Chapter Four 1

What can companies do to keep customers from leaving theircompany for another? The following nine “exit barriers,”based on Diorio (2002), help bolster e-loyalty.

1. Customers’ learning curve. The company’s Web site shouldbe easier to use than any other company’s. To achieve thisgoal, use familiar and easy-to-negotiate menus, similar toeBay’s online help.

2. Process integration. Become a part of the way your cus-tomers work. Examples include shared equipment or con-tractual commitments such as volume purchases online.

3. Personalization. Use information to match offerings withcustomers’ tastes. Ways to do this include customer data-bases, tracking of customer behavior, and targetedmarketing.

4. Mass customization. Expand choice so customers willnot look elsewhere. The desire to change vendors is min-imized if you offer a large variety of mass-customizedproducts/services and a rapid product evolution. The Webcan be a great help in mass customization (see Chapters 1and 2).

5. Risk reduction and trust. Make yourself the “safer”choice. Generate trust through service history and superbcustomer support (like Amazon.com). Using TRUSTe andBBBOnLine can be helpful.

6. Loyalty programs. Give economic incentives for cus-tomers’ loyalty. Programs include “points” and frequency

plans (e.g., frequent-flyer miles), volume discounts, spe-cial promotions, and referral awards. These can be man-aged online with auto tracking via the Web.

7. Brand affinity. Give emotional incentives for customersto stay with your company. This can be accomplished byaggregating products and Web page content to particularinterest or lifestyle groups. Forums and chat rooms on theWeb site also are incentives to stay with a product orcompany.

8. Customer collaboration. Talk with and listen to cus-tomers more. Ongoing value-added dialog, including chatrooms and dynamic e-mail messaging, keeps customershappy.

9. Be the standard. Find ways to become the only, or thebest, choice. By dictating industry standards, you cankeep customers from going elsewhere. One prime exampleof this is Microsoft.

However, according to Coyles and Gokey (2002), cus-tomer retention is not enough in the long run. Customersmay change their buying patterns, and these changes canundo the power of loyalty. Therefore, to keep up with chang-ing markets and consumer tastes, companies must under-stand their customers—what they want now and what theymight want in the near future.

Sources: Diorio (2002) and Coyles and Gokey (2002).

Online File W4.1 Customer Exit Barriers

2 Part 2

Spyware Banners Advertisers use spyware to spy on your movements on theInternet and send you e-mail ads or pop-ups based on whatthey learn about you. Authors of shareware and freewareoften use spyware in order to make money on their products,which are often offered free to users. Shareware authors mayinsert banner ads of advertisers into their products inexchange for a portion of the revenues generated by the ban-ners. This way, end users do not have to pay for the software,and the developers get paid for their efforts. End users whofind the banners to be annoying can usually obtain a banner-free copy of the shareware by paying a regular licensing feeto the developer. Increasingly, however, spyware is beingembedded in purchased software.

Spyware Categories Spyware threats come in different flavors. One variety of spy-ware is malware. This type of spyware can modify the systemsettings on the user’s computer (see Chapter 11) and performother undesirable tasks. A hijacker is a piece of spyware thatredirects the user’s browser to a particular Web site. A dialeris spyware that dials a service, most likely porn sites, with-out the user’s knowledge. In many cases, the user will beheld responsible for the bill! A trojan horse is spyware that isattached to a program and performs undesirable tasks on theuser’s system. Collectware is spyware that collects informa-tion about the user. This information is then provided to athird party without the user’s knowledge.

Simply the Best (simplythebest.net/info/spyware.html )identifies the following types of spyware:

◗ Adware networks. Ad serving networks such as DoubleClick,Web3000, Radiate, SaveNow, GAIN pay publishers of games,utilities and music/video players per download to includetheir ad serving programs.

◗ Stalking horses. A number of programs that collect infor-mation on the user, facilitate adware networks to functionon desktops. These programs are bundled in many popularprograms and often are presented in installation disclosurescreens as desirable add-ons to Trojan horse hosts. Ex-amples include TopText, Cydoor, OnFlow, Medialoads, Delfin,WebHancer, New.net.

◗ Trojan horses. These are popular Internet downloads thatusually come with the ad serving network basic software andat least one stalking horse. They are contained in KaZaa,Grokster, Morpheus, Limewire, AudioGalaxy, iMesh, DivX.

◗ Backdoor Santas. These are stand-alone programs thatincorporate similar approaches as above. However, theyhave no links to ad serving networks although their purposeit to collect information from users. They are located inAlexa, Hotbar, Comet Cursor, eWallet, CuteFTP, BonziBuddy.

◗ Cookies. Netscape Navigator and Internet Explorer stillsend out cookies even after cookies have been disabled inthe browser settings. Any and all cookie files must bemanually deleted on your system in order to eliminatebeing tracked by third-party ad networks or spyware oradware providers.

For more information about spyware, see spywareguide.comand Chapters 11 and 17.

Online File W4.2 Spyware

Chapter Four 3

ONLINE FILE W4.3

GOOGLE: A GIANT ADVERTISEMENT COMPANYMany consider Google to be the world’s best search engineprovider. But now Google is using its technology to generatebillions of dollars in advertising revenue (see Chapter 14 forGoogle’s strategy). How does it do it?

At the heart of Google is AdWords, a self-service adserver that uses relevance-ranking algorithms similar to theones that make the search engine so effective. Advertiserstell Google how much they want to spend and then they“buy” pertinent keywords. When Web surfers type in a termthat matches the advertiser’s keyword, the advertiser is listedin a banner near the search results with the heading“Sponsored Links.” Each time a user clicks the advertiser’sbanner ad, Google subtracts the cost-per-click from theadvertiser’s prepaid account. When the account’s daily adbudget is depleted, Google stops displaying the ad.

The system is easy to use and remarkably effective. Theclick-through rate is about 15 percent, which is more than10 times the click rate of the average banner ad. Accordingto industry experts, many Google advertisers experienced a20 to 25 percent increase in online sales.

AdWords has worked so well that in 2003 Google beganoffering the system to other sites. The new service, AdSense,places ads on popular Web sites, from ABC.com and The NewYork Times on the Web to popular blogs.

Pinpoint targeting has always been the goal for adver-tising, but has been unattainable for a couple of reasons.

First, content providers rarely have been able to deliver tomarketers enough well-described and narrowly focused con-sumers. Second, even when marketers get the perfect audi-ence, they rarely have targeted ads; it costs too much tomake multiple variations of an ad for microaudiences.AdSense strips away these barriers.

AdSense uses the relevance-scoring algorithms to findthe best match of an ad to a surfer by analyzing the contentof Web pages of the popular sites and key words of theadvertisers. Once a person clicks the banner ad, whichincludes a short sentence describing the advertiser’s product(service), he or she is transferred to the advertiser’s site. Allthis is done automatically.

Because people reading news sites are not likely to bein purchase mode, AdSense click-through rates are not ashigh as those generated by AdWords, but still far better thanregular banner clickthroughs. For Web site owners, theinvestment is minimal: A few minutes to sign up and a littlespace at the top, side, or bottom of the Web page. More than150,000 sites have joined this affiliate program. The moneygenerated from the ads is shared between Google and theaffiliate sites.

Source: adwords.google.com/select/ (accessed December 2004). Usedwith permission of Google Inc.

4 Part 2

Advertisers use dozens of innovative techniques to lure con-sumers into viewing online ads. The following list is only asample of the many interesting ideas companies have used toattract Web surfers. For more on promotions, visitpromomagazine.com.

◗ Retailers can provide online shoppers with special offers whilethey are purchasing or “checking out.” If a shopper’s profileor shopping history is known, the ads can be targeted.

◗ Netstakes runs sweepstakes that require no skill. Users reg-ister only once and win prizes at random (webstakes.com).The sponsors pay Netstakes to send them traffic. Netstakesalso runs online ads, both on the Web and through e-maillists that people subscribe to.

◗ Cydoor (cydoor.com) places ads, news, and other items onsoftware applications. Consumers who download the soft-ware receive a reward each time they use the software (andpresumably read the ads).

◗ CBS Marketwatch (cbsmarketwatch.com) uses animated beerbottles and interactive charts to attract viewers to its freefinancial site.

◗ Sometimes a catchy name draws Web surfers. For example,an old-economy seller of hard-to-find light bulbs changedits name to topbulb.com and created an online catalog,called the Bulbguy, through which it sells light bulbs onlineat a discount. The Web site is advertised both online andoff-line, and business is booming!

◗ Promotionbase.com is a magazine-format site dedicated toWeb site promotions. Users can find rich resources and pro-motions on how to increase Web traffic.

◗ To promote its sport utility vehicle, the 4Runner, Toyotawanted to reach as many Internet users as possible. Thecompany displayed Toyota banners on the search engineAltaVista (altavista.com). Whenever someone used AltaVistato search for anything related to automotives, they would

see the Toyota banner. Also, Kelly Blue Book’s new-car pric-ing catalog (kbb.com) had links to Toyota’s car. In the first2 months of the campaign, over 10,000 potential car buyersclicked on the banner ads looking for more detailed infor-mation about the Toyota 4Runner.

◗ Web surfers can play games, win prizes, and see “e-tractions”at uproar.com. Special promotion campaigns are also featured.

◗ To promote its job-recruiting visits on U.S. college cam-puses, IBM created over 75,000 college-specific bannerssuch as, “There is life after Boston College: click to seewhy.” The students clicked on the banners at a very highrate (5 to 30 percent). As a result of this success rate, IBMrestructured its traditional media plans using the “ClubCyberblue” scheme.

◗ Each year, almost 500,000 brides-to-be use theknot.com toplan their wedding. A “Knot Box” with insert folders is sentto users by regular mail. Each insert is linked to a corre-sponding page at theknot.com. Advertisers underwrite themail campaign. The Web site provides brides with informa-tion and help in planning the wedding and selecting ven-dors. Orders can be placed by phone or online (not all prod-ucts can be ordered online). Weddings411.com is a similarservice, operating primarily online.

Bargain hunters can find lots of bargains on theInternet. Special sales, auctions, and downloading ofcoupons are frequently combined with ads. Of special inter-est are sites such as coolsavings.com, hotcoupons.com,supercoups.com, clickrewards.com, and mypoints.com. A popu-lar lottery site is world-widelotto.com. In addition to lotteriesand coupons, free samples are of interest to many consumers,and “try-before-you buy” gives consumers confidence in whatthey are buying. Freesamples.com began to offer free samplesin June 2000.

Online File W4.4 How to Attract Web Surfers

ONLINE FILE W4.5

ECONOMICS OF ADVERTISEMENTOne of the major issues in advertising is the cost-benefit (the relationship of its cost to its benefitsit provides) it provides to the advertisers. The cost depends mostly on the method of payment.

PRICING OF ADVERTISINGJustifying the cost of Internet advertising is more difficult than doing so for conventionaladvertising for two reasons: (1) the difficulty in measuring the effectiveness of online adver-tising and (2) disagreements on pricing methods. Several methods are available for measur-ing advertising effectiveness, conducting cost-benefit analyses, and pricing ads.

Pricing Based on Ad Views, Using CPMTraditional ad pricing is based on exposure or circulation. So far, this model has been thestandard advertising rate-pricing tool for Web sites as well, usually using ad views to measurecirculation. Because advertisers pay an agreed-upon multiple of the number of “guaranteed”ad views using a CPM formula, it is very important that ad views are measured accurately inthe context of the advertising business model. Generally, CPMs seem to average on the orderof $40 (per 1,000 ad viewers), resulting in a cost of $0.04 per impression viewed (per ad view).

Some companies, such as USA Today Online, charge their clients according to the num-ber of hits (about $0.035 per hit in 2005). However, hits are not an accurate measure of visi-tation, because one ad view may have several hits.

Pricing Based on Click-ThroughAd pricing based on click-through is an attempt to develop a more accountable way of charg-ing for Web advertising. In this model, the payment for a banner ad is based on the numberof times visitors actually click on the banner. Payment based on click-through guarantees notonly that the visitor was exposed to the banner ad, but also that the visitor was sufficientlyinterested to click on the banner and view the target ad (Hoffman and Novak 2000).

However, a relatively small proportion of those exposed to a banner ad (about 1 to 3 per-cent of viewers) actually click on the banner. Therefore, space providers usually object to thismethod, claiming that simply viewing a banner ad may lead to a purchase later or to an off-line purchase, much as newspaper or TV ads do. Advertisers, on the other hand, do not liketo pay for ad views; they prefer the click-through method, which they feel is more accurate.Only large advertisers such as Procter & Gamble can pressure space sellers to accept click-through payment methods, or even better, interactivity.

Payment Based on InteractivityAlthough payment based on click-through guarantees exposure to target ads, it does notguarantee that the visitor liked the ad or even spent a substantial time viewing it. Theinteractivity model (Hoffman and Novak 1996) suggests basing ad pricing on how the visitorinteracts with the target ad. Such an interactivity measure could be based on the duration oftime spent viewing the ad, the number of pages of the target ad accessed, the number of addi-tional clicks generated, or the number of repeat visits to the target ad. Obviously, this methodis more complex to administer than the previous methods.

Payment Based on Actual Purchase: Affiliate Programs Many advertisers prefer to pay for ads only if an actual purchase has been made. Sucharrangements usually take place through affiliate programs. Merchants ask partners, known asaffiliates, to place the merchant’s logo on the affiliate’s Web site. The merchants promise topay the affiliate a commission of 5 to 15 percent whenever a customer clicks on the merchants’logo (banner) on the affiliate’s Web site and eventually moves to the merchant site and makesa purchase. For example, if a customer saw Amazon.com’s banner at AOL’s Web site, clickedit, moved to amazon.com, and completed the purchase, AOL would receive a referral fee of,say, 5 percent of the purchase price of the book. This method works only at sites where actual

Chapter Four 5

purchases can be made (e.g., see cdnow.com at Amazon.com and cattoys.com). At theRitchey Design or Coca-Cola (cocacola.com) sites, users only get information and brandawareness; thus this method would be inappropriate for these types of merchants.

Which Is the Most Appropriate Pricing Method?In addition to the four major methods just described, still other methods can be used to payfor ads. For example, some space providers charge a fixed monthly fee to host a banner,regardless of the traffic. Others use a hybrid approach, a combination of some of the previousmethods. The question is, which is the most appropriate method?

Web space providers, such as Yahoo!, push for CPM. They argue that the problem withactivity-based measures, such as click-through or interactivity, is that the Web space providercannot be held responsible for activity related to an advertisement. (If the customer sees anad, but it is a poor ad that does not inspire further activity, it is not the fault of the Web spaceprovider.) They also argue that traditional media, such as newspapers or television, charge forads whether or not they lead to sales. So why should the interactive condition be applied onthe Web?

Advertisers and their agencies, on the other hand, argue that because the Web mediumallows for accountability, models can and should be developed that measure actual consumeractivities. The answer to the question of the most appropriate method has not been settled.

ADVERTISEMENT AS A REVENUE MODELMany of the dot-com failures in 2000 to 2002 were caused by a revenue model that con-tained advertising income as the major or only revenue source (e.g., see the story of Go.comin Chapter 3). Many small portals failed, but three large ones are dominating the field: AOL,Yahoo!, and MSN. However, even these heavy-traffic sites reported only small revenuegrowth in 2001 and 2002. There are simply too many Web sites competing for advertisingmoney. For these reasons, almost all portals are adding other sources of revenue.

However, if careful, a small site can survive by concentrating on a niche area. For example,Playfootball.com is doing well. It pulls millions of dollars in advertising and sponsorship byconcentrating on NFL fans. The site provides comprehensive and interactive content,attracting millions of visitors.

MEASURING ADVERTISING EFFECTIVENESSDetermining the cost of advertising is easier than assessing its benefits. Most of the benefits ofadvertising are intangible. However, more and more companies are requiring that some mea-sure be made of the effectiveness of advertising. One financial measure is the return on invest-ment. Another way to measure advertising effectiveness is to measure and analyze Web traffic.

Return on InvestmentAn increasing number of companies are requiring that the rate of return on investment (ROI)be used to measure the benefits received from their online advertising campaigns. ROI can becalculated in several ways. One popular formula for ROI is the net benefit (the total benefitsminus the total cost) divided by the required investment. Obviously, the difficult part is to puta dollar amount on the total benefits. Nevertheless, Forrester Research (forrester.com) hasdeveloped an interactive ROI model (see Scheirer et al. 2001) for the “word-of-mouth” adapproach. Also, ad management companies such as Worldata.com generate reports that canhelp an organization to calculate the ROI of its e-mail ads (see the demo at worldata.com).Many other vendors offer ROI support services. For example, Advertising.com offers opti-mization services that analyze campaign data in real time, helping the advertising team makenecessary adjustments to the campaign.

The cost of advertising is perhaps the key component of ROI. (It certainly is the onethat companies and advertisers have the most control over.) One of the ways to improve ROIis to lower advertising costs. Various vendors offer services to help do that. To reduceexpenses, a company can negotiate ad purchasing at valueclick.com. In addition, companiescan use reverse auctions to solicit bids from space providers.

6 Part 2

Measuring, Auditing, and Analyzing Web TrafficBefore a company decides to advertise on someone’s Web site, it should verify the number ofad views, hits, clicks-through, or other data reported by the site’s owner. A site audit validatesthe data claimed by the site, assuring advertisers that they are getting their money’s worth.An impartial, external analysis is crucial to advertisers to verify the accuracy of any countsclaimed by sites.

The Audit Bureau of Circulation (ABC) (see abc.org.uk) is a nonprofit association cre-ated by advertisers, advertising agencies, and publishers, who came together to establishadvertising standards and rules. The ABC verifies circulation reports by auditing circulationfigures of newspapers, TV, radio, and now the Internet. It provides credible and objectiveinformation to the buyers and sellers of advertising. Several other independent third-partyInternet auditing companies also are in operation, such as BPA International (bpai.com), theInternet Advertisement Bureau (iab.net), and I/PRO (ipro.com).

Related to auditing is the rating of sites. This is done by companies such as Accipiter,I/PRO, Netcount, Interse, Hotstats, and CNET. Rating is done by looking at multiple crite-ria such as content, attractiveness, ease of navigation, and privacy protection. Sites withhigher ratings can command higher prices for advertising placed on their sites. In addition tooutside independent monitoring, several vendors sell software that allows Webmasters toself-monitor traffic on their own Web sites. Examples are worldata.com, webtrends.com, andnetratings.com. Additionally, Webmasters can measure who is coming to a site and fromwhere (e.g., see leadspinner.com). Using such software, companies can assess if placing adsreally increases traffic to their sites.

Audience Tracking. Advertisers are interested in gathering as much information as pos-sible about the acceptance of ads, both online and off-line. Arbitron Corporation has devel-oped a portable, wearable meter (beeperlike) device. The device logs programming (includingTV, radio, and streaming media Internet broadcasts) seen or heard anytime, anywhere, bywhomever is wearing it. A motion detector on the device verifies that a person is wearing it(and does not just set it in front of the TV or computer screen). Each night the device isplaced in a docking station in the wearer’s home from which it transmits the day’s data toArbitron (arbitron.com).

Chapter Four 7