. However, the software vendorsand switch providers have the re-sources, and ultimately will be motivat-ed, to form alliances with financial in-stitutions to eliminate this artificialtrade-off. Bridges between financialmanagement software and the Web,combined with advances in reliability,security, digital signatures, and legallybinding electronic contracts, will en-able financial Web sites to provide thefull range of banking services.

If that happens, the trade-off between richnessand reach will be broken. Customers will be able tocontact any financial institution for any kind of ser-vice or information. They will be able to maintaina balance sheet on their desktop, drawing on datafrom multiple institutions. They will be able tocompare alternative product offerings and to sweepfunds automatically between accounts at differentinstitutions. Bulletin boards or auctioning softwarewill allow customers to announce their productrequirements and accept bids. Chat rooms willpermit customers to share information with eachother or get advice from experts.

The sheer breadth of choice available to potentialcustomers will create the need for third parties toplay the role of navigator or facilitating agent. Forexample, some companies will have an incentiveto create (or simply make available) databases oninterest rates, risk ratings, and service histories.Others will create insurance and mortgage calcula-tors or intelligent-agent software that can searchfor and evaluate products. Still other companieswill authenticate the identity of counterparties urserve as guarantors of performance, confidentiality,or creditworthiness. (See the diagram "The Trans-formation of Retail Banking.")

As it becomes easier for customers to switchfrom one supplier to another, the competitive valueof one-stop shopping and established relationships

will drop. Cross-selling will become more difficult.Information about customers' needs or behaviorwill be harder for companies to obtain. Competi-tive advantage will be determined product by prod-uct, and therefore providers with broad productlines will lose ground to focused specialists.

In this new world, distribution will be done bythe phone company, statements by financial man-agement software, facilitation by different kindsof agent software, and origination by any number ofdifferent kinds of product specialists. The integrat-ed value chain of retail banking will have been de-constructed.

Deconstructed but not destroyed. All the oldfunctions will still be performed, as well as somenew ones. Banks will not become obsolete, buttheir current business definition will-specifically,the concept that a bank is an integrated businesswhere multiple products are originated, packaged,sold, and cross-sold through proprietary distribu-tion channels.

Many bankers-like encyclopedia executives-deny all this. They argue that most customers donot have personal computers and that many whodo arc not choosing to use them for banking. Theypoint out that people worry about the security ofon-line transactions and that consumers trustbanks more than they trust software companies.All true. However, on-line technology is advancinginexorably. And because they generate a dispropor-



The Transformation of Retail BankingIn today's integrated business model, the retail bank stands between the customer and the full range of financialservices. But soon, through Intemet technologies, customers will have direa access to product providers. As choicesproliferate, totally new businesses will arise to help customers navigate through the expanded range of banking options.

Reconfigured Business ModelIntegrated Business Model

Product Originaiors Navigators

1 B DBank Nonbank SearthEnqrw Databaw

HimBank

Product origination and packaging

inn

Transaction processing and account origination

Retail saies and distribution

78 PHOTOS: LEFT, PHOTODISC; RIGHT, STEPHEN SIMPSON/FPG

STRATEGY AND OFJN!,LQMiAIimi»

tionate share of deposits and fees, the 10% of thepopulation that currently use personal-financial-management software probably account for 75% ofthe profits of the banking system.

Market research suggests that Quicken users aremore likely to be loyal to their software than totheir hanks. In one study, half of them said that ifthey were changing hanks anyway, they would re-quire their new hank to support the software - thatis, allow them to transact their business on-lineusing Quicken. Now, bank accounts churn at therate of about 10% per year. If a bank that doesn'tsupport Quicken loses half of the new Quicken-using customers it might otherwise attract everyyear, and if such customers churn at the averagerate, then it follows that the hank will lose 3% to5% of its retail-customer margin per year. Refusalto support Quicken (or provide an acceptahle alter-native) could undermine the entire value of a fran-chise within just a few years.

The deconstruction of the value chain in bankingis not unprecedented. Fifteen years ago, corporatebanking was a spread business -that is, banks mademoney by charging a higher interest rate for loansthan they paid for deposits. Their business modelrequired them to form deep relationships with theircorporate customers so that they could pump theirown products through that distrihution system. Butthen, thanks to technology, corporate customersgained access to the same financial markets thatthe banks used. Today, corporate hanking consistsof small businesses that largely stand alone (evenwhen they function under the umbrella of a bigbank) and compete product by product. Creditflows directly from the ultimate lender to the ulti-mate borrower, facilitated by bankers who rate therisk, give advice, make markets, and serve as custo-dians. The bankers make money through the feesthey charge for those individual services. Clientsno longer bundle their purchases, and relationshipsare more volatile. Once critical, an advantage indistribution today counts for little.

Newspapers and banking are not special cases.The value chains of scores of other industries willbecome ripe for unbundling. The logic is most com-pelling-and therefore likely to strike soonest-ininformation businesses where the cost of physicaldistribution is high: newspapers, ticket sales, insur-ance, financial information, scientific publishing,software, and of course encyclopedias. But in anybusiness whose physical value chain has been com-promised for the sake of delivering information,there is an opportunity to unbundle the two, creat-ing a separate information business and allowing(or compelling) the physical one to be streamlined.

All it will take to deconstruct a business is a com-petitor that focuses on the vulnerable sliver of in-formation in its value chain. (See the insert "WhatWill Happen to Your Business?")

Implications forCompetitive AdvantageDeconstructing a vertically integrated value chaindoes more than transform the structure of a busi-ness or an industry-it alters the sources of compet-itive advantage. The new economics of informationtherefore not only present threats to establishedbusinesses but also represent a new set of opportu-nities. Every industry will shift according to itsown dynamics, and those shifts will occur at differ-ent speeds and with varying intensity. No single setof predictions can be applied across the board, butsome fundamental strategic implications of thechanging economics of information can be drawn:

Existing value chains will fragment into multiplebusinesses, each oi which will have its own sourcesof competitive advantage. When individual func-tions having different economies of scale or scopeare bundled together, the result is a compromise ofeach-an averaging of the effects. When the bundlesof functions are free to re-form as separate busi-nesses, however, each can exploit its own sourcesof competitive advantage to the fullest.

What Will Happen to Your Business?All businesses will eventually be affected by the shifting eco-nomics of information, but not all at the same rate or in thesame way. Answers to the following questions are a first stepin determining how a business could be restructured:

1, How and where in the current value chain of this businessis information a component of value?2, Where are trade-offs currently being made between rich-ness and reach in this business?3, In what situations will these trade-offs be eliminated?4,Which critical activities-especialiy informational activities-could be peeled off as stand-alone businesses?5, Could the underlying physical business be run more effi-ciently if the information functions were stripped away?6, What new activities-especialiy facilitating-agent roles-might be required?7, Among the successor businesses, how would risks andrewards be distributed?8, How would losing control over key activities affect the prof-itability of the current business model?9,Which current strategic assets could become liabilities?10. What new capabilities are needed to dominate the newbusinesses that will emerge?



Take, for example, car retailing in the UnitedStates. Dealerships provide information about prod-ucts in showrooms and through test-drives. Theyhold inventory and distribute cars. They brokerfinancing. They make a market in secondhand cars.They operate maintenance and repair services. Al-though most of these activities are physical, thebundle of functions is held together by the classicinfoTmational logic of one-stop shopping. A deal-er's competitive advantage is therefore based on amixture of location, scale, cost, sales force manage-ment, quality of service, and affiliations with carmanufacturers and banks.

Bundling these functions creates compromises.Each step in the value chain has different econo-mies of scale. If the functions were unbundled, spe-cialty companies that offer test-drives could takecars to prospective buyers' homes. Distributors ofnew cars could have fewer and larger sites in orderto minimize inventory and transportation costs.Providers of after-sales service would be free to op-erate more and smaller local facilities to furnishbetter service. Auto manufacturers could deliverproduct information over the Internet. And car pur-chasers could obtain financing by putting theirbusiness out for bid via an electronic broker. Elimi-nate the informational glue that combines all thesefunctions in a single, compromised husiness model,and the multiple businesses that emerge will evolvein radically different directions.

Some new businesses will benefitfiom network economies of scale,which can give rise to monopolies. In anetworked market, the greater the num-ber of people connected, the greater thevalue of being connected, thus creatingnetwork economies of scale. There is nopoint, for example, in heing the onlyperson in the world who owns a tele-phone. As the number of people whoown telephones rises, the value to anyone individual of hooking up increasesprogressively.

This self-reinforcing dynamic buildspowerful monopolies. Businesses thatbroker information, make markets, orset standards are all taking advantage ofthis dynamic. The implication: the firstcompany to achieve critical mass willoften take all, or nearly a l l - althoughthe continuing battle between first-mover Netscape and Microsoft in themarket for network hrowsers illustratesthat the lead of the first mover is not al-ways definitive.

Reaching critical mass can be an enormous chal-lenge. General Electric may have solved the proh-1cm by using its own huge purchasing power. GE hasopened its internal electronic-procurement systemto other buyers of industrial goods, turning its ownsourcing system into a market-making business.

As value chains fragment and reconfigure, newopportunities will arise for purely physical busi-nesses. In many businesses today, the efficiency ofthe physical value chain is compromised for thepurpose of delivering information. Shops, for exam-ple, try to be efficient warehouses and effectivemerchandisers simultaneously and are often reallyneither. The new economics of information willcreate opportunities to rationalize the physicalvalue chain, often leading to businesses whosephysically based sources of competitive advantagewill be more sustainable.

Gonsider the current battle in bookselling. Ama-zon.com, an electronic retailer on the Web, has nophysical stores and very little inventory. It offers anelectronic list of 2.5 million hooks, ten times largerthan that of the largest chain store, and customerscan search through that list by just about any crite-rion. Amazon orders most of its books from two in-dustry wholesalers in response to customers' re-quests. It then repacks and mails them from acentral facility.

Amazon cannot offer instant delivery,- nor cancustomers physically browse the shelves the way

80 HARVARD BUSINESS REVIEW Stpttmber-October 1997

STRATEGY AND THE NEW ECONOMICS OF INFORMATION

they can in a traditional bookstore. Its advantagesare based on superior information and lower physi-cal costs. Customers can, for example, aeeess bookreviews. They have greater choice and bettersearching capabilities. And Amazon saves moneyon inventory and retail space.

But Amazon's success is not a given. The dis-count chains are aggressively launching their ownWeb businesses. There is nothing defensible aboutAmazon's wide selection since it really comes frompublishers' and wholesalers' databases. By double-handling the books, Amazon still incurs unneces-sary costs.

In fact, the wholesalers in the book industrycould probably create the lowest-cost distrihutionsystem by filling customers' orders directly. If com-petition pushes the industry in that direction, elee-tronic retailers would become mere search enginesconnected to somebody else's database-and thatwould not add much value or confer on them muchof a competitive advantage. The wholesalers couldbe the big winners.

When a company focuses on different activities,the value proposition underlying its brand identitywill change. Because a hrand reflects its company'svalue chain, deconstruction will require new brandstrategies. For instance, the importance of branchesand automated teller machines today leads manybanks to emphasize ubiquity in their brand image(Citibank, for example). However, the reconfigura-tion of financial services might lead a company tofocus on being a product provider. For such a strate-gy, performance becomes the key message, as it isfor Fidelity. Another brand strategy might focus onhelping customers navigate the universe of third-party products. The key message would be trust, asit is for Charles Schwab.

New branding opportunities will emerge for thirdparties that neither produce a product nor deliver aprimary service. Navigator or agent brands havebeen around for a long time. The Zagat guide torestaurants and Consumer Reports are two obviousexamples. It's Zagat's own brand-its credibility inrestaurant reviewing- that steers its readers toward

Where the New Businesses Will EmergeIn a world of limited conneaivity, choices at each point in thevalue chain are, by definition, finite. In contrast, broadband con-neaivity means infinite choice. But infinite choice also meansinfinite bewilderment,This navigation problem can be solved inall sorts of ways,and each solution is a potential business.

The navigator could be a database. The navigator could bea search engine,The navigator could be intelligent-agent soft-v\/are,The navigator could be somebody giving advice. Thenavigator could be a brand providing recommendations orendorsements.

The logic of navigation can be observed in a number of busi-nesses in which choice has proliferated. People often react toclutter by going back to the tried and true. Customer researchindicates that peopie faced with complex choices either gravi-tate tov\/ard dominant brands or confine their search to narrowformats, each offering a presorted set of alternatives. In thegrocery store, for example, where the number of products hasquadrupled over the last 15 years, hundreds of segmentedspecialty brands have gained market share in almost every cat-egory. But so have the one or two leading brands.The prolifera-tion of choice has led to the fragmentation of the small brandsand the simultaneous concentration of the large ones. Thelosers are the brands in the middle.

Similarly, television viewers seem to flock to the hit showswithout caring which network those shows are on. But they se-lect specialty programming, such as nature documentaries ormusic videos, by tuning in to a cable channel offering that for-mat.ln essence,the viewer selects the channel,and the channelselects the content. In the first case, the product's brand pullsvolume through the channel; in the second, the channel'sbrand pushes content toward receptive viewers.

Those two approaches by the consumer yield different pat-terns of competitive advantage and profitability. Networksneed hit shows more than the hit shows need any network:the producers have the bargaining power and therefore re-ceive the higher return. Conversely, producers of low-budgetnature documentaries need a distributor more than the dis-tributor needs any program, and the profit pattern is, there-fore, the reverse. In one year, the popular comedian Bill Cosbyearned more than the entire CBS network; the DiscoveryChannel probably earns more than all of its content provid-ers put together. Despite the fact that CBS's 1996 revenueswere about six times those of the Discovery Channel, Discov-ery's 52% profit margin dwarfed CBS's 4%.

The economics playing out in the television industry are amodel for what will likely emerge in the world of universal con-nectivity. Think of it as two different value propositions: one isa focus on popular content; the other, a focus on navigation.

Navigation might have been the right strategy for Ency-clopaedia Britannica in responding to the threat from CD-ROMs. Its greatest competitive asset, after all, was a brand thatcertified high-quality,objective information,Given the clutter ofcyberspace, what could be more compelling than a Britannica-branded guide to valuable information on the Internet?

If Britannica's executives had written off their sales force, ifthey had built alliances with libraries and scientific journals,if they had built a Web site that had hot iinks directly to originalsources, if they had created a universal navigator to valuableand definitive information validated by the Encyclopaedia Bri-tannica brand, they would have been heroes,They might haveestablished a monopoly, following the example of Bill Gates, Infaa,he might have been forced to acquire them.

HARVARD BUSINESS REVIEW September-October 1997


a particular establishment. A more recent exampleis the Platform for Internet Content Selection(PICS), a programming standard that allowsbrowsers to interpret third-party rating labels onWeb sites. With it a parent might search for sitesthat have been labeled "safe for children" by Evalu-Web. PICS enables anybody to rate anything, andit makes those ratings ubiquitous, searchable,sortable, and costless. The dramatic proliferation ofnetworked markets increases the need for suchnavigators and other facilitating agents, those thatguarantee a product's performance or assume risk,for example. Thus there will be many new opportu-nities to develop brands. (See the insert "Where theNew Businesses Will Emerge.")

Bargaining power will shift as a result of a radicalreduction in the ability to monopolize the controlof information. Market power often comes fromcontrolling a choke point in an information chan-nel and extracting tolls from those dependent onthe flow of information through it. For example,sellers to retail customers today use their controlover the information available to those eustomersto minimize comparison shopping and maximizecross-selling. But when richness and reach extendto the point where such channels are unnecessary,that game will stop. Any choke point could then hecircumvented. Buyers will know their alternativesas well as the seller does. Some new intermedi-aries-organizers of virtual markets-may evenevolve into aggregators of buying power, playingsuppliers off against one another for the benefit ofthe purchasers they represent.

Customers' switching costs will drop, and com-panies will have to develop new ways oE generatingcustomer loyalty. Common standards for exchang-ing and processing information and the growingnumbers of individuals accessing networks willdrastically reduce switching costs.

Proprietary EDI systems, for example, lock com-panies into supply relationships. But extranetslinking companies with their suppliers using theInternet's standard protocols make switching al-most costless. The U.S. auto industry is creatingsuch an extranet called the Automotive Networkexchange (ANX). Linking together auto manufac-turers with several thousand automotive suppliers,the system is expected to save its participantsaround a billion dollars a year, dramatically reduce

ordering and billing errors, and speed the flow ofinformation to second- and third-tier suppliers. Byreducing switching costs and creating greater sym-metry of information, ANX will intensify competi-tion at every level of the supply chain.

Incumbents could easily become victims of theiiobsolete physical infrastructures and their ownpsychology. Assets that traditionally offered com-petitive advantages and served as harriers to entrywill become liabilities. The most vulnerable com-panies are those currently providing informationthat could be delivered more effectively and inex-pensively electronically-for example, the physicalparts of sales and distribution systems, such asbranches, shops, and sales forces. As with newspa-pers, the loss of even a small portion of customersto new distribution channels or the migration of ahigh-margin product to the electronic domain canthrow a business with high fixed costs into a down-ward spiral.

It may be easy to grasp this point intellectually,but it is much harder for managers to act on its im-plications. In many businesses, the assets in ques-tion are integral to a company's core competence. Itis not easy psychologically to withdraw from assetsso central to a company's identity. It is not easystrategically to downsize assets that have highfixed costs when so many customers still prefer thecurrent business model. It is not easy financially tocannibalize current profits. And it is certainly noteasy to squeeze the profits of distributors to whomone is tied by long-standing customer relationshipsor by franchise laws.

Newcomers suffer from none of these inhibi-tions. They are unconstrained by management tra-ditions, organizational structures, customer rela-tionships, or fixed assets. Recall the cautionary taleof Encyclopaedia Britannica. Executives must men-tally deconstruct their own businesses. If theydon't, someone else will.1. FOI a complete discussion of the value chain concept, seeMichael Poitei's Competitive Advantage (New York: The FreePress. 1985). Differences in value chains-that is, differences inhovf competitors perform strategic activities or differencesin v^hich activities they choose to perform - are the basis forcompetitive advantage.2. Ronald H. Coase, "The Nature of the Firm." Economica, vol. 4.no. 4. 1937. p. 386: Oliver E. Williamson, Markets and Hierar-chies: Analysis and Antitrust Implications (New York: FreePress. 1975).

Reprint 97504 To order reprints, see the last page of this issue.
