Rappa_2004.pdf

7/29/2019 Rappa_2004.pdf

1/11

The utility businessmodel and the future of

computing services

by M. A. Rappa

The utility business model is shaped by anumber of characteristics that are typical inpublic services: users consider the service anecessity, high reliability of service is critical,the ability to fully utilize capacity is limited,and services are scalable and benefit fromeconomies of scale. This paper examines theutility business model and its future role in theprovision of computing services.

The idea of utility computing has received attention

recently and for good reason. The use of computerscontinues to be a rapidly expanding feature of mod-ern society, and industry has come to rely on com-puters to perform a multitude of tasks beyond sim-ple data processing and storage. Computer networkshave extended the reach of computing to connectbusinesses across the supply chain and, in many in-stances, directly to the consumer. With the growthof the Internet, the computer has come to play aneven greater role in commerce.

Computing has also become a larger and more in-timate part of daily life for many people. Individ-

uals now use computers to accomplish a wide arrayof tasks, from the complex to the mundane. 13

Whether it is used for communicating by e-mail andinstant messaging, paying bills and managing per-sonal finances, or the pursuit of hobbies and enter-tainment, the computerhas become an essential tool.Indeed, the variety of tasks performed with comput-ers today would have been difficult to foresee as lit-tle as two decades ago.

With all this progress has come a greater degree ofreliance on computers and their connectivity to net-works, and this reliance has bred high expectations

for the availability and performance of computingand networking services. This expectation is not un-like that seen in other areas of technology to whichmodern society has grown accustomed; for example,the dependence on a ready availability of affordablypriced electricity. Long ago a curiosity and a luxury,over the last century we have seen electricity growbeyond a modern everyday convenience to becomea necessity in the lives of most people.

The prominence of computers in society and ourgrowing reliance on them raises an interesting ques-

tion: Is computing the next utility? The answer tothis question has broad implications for the futureof computing. Already, the idea of utility comput-ing has begun to influence the development of com-puter technology in such areas as the auto-provision-ing of computing resources and resource sharingacross a computing grid. 4 6 Its potential role in theevolution of business models for computing servicesis of equal importance, and that role is addressed inthis paper.

Common characteristics of utilities

In many parts of the world, although by no meanseverywhere, services such as water, power, heat, light,common carrier transportation (airlines, buses, andrailroads), and telephone access are typically pro-vided by a public utility. What makes any particular

Copyright 2004 by International Business Machines Corpora-tion. Copying in printed form for private use is permitted with-outpayment of royalty provided that (1)each reproduction is donewithout alteration and (2) the Journal reference and IBM copy-right notice are included on the first page. The title and abstract,but no other portions, of this paper may be copied or distributedroyalty free without further permission by computer-based andother information-service systems. Permission to republish anyother portion of this paper must be obtained from the Editor.

RAPPA 0018-8670/04/$5.00 2004 IBM IBM SYSTEMS JOURNAL, VOL 43, NO 1, 200432

7/29/2019 Rappa_2004.pdf

2/11

service a utility is shaped by a combinationof require-ments (see Table 1), most notably: users consider ita necessity; high reliability of service is critical; easeof use is a significant factor; the full utilization ofcapacity is limited; services are scalable (leading toeconomies of scale); and exclusive rights are grantedfor providing service in a given area.

Necessity. Users depend on utility services to fulfilltheir day-to-day needs. Doing without service is anunwelcome option for them. Of course, seldom doutility services start out as essential. Its takes timefor distribution networks to spread and costs to de-cline. It also may take time for users to adapt to theservice. Once a service does take hold, it may growin importance as users discover new ways to use itto their benefit. How crucial a service becomes mayultimately depend on the circumstances of the in-dividual user. But once users do come to depend ona service, it can become a transparent part of their

everyday reality.

Reliability. The service provided by a utility must bereadily available when and where the user needs it.A temporary or intermittent loss of service may causemore than a trivial inconvenience to the user; a pro-longed loss of service maycause severe hardship. Be-cause a failure in service has undesirable conse-quences, utilities must operate with an exceptionallyhigh degree of reliability.

Providing continuous service in the face of variouscontingencies is a huge technological challenge that

utilitiesface. Because some kinds of services maynotbe easily or cheaply inventoried, if at all, redundancymust be built into production capacity to make upfor the inevitable equipment failure. Furthermore,because utilities provide on demand services, theymust deploy transparent failover mechanisms andstandby services to ensure continuous availability to

the user. If one area of a service grid fails, the sys-temmust be able to compensate and respond instan-taneously to the shortfall, thereby preventing the dis-ruption of the service.

Necessity drives user expectations of utility servicesbeyond what may be typical in other industries.Whether or not these expectations are realistic, util-ities must do their best to buffer users from the pre-dictable problems that could cause a discontinuityin service.

Usability. No matter how technologically complex

they may be on the production end, utility servicesare characteristically simple at the point of use. Usershave what could be called a plug-and-play men-tality. This is not to say that devices connected to aservice are unsophisticated, but the utility service it-self tends to exist only in thebackground. Users maybecome mindful of a utility only in those rare in-stances when the service fails to meet their expec-tations. This may explain why the public perceptionof a utility is not always positive.

One ingredient in makinga service simple at the userinterface is a high level of technical standardization.

Table 1 Requirements common to utility services

Necessity Reliability Usability Util ization Scalability Exclusivity

Water H H H M H HElectricity H H H H H HCommon Carrier Transportation M H H H M H

Telephone:POTS H H H H H MCellular M M H H H L

Radio and Television Broadcasting:Terrestrial M H H L H MSatellite M M M L H L Cable M H H L H H

Internet Access:

DSL H H H H H HCable H H H H H HDial-up M M H H H L

L Low relevance

M Medium relevance

H High relevance

IBM SYSTEMS JOURNAL, VOL 43, NO 1, 2004 RAPPA 33

7/29/2019 Rappa_2004.pdf

3/11

Devices that add user functionality to the service

must conform to the specifications of the network.Plug compatibility, independent of the vendor, is acommon feature of utility services. Even so, tech-nical standardization can be extremely difficult toachieve. In marketplaces where proprietary innova-tion is strong, the incentive for competitors to agreeon standards is weak. Although a lack of standard-ization is costly and inconvenient, premature con-sensus on a standard may forestall significant inno-vation that can be of benefit to users.

To the extent that incompatible standards take hold,in some cases the consequences can endure for longperiods, as fixed investments in infrastructure grow.Just how long this condition can last is illustrated bythe case of the differencein voltage standards aroundthe world. In such situations, technologies that en-able the conversion betweenstandards become a reg-ular and cumbersome aspect of the user experience.

Utilization rates. Utilities aredriven by a need to care-fully manage utilization rates. User demand for util-ity services can fluctuate widely over time and acrossthe service region. Because sufficient production ca-pacity must be installed to handle periods of peakdemand, overall utilization rates are typically wellbelow full capacity.

In addition to fluctuations in usage, there may bediscrete incidents of an exceptional nature when de-mand spikes sharply upward. Such spikes can occurwhen large numbers of users suddenly want to usethe service simultaneously. Other spikes may occurwhen usersfear a shortage in supplyand beginhoard-ing, to the extent that it is possible.

Underutilization in off-peak periods provides astrong economic rationale for service providers toshift user demand from peak to off-peak periods. Bypricing services according to actual metered usage

and by providing off-peak price discounts, fluctua-tions in user demand can be smoothed out over thecycle. How a service is billed may also create incen-tives for users to limit their usage.

Scalability. Utilities are commodity businesses.Therefore, utility services can exhibit significanteconomies of scale that favor larger producers oversmaller ones. As production capacity rises, the unitcost of production falls. There may be other size-related benefits as well. It might be expected that asthe demand for a service increases beyond somethreshold, the quality of service may decline as users

begin to compete with each other. However, with

some types of utilities, service can become more andmore useful as the number of users of increases.

Service exclusivity. The economies of scale in a util-ity can benefit from a monopolistic provision of ser-vices. When this is the case, the government may stepin to grant an exclusive franchise in a geographic re-gion. Government regulation of the service and howit is priced typically accompanies such a sanction.Cost-based pricing is a common formula. With thebenefits of an exclusive franchise comes the obliga-tion to serve anyand all users regardless of how prof-itable it may be for the utility.

Some of the common characteristics of a utility de-rive from its relationship with its customers. Othercharacteristics are derived from technological andbusiness aspects of how the service is produced anddistributed. The preceding list of characteristics,while important, is not meant to exclude other pos-sible factors that may be relevant to particular typesof utility.

Each of the characteristics described here may ormay not play an equal role in shaping any particulartype of utility service. Table 1 provides an evalua-tion in the most general terms of the potential rel-

evance of each factor for public utility services, in-cluding water, electricity, and common carrier (orpublic) transportation. In addition, the comparisonis extended to examine a few businesses that havesome characteristics in common with public utilities,namely radio and television broadcasting and Inter-net access services.

The utility business model

The factors of usernecessity, reliability, usability, uti-lization, scalability, and exclusivity, when taken to-gether, shape the business model for utility services.To understand the nature of the utility model, it is

useful to place it in the context of business modelsin general. A business model is a method of doingbusiness. All business models specify what a com-pany does to create value, how it is situated amongupstream and downstream partners in the valuechain, and the type of arrangement it has with itscustomers to generate revenue. In any given indus-try, the methods of doingbusiness may vary, but thereare limits imposed by technological factors, by thecompetitive dynamic among companies and betweencompanies and their channel partners, and by cus-tomer expectations and preferences, among otherthings.

RAPPA IBM SYSTEMS JOURNAL, VOL 43, NO 1, 200434

7/29/2019 Rappa_2004.pdf

4/11

There have been a number of attempts to create

schema for classifying the various types of businessmodels seen in practice, particularly in relation tothe Internet. 711 The commercialization of the In-ternet during the 1990s drew a great deal of atten-tion to business models. The Internet opened thedoor to new business opportunities, but many Inter-net-based enterprises failed because they had notclearly thought through their model particularly,

how money would be made. Nonetheless, given therapid adoption of the Internet, it may no longer bepossible to discuss business models without takingit fully into account.

One approach to the classification of e-business mod-

els is a comprehensive taxonomy using the customerrelationship as the primary dimension for definingcategories.7 Although by no means the only ap-proach, this has proven to be a useful framework be-cause it builds upon a common parlance already usedin many industries to describe methods of business.Although other approaches may be more suitablefor other purposes, it is unreasonable to expect thatany single taxonomy can account for the vast diver-sity of business models found in practice without be-coming unwieldy.

Nine major categories are used to classify a number

of different types of business models that have beenidentified in practice among Web-based enterprises(see Table 2):

Brokerage model. Brokers are market makers: theybring buyers and sellers together and facilitate trans-actions. Brokers play a frequent role in business-to-business (B2B), business-to-consumer (B2C), or con-sumer-to-consumer (C2C) markets. Usually, a bro-ker charges a fee or commission for each transac-tion it enables. The formula for fees can vary.Brokerage models include exchanges, demand col-lection systems, and auction brokerages.

Advertising model. The advertising model on the Web

is an extension of the traditional media broadcastmodel. The broadcaster, in this case a Web site, pro-vides content (usually, but not necessarily, for free)and services (like e-mail, chat, forums) mixed withadvertising messages in the form of banner ads. Thebanner ads may be the major or sole source of rev-enue for the broadcaster. The broadcaster may bea content creator or a distributor of content createdelsewhere. The advertising model works best whenthe volume of traffic is large or highly specialized.Advertisingmodelsincludeportals, query-based paidplacement, contextual advertising, and content-tar-geted advertising.

Information-intermediary model. Dataabout consum-ers and their consumption habits are valuable, es-pecially when that information is carefully analyzedand used to target marketing campaigns. Indepen-dently collected dataabout producers and their prod-ucts are useful to consumers who are considering apurchase. Some firms function as infomediaries(information intermediaries) assistingbuyers and/orsellers to understand a given market.

Merchant model. Merchants are wholesalers and re-tailers of goods and services. Sales may be madebased on list prices or through auctioning. Merchant

models include virtualmerchants or e-tailers, mail-order businesses with a Web-based catalog, and tra-ditional brick-and-mortar retail establishments withWeb storefronts.

Manufacturer Direct model. The maker of a productor service may sell (by purchase, lease, or license)directly to the consumer. The manufacturer or di-rect model is based on the power of the Web to al-low a manufacturer to reach buyers directly andthereby compress the distribution channel. The man-ufacturer model may be chosen for its efficiency, im-proved customer service, or due to a better under-

standing of customer preferences.

Affiliate model. The affiliate model providespurchaseopportunities wherever people may be surfing theWeb. Financial incentives (in the form of a percent-age of revenue) are offered to affiliated partner sites.The affiliates provide purchase-point click-through(i.e. direct linking) from their Web sites to the mer-chants Web site. It is a pay-for-performance model if an affiliate does not generate sales, no cost tothe merchant is incurred. The affiliate model is in-herently well suited to the Web, which explains itspopularity. Variations of this model include banner

To understand the

nature of the utility model,

it is useful to place it

in the context of business

models in general.


7/29/2019 Rappa_2004.pdf

5/11

Table 2 Taxonomy of e-business models

Type Model

Brokerage Marketplace ExchangeOffers a full range of services covering the transaction process, from marketassessment to negotiation and fulfillment. Exchanges operate independently or are backed by anindustry consortium.

Buy/Sell FulfillmentTakes customer orders to buy or sell a product or service, including terms like priceand delivery.

Demand Collection SystemThe patented name-your-price model pioneered by Priceline.com**.Prospective buyer makes a final (binding) bid for a specified good or service, and the broker arrangesfulfillment.

Auction BrokerConducts auctions for sellers (individuals or merchants). Broker charges the seller alisting fee and commission based on the value of the transaction. Auctions vary widely in terms of theoffering and bidding rules.

Transaction BrokerProvides a third-party payment mechanism for buyers and sellers to settle atransaction.

DistributorA catalog operation that connects a large number of product manufacturers with volume andretail buyers. Broker facilitates business transactions between franchised distributors and their tradingpartners.

Search AgentA software agent used to search for the price and availability of goods or a servicespecified by the buyer or to locate hard-to-find information.

Virtual MallA hosting service for on-line merchants that charges setup, monthly listing, and/ortransaction fees. May also provide automated transaction and relationship marketing services.

Advertising PortalUsually a search engine that may include varied content or services. A high volume of user trafficmakes advertising profitable and permits further diversification of site services. A personalized portalallows customization of the interface and content to the user. A niche portal cultivates a well-defineduser demographic.

ClassifiedsList of items for sale or wanted for purchase. Listing fees are common, but there also may bea membership fee.

Registered UserContent-based sites that are free to access but require users to register and providedemographic data. Registration allows inter-session tracking of user surfing habits and thereby generatesdata of potential value in targeted advertising campaigns.

Query-based Paid PlacementSells favorable link positioning (i.e., sponsored links) or advertising keyed toparticular search terms in a user query, such as the Overture** trademark pay-for-performance model.Contextual AdvertisingFreeware developers who bundle ads with their product. For example, a browser

extension that automates authentication and form fill-ins may also deliver advertising links or pop-upsas the user surfs the Web. Contextual advertisers can sell targeted advertising based on an individual ssurfing behavior.

Content-Targeted AdvertisingPioneered by Google**, the precision of search advertising is extended tothe rest of the Web. Google identifies the content of a Web page and then automatically deliversrelevant ads when a user visits that page.

Ultramercials**Interactive online ads that require user interaction to reach the intended content.

InformationIntermediary

Advertising NetworksA service feeding banner ads to a network of member sites, thereby enablingadvertisers to deploy large marketing campaigns. Ad networks collect data about Web users that can beused to analyze marketing effectiveness.

Audience Measurement ServiceOn-line audience market research.Incentive MarketingCustomer loyalty programs providing incentives to customers such as redeemable

points or coupons for making purchases from associated retailers. Data collected about users are sold

for targeted advertising.Merchant Virtual MerchantA retail merchant that operates solely over the Web (also known as an e-tailer).

Catalog MerchantMail-order business with a Web-based catalog which combines mail, telephone, andon-line ordering.

Click and MortarTraditional brick-and-mortar retail establishment with a Web storefront.Bit VendorA merchant who deals strictly in digital products and services and, in its purest form,

conducts both sales and distribution over the Web.

Manufacturer Purchase ModelA manufacturer that sells its products or services directly to the consumer.Direct Lease ModelA manufacturer that finances the sale or rental of its products directly to the consumer.

Licensing ModelA manufacturer, such as a software maker, that licenses its product directly to theconsumer.

Brand-Integrated ContentIn contrast to the sponsored-content approach (i.e., the advertising model),brand-integrated content is created by the manufacturer itself for the sole purpose of productplacement.


7/29/2019 Rappa_2004.pdf

6/11

exchange, pay-per-click, and revenue sharing pro-grams.

Community model. The community model is basedon user loyalty. Loyal users invest both their timeand emotions in a business. Revenue can be gener-ated based on the sale of ancillary products and ser-vices or voluntary contributions. The best known ex-ample of a community model is that ofopensourcecomputing. The businesses that have emergedaround open source products rely on revenue gen-erated from related services such as systems integra-tion, product support, tutorials, and user documen-tation. Another example is the traditional publicbroadcasting model, the listener or viewer-contrib-

utor method used in not-for-profit radio and tele-vision broadcasting. The model is based on the cre-ation of a community of interestedusers whosupportthe site through voluntary donations.

Subscription model. Users are charged a periodicdaily, monthly, or annual fee to subscribe to a ser-vice. It is not uncommon for sites to combine freecontent with premium (i.e., subscriber only ormember only) content. Subscription feesare incurredregardless of actual usage rates. Subscription and ad-vertising models are frequently combined. Examplesinclude content services, person-to-person network-

ing services, trust services, and Internet service pro-viders (ISPs).

Utility and hybrid models. The utility model is basedon metering usage and constitutes a pay as yougo approach. Unlike subscription services, meteredservices arebased on actual usage rates.Forexample,an ISP may use a utility model, charging custo-mers for connection minutes, though the subscrip-tion model is more common among ISPs operat-ing in the United States. An interesting hybridmodel on the Web, the metered subscription,allows subscribers to purchase access to content inmetered portions, such as the number of pagesviewed.

Metering customer usage is one characteristic thatfigures prominently in the utility business model andsets it apart from other models. But utilities in theoff-line world are not limited to the approach of me-tering usage (see Table 3 and Figure 1). One exam-ple is residential telephone services. For some time,the so-called plain old telephone system (or POTS)adopted a combination of metered usage for longdistance services, a subscription model for local call-ing services, and a lease model for the usage of tele-phone equipment (though nowadays equipment istypically purchased outright). Under a subscription

Table 2 Taxonomy of e-business models (continued)

Type Model

Affiliate Banner ExchangeTrades banner placement among a network of affiliated sites.Pay-per-ClickSite that pays affiliates for a user click-through.Revenue SharingOffers a percent-of-sale commission based on a user click-through in which the user

subsequently purchases a product.

Community Open SourceSoftware developed voluntarily by a global community of programmers who share codeopenly. Instead of licensing code for a fee, open source relies on revenue generated from relatedservices like systems integration, product support, tutorials, and user documentation.

Public BroadcastingUser contributor model used by not-for-profit radio and television broadcastingextended to the Web. The model is based on the creation of a community of users who support thesite through voluntary donations.

Knowledge NetworksDiscussion sites that provide a source of information based on the sharing ofexpertise among professionals.

Subscription Content ServiceProvides text, audio, or video content to users who subscribe for a fee to gain access

to the service.Person-to-Person Networking ServiceConduit for the distribution of user-submitted information, for

example, individuals searching for former schoolmates.Trust ServiceMembership association that abides by an explicit code of conduct and to which

members pay a subscription fee.Internet Service ProviderProvides network connectivity and related services.

Utility Metered UsageMeasures and bills users based on actual usage of a service.Metered SubscriptionAllows subscribers to purchase access to content in metered amounts (e.g.,

numbers of pages viewed).

Source: Rappa, M. Business Models on the Web, Managing the Digital Enterprise (http://digitalenterprise.org), May 2003.


7/29/2019 Rappa_2004.pdf

7/11

model, users pay a flat rate for monthly serviceregardless of actual usage levels.

Cellular phone services have adopted yet anothercombination of the subscription and utility models.A monthly subscriber fee for both local and long dis-tance serviceis tied to a maximum level of usage(i.e.,connection minutes), beyond which usage is meteredand billed accordingly. The subscription may comewith a minimum-length service contract, and mayalso include equipment as part of the agreement. Thepopularity of the cellular business model has recentlyled telephone serviceproviders to considerthe adop-tion of a similar approach with the introduction ofa flat-rate subscription for both local and long dis-

tance calling services bundled together.

Radio and television broadcasting offers anotherexample of how different business models can becombined in a service that exhibits characteristicssimilar to a utility. Terrestrial broadcasting ser-vices have typically depended on sponsorship inthe form of commercial advertisements that areinterspersed with programming. There is also aform of terrestrial broadcasting, which is publiclysponsored, that could be classified with commu-nity business models. Satellite and cable broad-casters use a subscription model in which the user

is charged for a basic package of bundled servicesand can also choose from a menu of premium con-

tent.

Internet access services provide an example of howbusiness models adapt to technological changes inhow a service is provided. Early in the commer-cial rollout of the Internet, services offered dial-upaccess by using a combination of business modelsnot unlike the telephone service on which that ac-cess depended. A user could pay according to ac-tual usage or pay a subscriber fee for limited (andlater unlimited) service. With the advent of res-idential broadband services, Internet access pro-viders offer always on service using a subscrip-

tion model.

Utility computing

Recent projections from IBM have envisioned utilitycomputing as an integral part of the future of infor-mation technology. IBM Global Services provides thefollowing definition:

Utility computing is the on demand delivery of in-frastructure, applications, and business processesin a security-rich, shared,scaleable, and standards-based computerenvironment overthe Internet for

Table 3 Business models of utility services

Type of Service Business Models

Water Metered usage of serviceElectricity Metered usage of serviceCommon Carrier Transportation Basic pay-as-you-go fare for one-way or roundtrip service; subscription for commuter

serviceTelephone:

POTS Subscription for local service; metered usage of long distance service; equipment isleased or purchased

Cellular Subscription with usage limits; metered usage in excess of the subscription limit;equipment purchased or bundled with subscription

Radio and Television Broadcasting:Terrestrial Advertiser-sponsored, community-sponsoredSatellite Subscription with basic package and premium services

Lease or purchase equipment

Cable Subscription with basic package and premium servicesPay-per-view for special event programming and movie selectionsLeased equipment is bundled with service

Internet Access:DSL Subscription for unlimited (always on) service

Leased equipment is bundled with serviceCable Subscription for unlimited (always on) service

Leased equipment is bundled with serviceDial-up Subscription for limited service or metered usage based upon connection time

Equipment is purchased


7/29/2019 Rappa_2004.pdf

8/11

a fee. Customers will tap into IT resourcesandpay for themas easily as they now get their elec-tricity or water. 12

In a recent interview, Irving Wladawsky-Berger,General Manager for e-business on demand, said:

The idea of accessing computing and data with-out having to own the computers probably is evenolder than the 1980s because at some level thatswhat time-sharing computing was about, and thatprobably was back in the 1970s. Its a very nicethought, which comes from watching other suc-

cessful utilities. You know, people are used to thefact that they use electricity, they turn on waterfaucets, they use the telephone, they get access totelevision, and in none of these cases do they haveto own the equipment that generates electricity.They dont have to own the water supply, or thetelephone switches, or the broadcasting compa-nies. Everybody says, Gee wouldnt it be nice ifI can similarly plug in to get access to applicationsand information? 13

Recent discussionsof computing have begun to pointto its similarity to conventional public service util-

Figure 1 Utility-service characteristics and business models

WATERMETERED SERVICE

PREPAID SERVICE

AND SUBSCRIPTION

METERED, PREPAID,

AND SUBSCRIPTION

ADVERTISING AND

COMMUNITY SPONSOR

(TERRESTRIAL)

PREPAID AND

SUBSCRIPTION(CABLE AND SATELLITE)

METERED SERVICE

AND SUBSCRIPTION

ELECTRICITY

COMMON CARRIER

TRANSPORTATION

TELEPHONE

BROADCASTING

INTERNET ACCESS

USER NECESSITY

RELIABILITY

USABILITY

UTILIZATION RATES

SCALABILITY

SERVICE EXCLUSIVITY

CHARACTERISTICS SERVICES BUSINESS MODELS


7/29/2019 Rappa_2004.pdf

9/11

ities. 14,15 To be sure, the analogy can be taken too

far, but it is instructive to examine the general ser-vice requirements of utilities in the context of com-puting to better gauge where business models maybe heading in the future. It may very well be thatcurrent business models based on purchasing, leas-ing, and licensingof products will ultimately give wayto a utility model of computing based on subscriberfees and metered services.

The importance of computers for business has be-come an indisputable fact. Computer and networkservices are an end-to-end component of many bus-iness processes. To be without service is not merelyan inconvenience; it is a potential financial disaster.It should therefore come as no surprise that busi-nesses both large and small have come to view com-puters as a necessity in the same manner as theymight view utility services. Furthermore, it is likelythat the reliance of individual users on computingservices will soon rival if not surpass their relianceon public utilities like the residential telephone asan essential service.

As reliance on computing grows, so will the expec-tation among users for reliable service. Improve-ments in computer reliability have been made, butthere is still much work to be done. Software remains

a source of instabilityin the heterogeneous computerenvironments that exist today. As software has be-come more powerful, it has also become more com-plex in terms of the underlying code. This complex-ity creates a major challenge in engineering reliablesoftware even under benign conditions, and this isexacerbated by the unrelenting security threat tocomputer networks. Current approaches to secur-ity, which can rely heavily on patching software af-terit is deployed, maybe severelyflawed asa methodfor achieving high reliability.

The growth of personal computers and the Internet

have made computing a mainstream activity. Todaythe computer user population cuts across a widespectrum in terms of age, education, and other de-mographic dimensions. Fortunately enormous pro-gress has been made to improve computer usability.The user interface for personal computing is goodevidence of success in making computers easier touse. Both hardware and software makers are mov-ing quickly in the direction of creating products withplug and play convenience. But even so, comput-ers have not yet achieved the goal of becoming assimple to useas a commonhousehold appliance. Ad-vances in usability have been offset by the rapid tech-

nological innovation that has kept designs and stan-

dards in flux.

The substantial investment made in computing in-frastructure has spurred an interest in increasing theoverall rate of its utilization. Current progress intechnology is making this possible. An example isthe virtualization of server and storage capacity andthe advent of grid computing supported by open stan-

dards. Grid computingwill ultimately present utility-computing service providers with capacity planningissues similar to those faced by public utilities. Man-aging peak demand and the economical utilizationof capacity will require incentives to modify usagepatterns. This will favor the adoption of metered us-age as a core element in the business model for util-

ity computing.

To the extent thattechnology enables computing ser-vices to be scalable, the economies of scale typicalof public utility services should also apply to utilitycomputing. The benefits maycome on several fronts,because the fixed cost of services of the utility canbe amortized across a larger population of users,thereby reducing the unit cost per user. Utility com-puting mayalso benefit from the ability to retain thenecessary skilled workforce to manage and maintaincomputing services in a way that is difficult for smalland medium-sized enterprises. This may be of par-

ticular importance in dealing with network security,a field where there is a scarcity of talent.

One manner in which utility computing may differfrom a public utility is with respect to service exclu-sivity. The trend in recent decades has been to de-regulate and encourage competition in the publicutility sector, for example, telephone and electricityservice. It may be too soon to judge the overall ef-fect deregulation has had on the provision of publicservices. At least some of the challenge of deregu-lation is the result of the tumultuous transition froma regulated environment. Utilitycomputing may ben-

It is likely that the reliance

of users on computing services

will soon rival, if not surpass,their reliance on

public utilities.


7/29/2019 Rappa_2004.pdf

10/11

efit from an opportunity to grow in a highly com-petitive marketplace. To the extent that technologyevolves in a directionthat permits competition in theprovision of services, such as with open network pro-tocols, the need for service exclusivity may be less-ened. However, it should be expected that de factotechnical standards and competitive advantagesamong enterprises might eventuallylead to the emer-gence of dominant (if not exclusive) utility comput-ing service providers.

It is interesting to speculate on what shapethe utility-computing business model might take in the fore-seeable future. The provision of computing servicespresents a matrix of opportunities that goes well be-yond any comparison to traditional public utilities

like electricity. Although it may be technically fea-sible to meter some kinds of computing services,there remains the question of which services to meterand how this can be done. At the level of computinginfrastructure, it is possible to envision the meteredusage ofCPU resources, for example. At the appli-cation layer, there is already a move away from apure license model toward subscription-based ser-vices. It is also conceivable that some kinds of ap-plications could be adapted to a metered usagemodel, or a combination of subscription and meter-ing. Lastly, it may be advantageous to meter com-puting services based on the completion rate of dis-

crete business processes, such as the number ofcustomer transactions.

Fromthe customers point of view, the business logicof metering usage can be compelling: one pays onlyfor what one uses. This is something that IT man-agers who are faced with an escalating cost-of-own-ership can appreciate. But theinitial move away fromthe ownership model to the utility computing modelwill be hampered unless there exist clear operationalmeasures of the underlying demand function and,therefore, a way to determine the cost to an orga-nization when the meter is turned on. Depending

on the kind of metering employed, end users mayhave to adjust their computing habits to cultivate anethic of resource conservation and be mindful of theusage costs they incur. It is likely that commoditizedservices will be more easily adapted to a metered us-age model.

In comparison, the subscription model provides userswith more flexibility, and provides managers with amore accountable, if intermediate, approach on thepath toward utility computing. A subscription-basedservice is more amenable to high value-added pro-prietary services that require more elaborate servicelevel agreements. In a situation where there remainsuncertainty about the upside demand for services orwhere there are weak controls on usage levels, weare more likely to see adoption of a subscription ap-proach. Table 4 provides a summary comparing thefactors that may favor deployment of a subscriptionmodel versusthose that favor metered usage of com-puting services.

Conclusion

A vision of the future of computing services basedon the utility-computing business model has alreadybegun to take shape. Application service providers,managed services, and hosting are an increasinglycommon part of the computing landscape. Users

have come to depend on computers and have highexpectations of their reliability; they look toward aday when the use of computers matches the ease ofother everyday appliances. Furthermore, the provi-sion of computing services is increasingly driven byeconomies of scale and the effective utilization ofresources.

The kind of utility-computing business model thatwill find favor with the customerremainsto be seen.Already large enterprise customers are taking thefirst steps toward a model based on multiyear sub-scriptioncontracts. However the metered useof com-

Table 4 Factors favoring subscription versus metered utility model for computing services

Metered Model Subscription Model

Usage measures are easy to define, monitor, and verify Usage measures are difficult to define, monitor, or verifyStrong managerial controls on usage patterns Weak managerial controls on usage patternsCommoditized, low value-added services Proprietary, high value-added servicesFavored by cost-conscious users with an ethic to conserve

resourcesFavored by users who are less conscious of resource costs and

the need to conserveEasy to forecast resource usage patterns Hard to forecast resource usage patterns


7/29/2019 Rappa_2004.pdf

11/11

puting services is a significant leap from the current

model of purchasing or leasing computer hardware,accompanied by software licensing.

Clearly, there are trade-offs involved in migrating tonew business models. Customers must be convincedof the tangible benefits in making a change. Whatis ceded in terms of ownership and control must bemore than modestly offset by advantages in procur-ing computing services, such as flexibility, speed ofdeployment, and cost savings.

Acknowledgments

The author is grateful to Meeta Yadav for her re-search assistance, John Killela for his encourage-ment, and thehelpful commentsof three anonymousreviewers.

**Trademark or registered trademark of priceline.com Incorpo-rated, Google Inc., Ultramercial, Inc., or Overture Services Inc.

Cited references

1. A Nation Online, U.S. Department of Commerce (February2002).

2. Surveying the Digital Future, The UCLA Internet Report,UCLA Center for Communication Policy (January 2003).

3. J. B. Horrigan and L. Raine, Counting on the Internet, The

Pew Internet and American Life Project (December 2002).4. V. Berstis, Fundamentals of Grid Computing, IBM Redbook

(October 2002).5. M. Milenkovic, S. H. Robinson, R. C. Knauerhase, D. Bar-

kai, S. Garg, V. Tewari, T. A. Anderson, and M. Bowman,Toward Internet DistributedComputing, Computer, 36,No.5, 38 46 (May 2003).

6. I. Foster, C. Kesselman, J. M. Nick, and S. Tueckel., ThePhysiology of theGrid: An Open Grid Services Architecturefor Distributed Systems Integration, http://www.globus.org/research/papers/ogsa.pdf (June 22, 2002).

7. M. Rappa, Business Models on the Web,Managingthe Dig-ital Enterprise, http://digitalenterprise.org (May 2003).

8. M. Vitale, P. Weill, and J. Ross, From Place to Space: Mi-grating to Profitable Electronic Commerce Business Mod-els, Center for Information Systems Research Working Pa-

per No. 324, Massachusetts Institute of Technology (June2002).9. A. Osterwalder and Y. Pigneur, An e-Business Model On-

tology for Modeling e-Business,Proceedings of the 15th BledElectronic Commerce Conference, 112, Bled, Slovenia (June2002).

10. P. Timmers,BusinessModelsfor Electronic Markets,Elec-tronic Markets, 8, No. 2, 3 8 (1998).

11. P. Bambury, A Taxonomy of Internet Commerce,FirstMon-day, Issue 3.10 (1998).

12. FAQs: IBM e-Business On Demand, IBM Global Services(March 2002), http://www-1.ibm.com/services/strategy/files2/FAQ.ondemand.pdf

13. R. Jernigan, Interview: Irving Wladawsky-Berger ExplainsGrid Computing, IBM Linux Portal, http://www-1.ibm.com/linux/news31.shtml (March 2002).

14. e-Business On Demand; The Next Wave of IT Services, IBMGlobal Services, http://www-1.ibm.com/services/ondemand/files/nextwave.pdf (2002).

15. L. Siegele, At Your Service, The Economist, May 10,2003.16. T. Eilam, K. Appleby, J. Breh, G. Breiter, H. Daur, S. A.

Fakhouri, G. D. H. Hunt, T. Lu, S. D. Miller, L. B. Mum-mert, J. A. Pershing, and H. Wagner, Using a Utility Com-puting Framework to Develop Utility Systems,IBM Systems

Journal 43, No. 1, 97120 (2003, this issue)

Accepted for publication September 5, 2003.

Michael A. Rappa North Carolina State University, Box 7229,Raleigh, NC 27695 ([email protected]). Dr. Rappa is theAlan T. Dickson Distinguished University Professor of Technol-ogy Management and Director of the Open Courseware Labo-ratory at North CarolinaState University. Prior to joining NCSU,

for nine years he was a professor at the Massachusetts Instituteof Technology. In addition to overseeing the Open CoursewareLab, Dr. Rappa is co-director of North Carolina States E-Com-merce initiative. He is thecreator of theaward-winningWeb site,Managing the Digital Enterprise, digitalenterprise.org, one of theWebs most extensive and widely used sites devoted to under-standing the digital world and a definitive source for research onInternet business models.


Documents

Rappa_2004.pdf