The number of people connected to the Internet has been increasing very rapidly. According to some estimates, there are now more than 500 million users with Internet access, almost all of whom have access to email services [4]. This electronically connected Internet community has attracted direct marketers who view it as a medium that provides quick and inexpensive access to potential customers.
Direct marketers can easily get email lists of millions of potential consumers. These lists can be obtained from newsgroups or individual companies that compile this information. Solicitations to individuals on these lists via email have resulted in a problem commonly described as unsolicited commercial email (UCE), or spam. Fundamental to the problem of UCE is that senders find it more expensive to target their email messages to likely consumers than to simply send email to the entire list. Often, the information necessary to identify likely or interested consumers is either not available or expensive to procure. Since the cost of sending an additional email message is close to zero for the senders, everyone in the list gets bombarded with messages. The key operating business model is, "the more people you send email to, the more responses you will get."
The problem has become so severe that spamming is now the number-one source of customer complaints to ISPs. America Online (AOL), the largest ISP in the world, recently stated that over 30% of the estimated 30 million email messages it processes daily constitute UCE [6]. In a survey by World Research, over 65% of the respondents considered the amount of junk mail they received to be moderate to heavy, and over 68% found UCE to be not useful [1]. While individual users view email as a way to "reach out and touch someone," it can be abused to "reach out and harass everyone," which has incited public outrage toward UCE. As a result, the effectiveness of email advertising is severely impaired.
Here, we argue that effective legislation of UCE is necessary to realize the full potential of email advertising. When direct commercial solicitations become illegal, legislation will foster the development of new business initiatives, which, through market incentives, will ensure an optimal use of email channels for a variety of advertising purposes. We provide the framework for a business model that includes both consumers and sellers, and facilitates effective and mutually desirable email communication between them. We begin by investigating asymmetries in the cost-benefit structures that contribute to the problems with UCE and current legislative efforts to curb these problems.
Despite the risks associated with negative public reaction, the economic incentive for sending UCE is strong. This is especially the case for small- to medium-sized businesses that find it difficult to attract consumers to their Web sites. Although banner advertisements have become a popular option for those who shun the stigma associated with spamming, they can be very expensive and are viable only on Web sites that attract heavy user traffic. Even then, banner ads are a passive form of advertisement and their effectiveness is increasingly being questioned. Recent studies indicate most users are beginning to ignore banner advertisements [3]. By using email as an advertising mechanism, a consumer has to take some action, either read it or discard it. In effect, spamming messages do capture some amount of user attention because a user usually does not discard a message before reading or, at least, skimming it.
Economics associated with spamming reveal the key underlying problemnegative network externalitiescaused by the asymmetries in the cost-benefit structures. While the beneficiaries, the sellers, reap the majority of the benefits of spamming, the costs associated are borne mainly by ISPs and individual email users.
Costs borne by UCE senders are minimal. Software for extracting email addresses from the Web and list servers typically costs $100$600 and some is offered at no cost (see dir.yahoo.com/Business_and_Economy/Business_to_Business/Marketing_and_Advertising/Direct_Marketing/ Direct_Email/Software/). Using UCE makes perfect business sense to some sellers, even if only a tiny fraction of email recipients become customers. While the effects from the negative publicity prevent established firms from utilizing UCE, it is a small cost to pay for some small firms who can ill-afford traditional means of advertising to their potential customers. Email address-extraction software, which costs as little as $300, is able to collect as many as 10,000 to 15,000 addresses per hour (see dir.yahoo.com/Business_and_Economy/Business_to_Business/ Marketing_and_Advertising/Direct_Marketing/Direct_Email/Software/). Since the marginal cost of obtaining an address is close to zero, it can be used to extract as many email addresses as desired. For example, with a profit margin of $1 per sale, UCE becomes cost-effective at a response rate of as low as 0.5%, if a message is sent to 60,000 addresses. The larger the address list, the greater the potential profit.
The true underlying costs are borne by the ISPs and email users. The major elements of these costs are illustrated in Table 1. In 1997, Priori Networks spent $150,000 to combat spam. AOL has a half-dozen full-time staff dealing with the UCE problem [1]. In the long run, ISPs pass these costs on to individual subscribers, who, in addition, have to deal with productivity losses that result from reading messages, deleting messages, and lodging complaints against UCE senders.
A recent work provides a stylized model of the economic dynamics of UCE under the condition that no effective legislative or technical controls are imposed [5]. Fundamental to these dynamics are the search costs incurred by consumers. These costs include obtaining the knowledge about the existence of a product or service, contact information, information on pricing, features, and performance, and information regarding discounts and other special offers. Well-established firms typically impose lower search costs on consumers than smaller establishments impose on consumers. Thus, sellers of less well-known products and services have a significantly higher incentive than more established firms to utilize UCE in order to lower the search costs and thereby increase sales. Current evidence clearly indicates that a significant portion of junk email originates from such establishments. However, as the search costs for products and services from smaller firms begin to drop, the incentive for well-established firms to spam (to further lower the search costs and thus maintain their competitive edge) will begin to rise.
A third party could serve as the conduit between consumers and sellers by providing relevant and small volumes of commercial messages to consumers.
Companies such as Microsoft, Amazon.com, and Barnes and Noble, which rely heavily on traditional advertising channels, have been accused of occasional forays into sending UCE. As a result, the total number of UCE messages will sharply increase. This causes further compounding effects where each firm begins to send multiple messages to increase the probability their particular message will stand out and get noticed by the email users. This process will eventually result in the point of explosion, where user mailboxes are completely jammed by UCE. At this point, all email services come to a halt and email will cease to exist as medium of communication. A key result provided in [5] is that pure market mechanisms are incapable of solving the asymmetries in the cost-benefit structures of commercial solicitations via email.
There are two types of potential solutions to the problem of UCE: technical and regulatory. The common technical solutions include filtering, counterattack solutions, channels, payments, referral network, and fee restructuring [2]. Technical solutions are not always successful mainly because senders of UCE messages often use forged reply addresses to hide their identities. They place extra burden on ISPs and individual email users to devise and implement these solutions. Some technical solutions, such as payment, require widespread adoption of new protocols. Difficulties in curbing UCE messages through technical solutions are triggering efforts to control UCE through legislative measures.
The global nature of the Internet makes the regulation of UCE difficult but not impossible. To curb the growing problem of UCE, some efforts by lawmakers have been taken in the U.S. and more recently in other countries. For a summary of pending federal legislation against UCE, see the Web site of the Coalition Against Unsolicited Email at www.cauce.org/. In addition, Professor David Sorkin's Law of Cyberspace Web site, at www.spamlaws.com, provides information on both federal and state laws in the U.S. and anti-spam laws in other countries. None of the federal bills have yet passed the Congress, but there is relatively more success at the state level. Most of state bills require UCE senders to identify themselves, to provide a valid reply address, and to honor opt-out requests from email users who wish to have their names removed from the mailing list. For example, Nevada Senate Bill 13, enacted in 1997, and the California Assembly Bill 1676, enacted in 1998, require UCE messages to identify the sender and include opt-out instructions. The state of Washington enacted Washington House Bill 2752 in March of 1998, which prohibits false headers and misleading subject lines in UCE messages.
These regulations have come under criticism for actually legitimizing spamming and removing the stigma associated with it. Some predict that, as a result of regulation, the volume of UCE messages is likely to increase as spammers are essentially given at least one free shot at every address. Each individual user would receive a large number of messages and would be burdened with opting out of each unwanted list. In addition, ISPs would be burdened with maintaining filter software and processing the increased volume of UCE messages. Marketing effectiveness would continue to be very poor because the messages are not targeted.
These criticisms have galvanized lawmakers to propose bills that would require all unsolicited email messages to be conspicuously and uniformly labeled with subject headlines like "ADV," "Unsolicited," or some other clear nomenclature. The rationale is to enable easy filtering of the UCE messages. California Assembly Bill 1676 now includes a clause that requires certain email advertisements to contain "ADV:" or "ADV:ADLT" at the beginning of the subject line.
Some lawmakers attempted to take an even more radical approach: a complete ban of all UCE messages. The only legislative success for this approach thus far is the clause in the California Assembly Bill 1629, which bans all UCE to ISPs with explicit anti-UCE policies. Some other states are trying to enact similar bills. Such efforts have received widespread support from the Internet community and the ISP association alike.
If all UCE messages can be filtered out easily or are made illegal, consumers who wish to obtain commercial information from a potential seller need to initiate the communication. Many reputable sellers have already voluntarily elected to use so-called opt-in methods to contact potential consumers: send commercial email messages to a consumer only when the consumer has solicited such mailing by signing up with the seller's mailing list.
While an opt-in mechanism enables utilization of email as an advertising tool when UCE messages are illegal or can be easily filtered, it is not without drawbacks. With opt-in, the onus shifts to individuals to seek information on products and services, and to contact establishments that may satisfy their consumption needs. However, individuals would have to provide similar personal information to a number of competing sellers in order to intelligently comparison-shop and select the optimal product or service. Thus, effective regulation against UCE would arguably create a new problem: high search costs and repetitive dissemination of personal information to a variety of potentially questionable organizations.
A new form of business initiative is emerging on the Internet to enable email advertising in the face of increasing resistance to UCE and its ensuing problems. This business concept derives its roots from the notion of intermediation, and is founded on employing incentive-based approaches to match interested buyers and sellers. We term the concept admediation, as this business entity primarily provides advertising-related information management services.
The survival of such an admediary is based on two premises: First, unsolicited messages are not necessarily unwanted. Rather, it is the nature, relevance (or more precisely, irrelevance), and the sheer volume of these messages that email users truly object to. Second, from a seller's perspective, harassing users with unwanted messages and generating negative publicity is not an intended marketing strategy, but an unfortunate side effect of attempts to market their services via the new technologies. A third party could serve as the conduit between consumers and sellers by providing relevant and small volumes of commercial messages to consumers when such messages are desired by consumers, and by providing a list of a willing and well-targeted consumer base for the sellers.
Figure 1 provides the framework for admediation. The fundamental role played by the admediary is as the 'go-between' for buyers and sellers. Consumers sign up with the admediary, indicating their preferences for products and services. In addition, consumers specify the nature, volume, and frequency of email messages they wish to receive. Sellers subscribe to the admediary in order to solicit potential customers. When requested by the seller, the admediary composes the email list for a seller's advertisement from consulting the customer database, and emails the solicitations. The responses are forwarded to the seller, and archived to generate detailed analysis reports on the success of the email advertising campaign. The business model of such intermediaries involves providing value-added services to consumers and sellers. An overview of the value-added services and the payment transfers is provided in Figure 2.
The admediary serves as the value provider for both the consumers and the sellers. From a consumer's perspective, signing up with an admediary represents revealing personal information and soliciting product information from all relevant producers participating with this admediary. The prospect of reduced search costs for product information entices consumers to sign up. Further incentives, such as direct payment for signing up, coupons, and price discounts are also provided, along with guarantees of safeguarding individual personal information. The sellers are provided access to a targeted list of consumers composed of those who have signed up. In return, sellers have to compensate the admediary (typically on a per-email basis), and may also offer special discounts to the individual subscribers. The significant appeal of this business model of admediation is that its cost and price structure alleviates the negative externality effect, which is the driving force of UCE messages.
Table 2, compiled from MarketersMarket.com (see market.emailresults.com), provides some examples of online companies operating under the umbrella of the admediation concept. To be able to survive as a business entity, an admediary provides services to both consumers and sellers, detailed in Table 3. This mode of advertising, born only a few years ago, is beginning to gain momentum as frustrations and calls for a complete ban of UCE grow. Admediaries such as WebStakes and Business Link already claim a subscription base of well over a million. This has attracted companies such as Disney, Kodak, Dell Computers, and others that rely heavily on traditional advertising routes. Properly implemented, this business model of business-to-buyer interactions can prove to be the most potent in the marketing arsenal.
Currently, admediaries are popular with legitimate and established firms that find the low cost (averaging about $0.1/email) attractive. These firms have realized the increasing public outrage toward UCE messages has a negative impact on the long-term profitability of their products in the marketplace. However, there remain business entities that continue to find UCE attractive. To these companies, the long-term image of their product factors less into the bottom line, and even the low cost of advertising via an admediary is astronomical compared to the close to zero marginal cost of UCE. Examples include companies peddling pornographic products, and those that solicit money for make-believe business opportunities, investment opportunities, credit card offers, and dubious medical products.
Effective legislation of UCE is critical for the long-term success of admediation. Without legislation, admediaries have to compete with UCE messages for user attention. As the nature of these commercials turns increasingly offensive and deceptive, email users become suspicious of any solicitations via email. As these solicitations become synonymous with scam, it severely inhibits the appeal of admediation. More ominously, filtering software may not be able to distinguish UCE from legitimate messages from admediaries, since UCE messages can be disguised as originating from a popular admediary. Negative externality effect would continue to paint the picture, with ISPs and email users bearing most costs of UCE and with marketing effectiveness of admediaries reduced. Many admediaries, realizing the potential damaging role of UCE, have joined the fight against UCE messages.
If consumers gain legislative protection against UCE, admediation will emerge as the only viable avenue for the dissemination of commercial email messages. Email can then be harvested as a powerful advertising tool where both sellers and email users would be better off because the mailings would be targeted, effective, desirable, and yet manageable.
As admediaries collect, group, and organize email messages sent to their consumer lists in a conscientious way (BulletMail, for example, promises no more than 55 lines of text per email), congestion would be eased and the load on ISPs' email servers reduced, which directly benefits ISPs. In addition, being gatekeepers to their subscribers, ISPs control channels that make business-to-buyer communication possible. As such, powerful incentives exist for them to participate in this value-added process. It is increasingly likely that business-to-buyer interactions would be mediated by admediaries and ISPs, with each providing important value-added services. Sellers and consumers, in turn, would be drawn increasingly to using admediaries as they both enjoy the benefits of targeted emailing.
1. Berry, D. Controlling unsolicited bulk e-mail. Sun World (Aug. 1997); www.sunworld.com/sunworldonline/swol-08-1997/swol-08-junkemail.html
2. Cranor, L.F. and LaMacchia, B.A. Spam! Commun. ACM 41, 8 (Aug. 1998), 7481.
3. Dietz, N. Survey: Banners losing effectiveness. Business Marketing (Sept. 1998).
4. Global Research; www.glreach.com/globstats/index.php3
5. Gopal, R.D., Walter, Z., and Tripathi, A.K. Economic analysis of unsolicited commercial email. Tech. Rep., Department of Operations and Information Management, University of Connecticut, 1998; www.sba.uconn.edu/users/ram/research/UCEPaper.pdf
This work was partially supported by the Treibick Electronic Commerce Initiative, Department of Operations and Information Management, School of Business, University of Connecticut.
©2001 ACM 0002-0782/01/1200 $5.00
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2001 ACM, Inc.
No entries found