Personalized mobile advertising: Its key attributes, trends, and social impact

Peng-Ting Chen a,⁎, Hsin-Pei Hsieh b

a Department of Business Administration, I-Shou University, No.1, Sec. 1, Syuecheng Rd., Dashu Township, Kaohsiung County 840 Taiwanb Postgraduate Programs in Management, I-Shou University, No.1, Sec. 1, Syuecheng Rd., Dashu Township, Kaohsiung County 840 Taiwan

a r t i c l e i n f o a b s t r a c t

Article history:Received 4 January 2011Received in revised form 1 August 2011Accepted 13 August 2011Available online 15 September 2011

Advertising media are a means of communication that creates different marketing and com-munication results among consumers. Over the years, newspaper, magazine, TV, and radiohave provided a one-way media where information is broadcast and communicated. Due tothe widespread application of the Internet, advertising has entered into an interactive commu-nications mode. In the advent of 3G broadband mobile communication systems and smart-phone devices, consumers' preferences can be pre-identified and advertising messages cantherefore be delivered to consumers in a multimedia format at the right time and at theright place with the right message. In light of this new advertisement possibility, designingpersonalized mobile advertising to meet consumers' needs becomes an important issue. Thisresearch uses the fuzzy Delphi method to identify the key personalized attributes in a person-alized mobile advertising message for different products. Results of the study identify six im-portant design attributes for personalized advertisements: price, preference, promotion,interest, brand, and type of mobile device. As personalized mobile advertising becomes moreintegrated in people's daily activities, its pros and cons and social impact are also discussed.The research result can serve as a guideline for the key parties in mobile marketing industryto facilitate the development of the industry and ensure that advertising resources are proper-ly used.

© 2011 Elsevier Inc. All rights reserved.

Keywords:PersonalizedAdvertisingMobile advertisingFuzzy Delphi methodPersonalized attribute

1. Introduction

Advertising media plays an important role for advertisers in product sales and marketing. Different media will create differentmarketing and communication results for consumers. In general, there are five major advertising media categories: magazine/newspaper, TV, radio broadcast, Internet, and mobile communications.

The traditional forms of advertising media are passive in nature. In other words, advertising messages are displayed in text,pictures, or graphics in media forms such as magazines and newspapers. In contrast, radio broadcasting and TV are dynamicmedia that can deliver voice and video advertising messages to consumers. However, magazines/newspapers, radio broadcasting,and TV advertising media cannot deliver personalized advertising for different target markets. Given the high penetration rate ofthe Internet and its interactive nature, instead of only receiving advertising messages, consumers can now proactively search fornecessary advertising information. Furthermore, personalized advertising can also be accomplished through interactive Internetcommunications.

In 2Gmobile communications, the typical data transfer rate achievable on aGlobal System forMobile Communications (GSM) net-work is 9.6 kilobits per second, which leads to user frustration because of long download times. The low transfer rate also severelylimits the richness of information and complexity of the wireless data services and applications that can be offered [1]. With thewide deployment of mobile communications and in the advent of 3G mobile communication systems, increased bandwidth enablesmobile advertisers to deliver multimedia advertising information to consumers. In addition, the enhanced communication speed and

Technological Forecasting & Social Change 79 (2012) 543–557

⁎ Corresponding author. Tel.: +886 7 6577711x5916; fax: +886 7 6578931.E-mail addresses: ptchen@isu.edu.tw (P.-T. Chen), m9782030@isu.edu.tw (H.-P. Hsieh).

0040-1625/$ – see front matter © 2011 Elsevier Inc. All rights reserved.doi:10.1016/j.techfore.2011.08.011

Contents lists available at SciVerse ScienceDirect

Technological Forecasting & Social Change

bandwidth allow mobile consumers to surf the Internet for advertising information. Consequently, the possibility of delivering thecorrect advertising message to the right people at the right time has become possible [2]. With the availability of mobile broadbandcommunication, mobile carriers and advertisers are all aggressively preparing for multimedia-related applications and services [3,4].

The number of handset subscribers worldwide exceeded 5 billion by the end of 2010 [5]. In Taiwan, there were 17.84 millionhandset subscribers by the end of the fourth quarter of 2010, resulting in an overall 120.2% mobile communication penetrationrate [6]. Among these handset subscribers, 18.2 million subscribers subscribed to Internet access service. These data show thatmobile devices not only serve as personal communication devices but have also become important Internet access devices. There-fore, mobile handsets can potentially serve as a new advertising channel. Advertising messages are also displayed in a multimediaformat as a result of the advancement of mobile communication. Right now mobile advertising only represents around$2.7 billion of the global $500 billion advertising market based on Interpublic Group of Cos.' ad forecaster Magna. According toeMarketer, in the US alone, mobile advertising spending is expected to be more than doubled by 2014 to $2.6 billion, from around$1.1 billion in 2011 [7]. In the past, because they were limited by mobile bandwidth, mobile advertising messages were typicallyin the short message service (SMS) text format. Given that SMS is not designed for personal consumer needs, it usually createsannoyance among SMS receivers and also gives a negative impression on the advertiser. With the development of global positionsystem (GPS) location services, bandwidth availability, and smartphone handsets, advertisers now have the capability to providepersonalized advertising messages to consumers at the right time and place. In light of this newmobile capability, how to design apersonalized advertising message to meet consumers' needs has become a critical issue.

There is little research on the subject of designing a personalized mobile advertising message. Not only is there a lack of studyon the attributes of personalized mobile advertising, but there is also a lack of study on its application, trends, and social impact.Therefore, the goals of this research are to identify the important attributes in designing a mobile advertising message for prod-ucts (goods and services), explore the potential applications of these attributes, and understand their trends and social impact.The research results can serve as a guideline for the members in the mobile advertising industry to assist the industry in allocatingthat advertising resources.

To ensure the practicality of the research results, this analysis researched mobile advertising attributes from academic studiesand industry case studies. We also conducted surveys among different parties in the mobile advertising value chain to solidify theresults. The traditional Delphi research method inherently has the limitation of experts' opinions not easily converging, high sur-vey cost, and true expert opinion being eliminated by survey analyzers [8]. Therefore, this research used the fuzzy Delphi researchmethod to collect, analyze, and identify the important mobile advertising attributes in designing an advertising message. The ad-vantages of using the fuzzy Delphi research method include reducing the number of survey iterations, reducing survey cost, in-creasing survey return rate, effectively capturing and analyzing expert opinions on the questionnaires, and capturing the fuzzynature of expert estimation and predictions in answering the questionnaires [9–11].

This paper is organized as follows. Section 2 reviews the development history of advertising media. Section 3 summarizes ad-vertising literature review and discusses the trend in personalized advertising. Section 4 describes relationship among the playersin the mobile advertising industry value chain. Section 5 establishes the fuzzy Delphi model and the survey questionnaire. Sec-tion 6 shows the practicality and usefulness of this method and its empirical results, and discusses the issues and social impactof personalized mobile advertising. The last two sections conclude and offer some recommendations, respectively.

2. Overview of advertising media development

The AmericanMarketing Association [12] defines advertising as “the placement of announcements and persuasive messages intime or space purchased in any of the mass media by business firms, nonprofit organizations, government agencies, and individ-uals who seek to inform and/or persuade members of a particular target market or audience about their products, services, orga-nizations, or ideas.” Given the different unique media characteristics, each advertisement medium conveys different marketingmessages to consumers. There are five advertisement media categories: magazines/newspapers, TV, radio, Internet, and mobileadvertisement.

Newspapers and magazines are a traditional and primary form of advertising media; they also have a long history. Generally,advertisements in newspapers or magazines are of a pull type because the message is transferred by the free will of the audience.This creates a high involvement level with audiences. TV and radio are a push type of advertisement medium. The audience sits infront of a TV and views what appears on the screen. The audience has less involvement during the process. Nonetheless, news-papers, magazines, TV, and radio have been the primary advertisement media for advertisers to promote their products. However,in using the aforementioned advertisement media, if a consumer is interested in purchasing the advertised product, another com-mercial medium is required to complete a business transaction.

Compared with traditional TV commercials, online advertising works interactively, which has been found to be more efficientthan one-way advertising because interactivity improves the comprehension of the message the advertisement tries to convey[13,14]. With the help of individual and interactive connections, an advertising agency can easily personalize the contents ofthe advertisement. By coupling with traditional advertisement media, the Internet allows the sender and receiver to managethe advertisement information effectively.

With the emergence of mobile communications, mobile advertisement enables advertisers to deliver personalized advertisinginformation to consumers at the right time and place. Consumers can not only receive advertisement information through thepush-type delivery method but can also proactively retrieve advertisement information via the pull-type delivery method. Mobilecarriers typically have personal information on their subscribers; therefore, designing a personalized advertisement based on a

544 P.-T. Chen, H.-P. Hsieh / Technological Forecasting & Social Change 79 (2012) 543–557

subscriber's profile and preferences is deemed conceivable. This personalized advertisement can then be delivered to the personalhandset device to maximize the advertisement andmarketing effect. Table 1 summarizes the advertisement media characteristicsand target markets.

3. Literature review

3.1. Mobile advertising

The Mobile Marketing Association [15] defines mobile advertising as a form of advertising that transmits advertisement mes-sages to users via mobile phones, personal digital assistants (PDAs), or other wireless communication devices. According to theWireless Advertising Association [16], mobile advertising uses mobile networks to transmit advertisement messages throughwireless communication devices such as mobile phones or PDAs to achieve advertising effectiveness. Given the vigorous promo-tion of the wireless broadband industry in Japan, South Korea, and Taiwan, the Institute for Information Industry [17] expects thatthe mobile advertising market in Asia will reach US$7.7 billion by 2012 [18].

Mobile users continue to subscribe the 3G and 3.5G (HSPA) broadband services. According to Paul Budde Communication sta-tistics, there are around 650 million 3G/3.5G subscribers worldwide [19]. Therefore, with the development of communicationtechnology, coupled with the gradual increase in 3G mobile device penetration, mobile advertising will become the advertisingmedium with the most potential business opportunities.

Currently, the most common forms of mobile advertising include text messaging (i.e., SMS) and multimedia messaging (i.e.,multimedia messaging service). SMS text messages are in plain text format, which is readable by nearly all cell phones. Therefore,SMS is currently the most widely used form of advertising. However, each message can only accommodate up to 160 characters;thus, designing compelling content is difficult. As the bandwidth increases, the speed of data transmission also increases; hence,3G wireless networks can handle multimedia, Web browsing, and e-business messaging services. Given the enhancement of mo-bile phone features, mobile content can break the SMS text-messaging limit, allowing message content to include rich sound,video, and multimedia effects [2].

There are many types of mobile advertising for different communication methods and advertising purposes, but the mobileadvertisement methods mainly used fall into the categories of push and pull advertising [20]. In push advertising, the advertisingcompany proactively sends information to the user, such as the UK mobile advertising platform Celltick, which presents adver-tisements when a user's cell phone is in standby mode. If the user is interested in the advertisement, it can be directly clickedon. In pull advertising, the company links users to a Web site and learns more about the users' habits and preferences throughan interactive process. This allows effective communication of relevant advertising messages, often marketing to customers viapromotions and coupons. For example, Turkish mobile operator Turkcell-Tonla Kazan allows users to register as members onthe site and agree to receive advertisements for certain goods or services. It then sends advertisements and information aboutrelevant product promotions and coupons in accordance with the member's personal information [21]. Compared with tradition-al coupons, mobile coupon advertisements are more likely to be read and saved but are also more easily lost [22].

Zoller, Housen, and Matthews [23] divide mobile advertisements into the following three types, mainly based on the mode ofinteraction with the consumer.

• Permission-based advertising: permission-based advertising sends messages on specific goods and services targeted at individ-uals who show a clear willingness to receive advertisements. Therefore, if mobile-advertisement-related businesses can obtainthe user's permission to send messages, then user acceptance will be relatively high [24].

• Incentive-based advertising: incentive-based advertising provides individuals with incentives to agree to receive advertise-ments for promotional events. Park et al. [20] indicate that consumer willingness to accept this type of advertising is relativelyhigh.

• Location-based advertising: location-based advertising sends advertisements associated with an individual's current location ordestination. Location-based services can ascertain the exact location of a user via mobile devices and wireless networks, allow-ing advertisers to provide location-related real-time messaging services [25]. Therefore, not only can mobile advertisements

Table 1Comparison of advertisement media.

Media type Magazine/newspaper TV Radio Internet Mobile

Transmitted media Text/picture/physical Video Voice Video/picture/text Picture/videoDelivery style Pull Push Push Pull and push Pull and pushInvolvement Medium Low Low High HighTarget market Specific target market Mass market Specific target market Mass market and segmented

personalized marketMass market andsegmented personalizedmarket

Content Detailed Limited Limited Detailed LimitedPersonalization Medium Difficult Medium Easy EasyCommunications One-way One-way One-way Two-way Two-way

545P.-T. Chen, H.-P. Hsieh / Technological Forecasting & Social Change 79 (2012) 543–557

provide rich and diverse multimedia content, but they are also no longer limited by time, place, or other factors when conduct-ing real-time interactive communication, giving the mobile advertising market huge potential business opportunities.

Other scholars have also conducted researches on various mobile advertising topics as the following:

• Contextualized mobile advertising: Yuan and Tsao [26] discussed how to design a recommendation mechanism for contextualizedmobile advertising.

• Location-based Services: Varshney [25] conducted research on how to develop location-based services in mobile advertising.• Consumer attitude: Tsang et al. [24] reviewed the general attitude of consumers toward mobile advertising.• Factors influencing consumers' acceptance: Leppaniemi and Karjaluoto [27] and Drossos, Giaglis, Lekakos, Kokkinaki, and Stavraki[28] published papers on the factors influencing consumers to accept mobile advertising.

• Advertising platform management: Mahmoud and Yu [29] conducted research on the mechanisms of mobile advertising plat-form management.

• Influence of content presentation methods: Lee, Lee, Lee, Kim, and Lee [30] published a paper on the influence of perception andrecall of mobile advertising content presentation methods.

• Mobile advertising influence on consumer perception: Nasco and Bruner II [31] discussed the influence of mobile advertising onconsumer memory and perception.

• Relationship between consumer attitude and presentation style:Merisavo et al. [32] published a paper on the relationship betweenmobile advertising presentation style and consumers' attitude.

• Business model: Park et al. [20] discussed how to design a mobile advertisement business model and its related developmentstrategy.

• Policy issues: Evelyne [33] reviewed the policy and regulatory issues of mobile advertising.• Personalized mobile advertisement: Xu Liao and Li [34] discussed personalized mobile advertisement applications issues in thecatering industry.

• Consumer behavior: lastly, Soroa-Koury and Yang [35] studied the factors influencing consumers' behavior in response to mo-bile advertisements from the perspective of social morals.

In the above research studies on mobile advertising, only Xu et al. [34] refer to a portion of the individual attributes of mobileadvertisements in the design of personalized mobile advertising applications. The personalized mobile advertising attributesmentioned in the study are designed based on the needs of the food industry. Aside from ignoring some attributes, the studydoes not address the different business needs in the mobile advertising industry value chain as well as the feasibility and appli-cation of the mobile advertising attributes. In particular, identifying attributes to design an effective personalized mobile adver-tising message was not addressed in the past research studies.

3.2. Trends in personalized advertising

Enpocket [36] discussed the survey conducted by the Intelligent Mobile Marketing Company on the attitudes toward mobileadvertising of 1200 mobile phone users in the United States, Europe, and India. The results showed that 50% of users are willing toreceive mobile advertisements, 78% of users are happy to receive mobile advertisements that catch their interest, and 64% of usersare willing to provide personal information to receive relevant mobile advertisements. Moreover, 81% of users will not delete mo-bile advertisements before viewing them.

For service providers to provide customers with better service, they must interact with customers to obtain personal informa-tion and analyze each customer's needs and preferences, so that they may provide services that meet those needs and preferences[37]. The changing lifestyle of consumers has strengthened the outlook of service providers toward providing personalized goodsand services to different customers, emphasizing that content and media should be specialized for different target groups or in-dividuals [38].

The ARC Group [39] surveyedmobile advertising experts in different countries. They found that a personalizedmedium is the keyto accelerating mobile advertising because mobile implements provide personalized services. Thus, the loyalty of subscribers can beestablished by one-to-one marketing. By understanding an individual's needs, specific subscribers may provide different advertisingcontexts to achieve satisfaction [27]. Most scholars believe that personalizedmobile advertising can not only attract consumers' will-ingness to purchase products and services, but also serve as the major differentiator between mobile advertising and other advertis-ing media [30,34,40]. Hoffman and Novak [41] indicate that consumers' permission to use their personal information is key to thesuccessful development of mobile advertising [42]. Compared with traditional broadcasting advertising, mobile advertising providesadvertisers with the opportunity to offer personalized and interactive messages according to a consumer's unique location and envi-ronment [43,44].

3.3. Literature review on advertising attributes

Several scholars have discussed the important attributes in designing a mobile advertising message. Dey, Salber, and Abowd[45] indicate that when personalizing applications based on personal background information, advertisers should pay attentionto user activities. The analysis of Lee et al. [30] on the response rates and effectiveness of the use of mobile advertising inSouth Korean mobile phones indicates that one should focus on time and location when designing mobile advertisements.

546 P.-T. Chen, H.-P. Hsieh / Technological Forecasting & Social Change 79 (2012) 543–557

Leppaniemi and Karjaluoto's [27] investigation of consumer acceptance of mobile advertising notes that in personalized mobileadvertising, time and location are important personal factors on which to focus. Park et al. [20] reviewed 53 cases to explorethe effectiveness of mobile advertising. They noted that strengthening customer acceptance of advertisements is easier by consid-ering time and location, and that location, promotion, price, and brand are all attributes that should be considered. In investigat-ing the attributes related to the proliferation and success of mobile advertising, Schmidt, Beigl, and Gellersen [46] believed thatmobile advertising can provide personalized information based on time, location, interests, and user activities. The investigationof Venkatesh, Ramesh, and Massey [47] on location-intensive mobile commerce applications shows that time and place are themain focus in the application of personalized service. Varshney and Vetter's [48] discussion on a framework for emerging mobilecommerce applications states that in personalized service in mobile commerce, attention must be given to promotion, price,brand, background information, and preferences.

Vatanparast's [49] study on mobile advertising states that personalized services in mobile advertising must consider the attri-butes of location, time, background information, preferences, search history, and virtual communities. The research of Xu et al.[34] into the application of personalized mobile advertising mentions that design attributes include user activity, location,time, type of mobile device, promotion, price, brand, preferences, and other factors. Zhang and Shijagurumayum's [40] researchinto personalized content delivery to mobile devices shows that for personalized services, the type of mobile device, backgroundinformation, and interests must be considered. Zhang's [50] study on how to deliver personalized and adaptive content to mobiledevices reports that background information, preferences, and interests should be taken into account. The study of Cheverst,Mitchell, and Davies [51] on travel guides reveals that weather should also be considered. The study of Scharl et al. [22] on suc-cessful mobile marketing reports that personalized content should take into account place, time, and search history, among otherfactors.

The various personalized mobile advertising attributes mentioned in the above literature review are summarized in Table 2.These attributes include weather, user activities, location, time, device type, promotion, price, brand name, background informa-tion, preference, interests, search history, virtual community, and so on.

According to the literature review of Xu et al. [34] for the catering industry, the design of personalized mobile advertisements mustconsider the consumers' context, advertising content, consumers' demographics, and user preference. Selecting the appropriate targetmarket segment based on personal information, sending the user's desired advertising message, and creating personalized advertisingcontent are important issues that will affect the degree of consumer acceptance towards mobile advertising and its effectiveness. Con-text design elements mainly involve using attributes such as the user's location and environment to determine the opportunities forsending advertisements. Other factors include weather conditions, user activity, location, time, and mobile devices. For product infor-mation to be included in the design of the main text of the advertisement, it may include promotions, price, and brand. The personalprofile used to understand user's preference formarket segmentation can include background information, preferences, search history,and virtual communities.

4. Mobile advertising industry value chain

In the current system of mobile advertising industry value chain, the key members include advertisers, advertising platformproviders, content providers, mobile service providers, and mobile phone users. Advertisers entrust advertising plans to mobileadvertising platform providers, and mobile advertising platform providers and mobile service providers in turn cooperate with

Table 2Compiled chart of personalized mobile advertising attributes.

Attribute Literature

A B C D E F G H I J K L M

Weather ●User activities ● ● ●Location ● ● ● ● ● ● ● ● ●Time ● ● ● ● ● ● ● ●Device type ● ●Promotion ● ● ●Price ● ● ●Brand name ● ● ●Background information ● ● ● ●Preference ● ● ● ●Interests ● ● ●Search history ● ●Virtual community ●A: Cheverest et al. (1998) F: Zhang (2003) J: Lee et al. (2006)B: Schmidt et al. (1999) G: Zhang and Shijagurumayum (2003) K: Vatanparast (2007)C: Dey et al. (2001) H: Leppaniemi and Karjaluoto (2005) L: Park et al. (2008)D: Varshnev and Vetter (2001) I: Scharl et al. (2005) M: Xu et al. (2008)E: Venkatesh et al. (2003)

547P.-T. Chen, H.-P. Hsieh / Technological Forecasting & Social Change 79 (2012) 543–557

content providers to conduct advertising content production and design-related marketing activities. Finally, the mobile serviceprovider delivers the advertisements to mobile phone users. If users are interested in the content, they can reply to the mobileadvertising platform provider or the service provider. Mobile phone users can also register to receive advertisements for relatedproducts regularly. The business relationship among these players is depicted in Fig. 1.

The details of the business functions and interplayer relationship in the mobile advertising industry value chain are describedfurther as follows.

4.1. Advertiser

Advertisers want to use mobile advertising to consolidate existing markets and develop new potential customers. Hence, asidefrom sending advertisements to their own members, they also use the mobile service provider's database to select appropriatetargets to which to send the additional advertisements. The advertisement text production and event planning are entrusted tomobile marketing agencies and content providers, who will then send the completed advertising messages to users via the mobileservice providers. Interested users can find interactive advertising and further access relevant information. McDonald's, Coca-Cola, and Adidas are examples of such advertisers.

4.2. Content provider

Content providers are responsible for producing content for mobile advertisements. However, they lack advertisement planningand other aspects of marketing knowledge and will therefore likely cooperate or form strategic alliances with advertising agencies.For example, Cyber Agenthxyc and Pentsu established a joint venture with NTT DoCoMo to form CA Mobile [52].

4.3. Mobile service provider

Mobile service providers provide the fundamental infrastructure to transmit mobile advertisements and use their large userdatabases to help advertisers select appropriate recipients to improve advertising effectiveness. Cooperating with mobile market-ing agencies and content providers, mobile service providers, such as Chunghwa Telecom, Taiwan Mobile, and FarEaston, createadvertisement texts and plan marketing activities.

4.4. Mobile advertising platform providers

Mobile advertising platform providers (i.e., advertising media agencies) play an important role in leading the mobile advertisingindustry value chain. They are responsible for planning content and marketing activities for advertisers, as well as assessing andreporting results, so that advertisers can maintain advertising effectiveness.

Advertiser

Mobile advertisingplatform

providers/advertisingmedia agencies

Mobile serviceprovider

Mobile phoneuser

Marketing and planning outsourcing

Advertisementsend/receive feedback

information

Marketing

Content providerContent production

Service charge

Marketing expense

Advertisement transmitand response

Transport fee or monthlycommunications fees

Advertisementsend/receive fees

Marketing expense

Profit sharing

Information flow Cash flow

Feedback information

Content production

Fig. 1. Mobile advertising industry value chain.

548 P.-T. Chen, H.-P. Hsieh / Technological Forecasting & Social Change 79 (2012) 543–557

4.5. Mobile phone users

Mobile phone users who anticipate mobile advertising content can proactively search for advertisements that fit their prefer-ences and interests, and can also register certain products to receive advertising information for related products on a regularbasis.

For mobile phone users to see the value of personalized mobile advertising, all players in the mobile marketing industry valuechain will need to work together in a seamless manner. As each player in the value chain has different business objectives andinterests, creating a collaborative model for all players to work together will be critical. The key successful factor is addingvalue to mobile phone users in their daily activities [4].

5. Research method

This study first compiled a list of attributes comprising personalizedmobile advertising based on the literature review. The objectiveof this research is to identify a subset of attributes from the list that are important in designing a mobile personal advertisingmessage.Our study used a fuzzy Delphi expert questionnaire to survey mobile advertising industry value chain experts to collect their opinions.Given that evaluatorsmay have different perceptions on different attributes in terms of their importance and the possible adverse con-sequence, the evaluation is conducted in an uncertain and fuzzy environment. This fuzzy evaluation design allows evaluators to expresstheir opinions in a fuzzy expressionmanner. For these reasons, the fuzzy Delphimethodwas selected to conduct this evaluation. Basedon the evaluators' opinions, a fuzzy Delphi thresholdwas calculated to sort out the attributes that are likely to improve the effectivenessof personalized mobile advertising.

5.1. Questionnaire attribute analysis

Our questionnaire is organized into three dimensions: context, content, and personal profiles. Under each dimension, appro-priate attributes are selected based on the aforementioned literature review. Fig. 2 depicts the questionnaire hierarchy structure.A detailed description of the dimensions and attributes is illustrated in the following sections.

5.1.1. ContextIf an advertiser knows the consumer's current environment (mobile device, weather conditions, and current location), a mo-

bile advertising message can be effectively designed to meet the consumer's personalized needs [20]. Time and place, however,are generally lacking in other advertising media, which is the biggest difference betweenmobile advertising and other advertisingmedia; therefore, effective use of time and place in mobile advertising has the potential to increase revenue greatly [20,22,30].This study integrates relevant context attributes (i.e., weather, user activity, location, time, and mobile device) proposed by anumber of scholars as the attributes under context [22,27,34,45–47,49,51].

• Weather: Cheverst, Mitchell, and Davies [51] point out that among personal contextual factors, the weather is an important con-sideration [34]. Moreover, in their study that develops a personalized mobile advertising system that can protect personal pri-vacy, Scharl et al. [22] note that when designing the context in personalized advertisements, the weather, that is, weather

Goal

Personalized MobileAdvertisement

Dimension

Context

Content

Personal Profile

Attribute

Weather

User Activity

Location

Time

Device Type

Promotion

Price

Brand Name

Background Information

Preference

Interests

Search History

Virtual Community

Fig. 2. Questionnaire hierarchy structure.

549P.-T. Chen, H.-P. Hsieh / Technological Forecasting & Social Change 79 (2012) 543–557

information for a city or country provided to the user, is an important factor because it will influence individuals when planningactivities.

• User activities: user activities refers to the interaction of users with their environment [45,46]. We can determine user activitiesthrough their mobile devices and obtain highly personalized, relevant information to enhance the accuracy of targeted adver-tising and marketing.

• Location: using mobile devices and wireless networks to determine the user's location or destination, personalized advertisementinformation relevant to the area can be transmitted in real time [20,22,23,25,27,47–49,53].

• Time: understanding and analyzing user behavioral patterns can transmit appropriate advertising information to users at theappropriate time, significantly improving advertisement performance and customer satisfaction [22,27,30,34,47,49].

• Mobile device: by understanding the type of mobile device held by the user, sending content that is supported by the device andconforms to its specifications is possible [17,27,54,55].

5.1.2. ContentIn Zhang and Shijagurumayum's [40] investigation of a mechanism to load personalized content on mobile devices, note that

cell phones are highly personalized devices and thus are more capable of accurately targeting and marketing to customers thanother media. Furthermore, the personal elements to be considered must include the user's background information, interests, andtype of mobile device. According to Varshney and Vetter's [48] research, mobile advertising content must provide store sales in-formation to the user, mainly promotions, prices, and brands [34]. Under content dimension, this study selected three attributes,namely, store promotion, price, and brand information, for questionnaire design [20,34,48].

• Promotion: by determining the user's preferred products and services, relevant content can be sent to the user when there isinformation about promotional activities available [20,48].

• Price: given that the size and capacity of the windows on mobile devices are limited, only a limited amount of information canbe shown [20]. Providing clear price information to users will allow users to better understand the product price and services[20,48].

• Brand name: brands allow the identification of product manufacturers, distinguishing them from their competitors and providingrelevant brand messages to consumers, whom, in turn, are enabled to learn more about the product [20,48].

5.1.3. Personal profilesManagement of user profile information is the core of personalization technology [27]. Advertisers can better grasp the needs

of users by examining search history and virtual communities. This study examined the literature and industry data and selectedthe following attributes under personal profiles: background information, preferences, interests, search history, and virtual com-munities [22,48–50,56].

• Background information: background information includes age, income, education, and other basic personal information; allthese affect consumer attitudes toward advertisements [34,48–50]. Analyzing this personal data can help increase advertisingeffectiveness.

• Preferences: by understanding a user's preferences with regard to products and services, more relevant advertisements can beprovided, customer loyalty can be built, and purchase intention increased [34,48,49,53].

• Interests: understanding the products and services that interest customers and sending advertising messages that meet theirneeds can efficiently satisfy the target customers [27,40,49,50,57,58].

• Search history: the research of Rao and Minakakis [56] suggests that the understanding of user files, search history, and needs isan essential marketing skill. If personalized mobile advertisers can create a search, research, and tracking system, then they canaccurately target the customers; the system can also be used to understand the interests of existing and potential customers[49,57,58].

• Virtual community: Hagel III and Armstrong [59] brought the meaning of a virtual community to the business applications level.As virtual communities provide members with consumer information on things such as the price or quality of trade goods, oncethe membership reaches critical mass, consumer bargaining power and the potential effect of consumer behavior will increase;personalization strategies can allow these communities to be built faster. According to the investigation of Vatanparast [49] onthe factors influencing the development of mobile advertising, mobile advertising can accurately target specific groups to sendrelevant content to, and virtual communities create an important developing market.

The questionnaire was given to senior executives and managers in the mobile advertising industry value chain to determinethe technical and market feasibility of the dimensions and attributes. Evaluators include advertisers, mobile advertising agencies,content providers, and mobile service providers.

5.2. Fuzzy questionnaire design

The questionnaire structure has four sections: basic personal information, questionnaire explanation, linguistic scale, and question-naire content. In consideration of each respondent's different interpretations of the words used in the scale, the linguistic scale wasgiven five levels, namely, “very great impact,” “great impact,” “general impact,” “little impact,” and “minimal impact.” The respondentswere asked to rate each level from 0 to 100 according to importance, with a higher rating indicating greater importance.

550 P.-T. Chen, H.-P. Hsieh / Technological Forecasting & Social Change 79 (2012) 543–557

5.3. Fuzzy Delphi method

The Delphi method was first developed by Dalkey and Helmer [60] for corporations and has been widely applied in manage-ment areas such as in forecasting, public policy analysis, and project planning [61]. The Delphi method is one of the effectivemethods that enable forecasting by converging a possibility value through the feedback mechanism of the results of the question-naires based on experts' judgments [9]. Delphi is an expert opinion survey method with three features: anonymous response, it-eration and controlled feedback, and statistical group response. However, some weaknesses of the method have been exposed: itneeds repetitive surveys to allow forecasting values to converge, which requires muchmore time and cost [8]. The Delphi methodhas always suffered from low convergence expert opinions, high execution cost, and the possibility that opinion organizers mayfilter out particular expert opinions [11].

Thus, Murry, Pipino, and Gigch [62] propose the concept of integrating the traditional Delphi method and the fuzzy theory to im-prove the shortcomings of vagueness and ambiguousness of Delphi. Group membership functions were used to compute theexpected values of fuzzy numbers, resulting in a forecasting value [63]. The fuzzy Delphi method effectively gathers information onthe developing critical technology selection criteria on the aspects of economic or social prospects and problems, and at the sametime reduces the uncertainty and ambiguity existing in experts' judgments [64]. The fuzzy Delphi method requires only one investi-gation, and all the opinions can be covered and respect the original opinions of all the experts. It gives a differentmembership degreefor each possible consensus. In addition to fuzzy parts in human thinking, uncertain and subjective messages can also be induced.Moreover, it reduces the time of investigation, cost, and time consumption [9–11]. Thus, the fuzzy Delphi method, which can bothdeal with fuzziness and vagueness in experts' expressions and reduce the number of rounds in facilitating the formation of a groupjudgment, is applied in our study to select the most critical attributes [65]. The fuzzy Delphi method steps are as follows [66–68].

Step 1:Each expert separately assesses each evaluation item, giving a possible range of values. The minimum value of this intervalindicates this expert's “most conservative cognitive value,” whereas the maximum value of this interval indicates the expert's“most optimistic cognitive value.”

Step 2:For each evaluated item i, analyze the “most conservative cognitive value” and “most optimistic cognitive value” and re-move any extreme values that fall outside two times the standard deviation. Among the remaining “most conservative cognitivevalues,” find the lowest value CL

i, the geometric mean CMi , and the maximum CU

i . Among the “most optimistic cognitive values,”find the minimum OL

i , geometric mean OMi , and maximum OU

i .

Step 3:Through the above steps, we can establish the triangular fuzzy numbers C i=(CLi,CMi ,CUi) for the “most conservative cogni-tive value” and Oi=(OL

i ,OMi ,OU

i ) for the “most optimistic cognitive value.”

Step 4:Finally, the level of expert consensus can be tested as follows:

1.There is no gray area.

If CUi ≤OLi , that is, the two triangular fuzzy numbers, do not overlap, then the experts' interval values have a consensus section.

This indicates that for the assessed item i, the most conservative cognitive values of all the experts have reached a consensus. Sim-ilarly, for the assessed item i, the most optimistic cognitive values of all the experts have also reached a consensus. Therefore, letthe assessed item i's “consensus importance value” G i be equal to the arithmetic mean of CMi and OM

i :

Gi¼ðCMi þOM

i Þ =2: ð1Þ

2.There is a gray area, but there is little difference between the views of experts.

If cUi NoLi , the two triangular fuzzy numbers overlap, and the fuzzy relationship's gray area Z i=CUi −OL

i is less than the intervalrange M i=OM

i −CMi of the geometric mean of the “optimistic cognitive value” and “conservative cognitive value,” then although

the interval value of the experts' opinions generates a fuzzy zone, the experts who gave extreme opinions do not have too great adifference in opinion from the views of other experts; thus, there is no opinion divergence. Therefore, let the “consensus impor-tance value” G i equal the fuzzy set F i(xj) calculated from the intersection of the gray area of the two triangular fuzzy numbers.Calculate the quantitative score μF4(xj) with the maximum degree of membership in the fuzzy set. The expression is as follows:

Fi ðxj Þ¼(∫x

nminZci ðxj Þ ;oi ðxj Þx

odx

)

Gi ¼nxj jmax μF

4 ðxj Þo:

3.There is a gray area, and there is a big difference between the views of experts.

If CUiNOLi, the two overlapping triangular fuzzy numbers overlap, and the fuzzy gray area Zi¼CUi−OLi is greater than theinterval rangeMi¼OMi−CMi of the geometric mean of the optimistic and conservative cognitive values, then the range of values

(2)

(3)

551P.-T. Chen, H.-P. Hsieh / Technological Forecasting & Social Change 79 (2012) 543–557

of the views of the experts creates a fuzzy zone without consensus. This means that the extreme values given by some expertshave too great a difference from the views of other experts, resulting in opinion divergence. Therefore, we must allow the expertsto assess the geometric mean of the “optimistic cognitive” and “conservative cognitive” for the items that did not achieve conver-gence. Repeat steps 1–4 to conduct another survey until all evaluated items achieve convergence and the “consensus importancevalue” Gi can be calculated.

The higher the calculated value of each “consensus importance value” Gi described above, the higher the level of expert con-sensus is. Afterwards, using the geometric mean of the most likely single value for each item, find the arithmetic mean and use itas the threshold value of this study to select an appropriate number of evaluation attributes that have attained expert consensus.

6. Empirical study

This survey was intended to be comprehensive and to cover most of the perspectives of every relevant type of expert in themobile marketing industry. Basic questionnaire data were analyzed, including the questionnaire recovery percentage and ques-tionnaire respondents' job classification and seniority level, to better understand the respondents' credential. Personalized mobileadvertising design attributes above the calculated threshold were selected and identified as important attributes in designing amobile marketing advertising message. Different perspectives based on the evaluator's role and job function were analyzed.The differences between current practices and future trends in mobile advertising were also explored.

Questionnaireswere distributed to experts in industries related tomobile advertising; 61 valid questionnaireswere obtained. Per-sonal demographic statistical analysis was performed on the basic information of the valid questionnaire respondents, including jobclassification and industry seniority, for the purpose of demonstrating thewide spectrumof representation of the survey respondents.

6.1. Analysis of the respondents' personal demographic statistics

The expert questionnaires are used in this study to collect experts' professional abilities and practical experience. The studywas conducted from 1 November 2009 to 5 March 2010. The evaluators include mid- to high-level executives in charge of busi-nesses related to mobile advertising from five major Taiwanese mobile service providers (i.e., Chunghwa Telecom, TaiwanMobile,Far Eastone, Vibo, and Asia Pacific Telecom), who returned 16 valid questionnaires. The Taiwanese advertising media agenciesreturned seven valid questionnaires. With regard to advertisers, approximately 100 local businesses with their own productbrands listed in the Taiwan top 1000 companies were given our questionnaire, but only 34 valid questionnaires were returned(Advertisers). Among the seven major Taiwanese mobile advertising platform providers (i.e., Diymad, Every8D, Freed, MediaPalete, MobiBon, Joymaster, and Broadfast), only four returned valid questionnaires. Note that the founders of the four companiesfilled out these four questionnaires. In total, there were 61 valid questionnaires filled out by executives and managers with an av-erage of 13 years' industry experience. Although the sample size was small, however, the questionnaires were answered by thekey parties in the mobile marketing industry.

6.1.1. Analysis of the survey based on the respondents' job levelsAmong the survey respondents, 28% were entry-level managers, 23% were frontline staff, 21% were high-level executives, 15%

weremiddlemanagers, and 13%weremid-level executives (Table 3). This analysis indicates that the return questionnaires representa wide distribution of job functions and are not only focused on a certain section of the job level in the mobile marketing industry.

6.1.2. Analysis of the respondents' seniorityThirty-one percent of the respondents had worked in the industry for 5–10 years, 25% had worked for 1–5 years, 20% had worked

for more than 20 years, 15% had worked for 10–15 years, and 10% had worked for 15–20 years (Table 4). Based on this respondent se-niority analysis, our research result is well represented by the different experience levels of people in the mobile marketing industry.

6.2. Analysis of attributes of personalized mobile advertising

This study used the fuzzy Delphi method and Excel to calculate a threshold value of 71.82, which was used as the standard forselecting the personalized mobile marketing attributes. Six important design attributes for personalized advertisements abovethe threshold value were identified: price (80.12), preference (78.00), promotion (76.81), interest (75.47), brand (74.88), and

Table 3Analysis of the survey respondents' job level.

Job level Questionnaires returned Percentage

High-level executives 13 21Mid-level executives 8 13Middle managers 9 15Entry-level managers 17 28Frontline staff 14 23Total 61 100

552 P.-T. Chen, H.-P. Hsieh / Technological Forecasting & Social Change 79 (2012) 543–557

type of mobile device (74.51) (Table 5). Note that time (70.09), location (71.49), and search history (70.49) were all close to the71.82 threshold value. The weather attribute received the lowest value (57.25) in the analysis, indicating that only limited mar-keting conditions require weather information. Other attributes that received a calculated value below 71.28 deserve further re-view under special business scenarios.

6.3. Comparison of expert group opinions

The subjects of this study are groups of experts with industry influence on the development of mobile advertising, includingthose working with mobile service providers, advertising media agencies, mobile advertising platform providers, and advertisers.Table 6 shows the opinion of each group of experts on the various design attributes for personalized mobile advertisements, in-dicating each group's different attributes to be important. The three most important attributes for mobile service providers areprice (84.05), interest (76.53), and promotions (73.54); for advertising media agencies, they are background information(78.56), mobile device type (77.20), and preferences (77.10); for mobile advertising platform providers, they are location(87.78), preferences (83.67), and promotions (83.59); and for advertisers, they are price (80.06), preferences (79.63), and mobiledevice type (78.34). Each player in the mobile marketing industry has its own priority in designing a personalized mobile adver-tising message; therefore, different expert groups have different attribute priorities in the survey results. However, price attributewas given the highest score by two groups, indicating the importance of price information in a personalized mobile marketingadvertising message.

Currently, among the personalized design attributes observed in practice in the industry, themost common are promotions, brand,and interests. However, questionnaire respondents generally believe that price, preferences, and promotions are the most important.As promotions can easily draw consumer attention and interest, this element plays a pivotal role in the design of personalized mobileadvertisements. Given the mobile and instantaneous nature of mobile advertising, advertisements can be sent to users anytime andanywhere. Thus, accommodating advertisers' promotional activities can enhance advertising effectiveness. Mobile advertising mes-sages can easily be saved and carried, satisfying consumer needs at all times. Unlike traditionalmodels, mobile advertising can employpersonalized design to enhance advertising effectiveness. Product brand is another important design element among respondents and

Table 5Fuzzy Delphi survey results.

C i O iG iCriterion

LC i

MC i UC i LO i MO i UO iM i Z i

LUC i − O i

Weather 0 48.82 90 20 65.68 100 16.86 70 57.25+User activities 0 60.97 90 15 76.88 100 15.91 75 68.93

Time 0 61.45 90 15 78.72 100 17.27 75 70.09

Mobile device 0 66.82 90 3073.91

3100 15.98 60 74.81

Location 0 63.98 95 6 79 100 15.02 89 71.49

Promotion 10 68.88 90 40 84.92 100 16.04 50 76.90

Price 35 72.43 90 40 88.17 100 15.74 50 80.30

Brand 0 66.52 90 40 83.67 100 17.15 50 75.10

Background information 20 61.60 90 40 78.22 100 16.62 50 69.91

Preference 35 70.33 90 20 85.30 100 14.97 70 77.82

Interest 16 69.35 90 28 85.83 100 16.48 62 77.59

Search history 0 61.02 90 28 80.17 100 19.15 62 70.60

Virtual community 0 57.13 90 15 73.63 100 16.50 75 65.38

Threshold value 71.82

++

+

+++++++++

Note 1: the “+” symbol in the table means CUi ≤OL

i , indicating that that the experts have consensus; the consensus value has been calculated with the formulaGi=(CM

i +OMi /2).

Note 2: rows colored gray indicate that the attributes are not selected.Note 3: Ci represents most conservative cognitive triangular fuzzy number.Note 4: Oi represents most optimistic cognitive triangular fuzzy number.

Table 4Seniority of the valid questionnaire respondents.

Years in Industry Questionnaires returned Percentage

1–5 15 255–10 19 3110–15 9 1515–20 6 10N20 12 20Total 61 100

553P.-T. Chen, H.-P. Hsieh / Technological Forecasting & Social Change 79 (2012) 543–557

in the current industry practice. Due to the limited nature ofmobile device interfaces, brand effect also plays an important role. Ifmobileadvertisements contain information about consumers' preferred brands, they can improve consumer acceptance and brand loyalty. Forexample, iAd and Nike worked together to release a series of advertisements documenting the events in Nike's history, packaging themarketing goals of the advertisement in a video and allowing consumers to learn about past and future Nike products.

7. Discussion and conclusion

Advertising continues to evolve from mass distribution to personalized messages. On the one hand, mass-distributed adver-tising messages continue to build a brand image in consumers' minds. On the other hand, personalized advertising messagescan proactively add value and support to consumers' daily lives and seamlessly be integrated into people's daily activities. For ex-ample, consumers' favorite products and services can now be linked to the consumers' itinerary and be presented to them at theright time and at the right place to meet their needs. The mass and personalized advertising messages can work hand in hand tocomplement each other in promoting products and services. However, as personalized advertising starts to integrate into people'sdaily lives, consumer privacy could become an issue, particularly if a security issue is caused by the misuse of private information.Moreover, personalized advertising could potentially promote a consumption-oriented society, which may not be in the best in-terest of the environment. One thing is certain: personalized advertising will affect people's future purchasing behavior and bemore interactive with advertisers and commercial vendors and suppliers.

In Japan, an analysis of successful mobile services and the lead users suggests that the integration of mobile sites with othermedia is a major driver of mobile shopping. For example, a growing fraction of Net Price's Xavel and Index's sales are for productsintroduced in television programs and magazines, respectively. Xavel has used the success of mobile shopping to open stores andbegin connecting the virtual and physical worlds with 2D bar codes and phones that read them with cameras and bar code rec-ognition software. Opposed to replacing other media as some argued the PC Internet would do in the late 1990s, this suggests thatthe mobile Internet will likely draw on the advantages of these media and the consumer behavior established from them. Inte-grating mobile sites with magazines and radio and TV programs enables the mobile sites to draw on existing consumer searchprocesses in other media [69].

Finally, personalized mobile advertising also highlights the importance of collaboration among all players in the industry valuechain. How to create a win–win model for all players in the value chain will be critical for the success of the mobile advertisingindustry. Issues related to the industry collaborative model include profit sharing mechanism, phone user information sharingand tracking, consumer social networking protection, and others. Concerted effort from all players is required to address theseissues and create a healthy and successful mobile marketing advertising industry.

This study reviewed relevant research onmobile marketing advertising and industry practices, and compiled a list of 13 designattributes of personalized mobile advertising. Using a fuzzy Delphi expert questionnaire, the following six design attributes,which received a calculated value above the threshold value (71.82), were selected: mobile device type (74.51), from the contextdimension; promotions (76.81), price (80.12), and brand (74.88), from the content dimension; and preferences (78.00) and in-terest (75.47) from the personal profile dimension. Therefore, to increase advertising effectiveness, when designing personalizedmobile advertisements, the user's preferences, interests, and mobile device should be considered, and message contents shouldalso include price and promotional information.

The large number of mobile devices with different specifications currently available affects consumers' ability to receive andview advertising messages properly. Therefore, the respondents believe that when transmitting mobile advertisements, consid-ering the type and specifications of the target user's mobile device is important to ensure that advertising messages can be properlydisplayed.

The respondents believe that advertisements must be able to present price, promotional activity, and brand information evenat the limited size of the mobile device user's interface. The value and meaning of the brand itself will assist consumers in judging

Table 6Expert group ratings of the design criteria.

Criterion Expert group

Mobile service providers Advertising media agencies Mobile advertising platform providers Advertisers

Weather 50.63 54.35 55.43 59.91User activities 67.09 56.76 76.94 70.02Time 71.59 55.09 73.99 71.45Mobile device 65.23 50.89 60.10 74.23Location 68.86 77.20 (2) 87.78 (1) 78.34 (3)Promotions 73.54 (3) 72.39 83.59 (3) 78.18Price 84.05 (1) 70.39 79.52 80.06 (1)Brand 71.75 70.51 69.95 77.15Background information 63.48 78.56 (1) 70.60 69.75Preferences 72.66 77.10 (3) 83.67 (2) 79.63 (2)Interests 76.53 (2) 73.71 77.03 78.25Search history 63.38 62.29 76.49 73.58Virtual community 59.28 58.44 68.97 68.22

Bold ratings are the top three important design criteria for each expert group.

554 P.-T. Chen, H.-P. Hsieh / Technological Forecasting & Social Change 79 (2012) 543–557

the merits of the product. The price of the product is also a common and important consideration for consumers when decidingwhether to buy the product. Prices and related promotional activities will draw consumers' attention and affect their desire topurchase the product. As consumers are currently unable to use their mobile devices to find and compare prices quickly, respon-dents believe that clearly displaying the prices and related promotional activities can increase consumers' willingness to receivemobile advertisements and convince them to take a purchasing action after viewing the message.

In order to avoid consumer resentment, the respondents believe that when designing personalized mobile advertisements,consumers' interests and preferences must be taken into consideration.

Location (71.26), search history (69.97), time (69.84), background information (69.77), user activity (68.72), virtual commu-nity (65.35), and weather (56.93) fell below the threshold value and were eliminated in the advertising attribute list by the fuzzyDelphi research model. By examining the current industry practices, many advertisers apparently use location when designingand sending personalized mobile advertisements. For example, when a consumer enters a particular shopping district, advertisersmay send him/her advertisements on the businesses in the area. Therefore, a possible reason for location attribute to receive a lowcalculated score is due to privacy and personal safety issues. As most mobile advertisers do not obtain consumers' permission be-fore sending advertisements, the respondents become annoyed by the mobile advertising message and are often reluctant tomake use of the mobile advertising information due to privacy concerns. However, if mobile advertisers can implement a mech-anism for obtaining consumer permission, a permission-based marketing model coupled with personal data protection measureswill aid in the development of mobile advertising.

Although background information, search history, virtual communities, and user activities are widely used in the design anddelivery of online advertisements, current mobile networks and mobile communities are not yet universal, and the databases ofinternet users and mobile users are not well connected. Therefore, it is difficult for members of the mobile advertising industry toobtain user data on these attributes. Consequently, some of the respondents may not consider them to be important in the mobileadvertising message. However, if internet and mobile user databases could be connected, or if niche communities were developedbased on users' personal interests and preferences, it could have the potential to increase the effectiveness of mobile advertisinggreatly.

8. Implications and recommendations

Personalizedmobile advertisement is expected to enhance the effectiveness of mobile advertising. By providing information onproducts that interest consumers, including price, promotional activity, and brand information, properly formatted for each con-sumer'smobile device, advertisers allow consumers to obtain detailed product information quickly. If consumers can properly receiveand view advertising content custom-tailored for them, it will leave a good impression and increase their desire to purchase the prod-uct. For advertisers, if they are able to direct their advertising resources to the appropriate target groups, they can reduce advertisingandmarketing costswhile improving advertising effectiveness at the same time. In addition to retaining their original customers, theymaymore accurately develop new potential markets. If mobile advertisers can design appropriate mechanisms to ensure the privacyof their customers and provide secure transaction mechanisms, they will be able to increase consumer acceptance of mobile adver-tisements and accelerate the purchase decision-making process.

The questionnaire was distributed to major members of the mobile advertising industry chain, including mobile service pro-viders, advertising media agencies, mobile advertising platform providers, and advertisers. Although small sample size of the ad-vertisers could possibly affect the accuracy of the advertisers' opinion, this study still provides a good reference model andguideline in designing a mobile advertising message. In addition, players in the industry value chain can also evaluate the impor-tance of each advertising attribute according to their unique function and role in the industry value chain.

With the rapid development of new technology, internet and mobile networks have become inseparable from the public'sdaily lives. Moreover, advertisements have moved from traditional newspapers, magazines, TV, and radio to internet and mobilenetworks. Although consumers use mobile devices mainly for communication, they have increasingly turned to mobile carriersfor additional value-added services. With the increase in mobile network bandwidth and the development of smartphones, mo-bile advertisements can now be more varied and personalized. However, personalizing mobile advertisements requires a largeamount of personal information. Members of the mobile advertising industry need to implement cross-platform media integra-tion so that they may effectively design integrated marketing campaigns. The vigorous development of mobile positioning ser-vices and community Web sites, combined with personal information that is properly collected and integrated, can enrichpersonalization design attributes, improving personalized advertisement design and allowing the accurate transmission of mobileadvertisements to the appropriate targets.

At the same time that personalized advertising is being developed, consumers have become increasingly aware of privacy anddata safety issues. Currently, several countries already have legislation stating that mobile advertisements cannot be sent withoutthe user's permission. However, Taiwan currently does not have any laws or regulations dealing with privacy or related mobilesecurity issues. As a result, when designing personalized advertisements for consumers, it is necessary to have a safe and reliablemechanism for maintaining their privacy and safety to gain their trust.

This study focuses on the general products and did not investigate the difference between goods and service in personalizedmobile advertisement. However, there are unique characteristic differences between them. Future studies could investigate ad-vertising attributes difference between goods and service. In addition, the opinions of consumers should also be integratedwith those of expert members in the industry, thereby increasing the value and representativeness of the research.

555P.-T. Chen, H.-P. Hsieh / Technological Forecasting & Social Change 79 (2012) 543–557

Acknowledgments

This research was made possible by the support and assistance of a number of people whom we would like to thank. We arevery grateful to the anonymous referees for their valuable comments and constructive suggestions. We would like to thank all therespondents for their valuable opinions. This research was supported by the National Science Council under grant number NSC99-2410-H-214-019-MY2.

References

[1] G. Preez, C. Pistorius, Analyzing technological threats and opportunities in wireless data services, Technol. Forecast Soc. Change 70 (1) (2003) 1–20.[2] M.E. Mostafa, MMS—the modern wireless solution for multimedia messaging, IEEE Int. Symp. Pers. Indoor Mobile Radio Commun. 5 (15) (2002) 2466–2472.[3] P.T. Chen, H.P. Hsieh, J.Z. Cheng, Y.S. Lin, Broadband mobile advertisement: what are the right ingredient and attributes for mobile subscribers, Conf. Technol.

Manage. Age. Fundam. Change, Portland, Oregon, USA, 2009.[4] P.T. Chen, J.Z. Cheng, Unlocking the promise of mobile value-added services by applying new collaborative business models, Technol. Forecast Soc. Change

77 (4) (2010) 678–693.[5] Y.C. Tsai, K.C. Chung, The worldwide handset subscribers forecasting, 2011~2015, Market Intelligence & Consulting Institute (MIC), Taiwan, 2011.[6] C.Y. Yang, The Observation of the Mobile Internet by the End of the Fourth Quarter of 2010, Foreseeing Innovative New Digiservices (FIND), 2011, http://

www.find.org.tw/find/home.aspx?page=many&id=283, Retrieval date: May 17, 2011.[7] R. Bender, Mobile-ad market still faces hurdles, Wall St. J. (2011), http://online.wsj.com/article/SB10001424052748704900004576152132113932442.html.

Retrieval date: May 17, 2011.[8] J.S. Hwang, G.D. Lin, Nonlinear approximation and moment problem, Nonlinear Anal. (1987) 337–353.[9] I.S. Chang, Y. Tsujimura, M. Gen, T. Tozawa, An efficient approach for large scale project planning based on fuzzy Delphi method, Fuzzy Sets. Syst. 76 (3)

(1995) 277–288.[10] P.T. Chang, L.C. Huang, H.J. Lin, The fuzzy Delphi method via fuzzy statistics and membership function fitting and an application to the human resources,

Fuzzy Sets. Syst. 112 (3) (2000) 511–520.[11] Y.F. Kuo, P.C. Chen, Constructing performance appraisal indicators for mobility of the service industries using fuzzy Delphi method, Expert Syst. Appl. 35 (4)

(2008) 1930–1939.[12] AmericanMarketing Association, AMADictionary ofMarketing Terms, 2003, http://www.marketingpower.com/live/mgdictionary. Retrieval date: September 10,

2008.[13] K. Risden, M. Czerwinski, S. Worley, L. Hamilton, J. Kubiniec, H. Hoffman, N. Mickel, E. Loftus, Interactive advertising: patterns of use and effectiveness, Pro-

ceedings of the SIGCHI Conference on Human Factors in Computing Systems, Los Angeles, California, United States, April 18–23, 1998, pp. 219–224.[14] W.Macias, A beginning look at the effects of interactivity, product involvement andweb experience on comprehension: brandweb sites as interactive advertising,

J. Curr. Issues Res. Advert 25 (2) (2003) 31–44.[15] Mobile Marketing Association, Wireless Definitions, 2003, http://www.mmaglobal.com/about/index.html. Retrieval date: October 28, 2003.[16] Wireless Advertising Association, 2002, http://www.imediaconnection.com/news/406.asp. Retrieval date: September 01, 2009.[17] Interactive Advertising Bureau [IAB], Platform Status Report: A Mobile Advertising Overview, Marketing & Media, 2008, pp. 1–21.[18] C.Y. Chen, The developing trend of mobile internet, Market Intelligence & Consulting Institute (MIC), Taiwan, 2008.[19] B. Paul, Communication Pty Ltd, Global — Key Telecoms, Mobile and Broadband Statistics, 2010, http://www.researchandmarkets.com/reportinfo.asp?

report_id=1059189&t=d&cat_id=. Retrieval date: May 17, 2011.[20] T. Park, R. Shenoy, G. Salvendy, Effective advertising onmobile phones: a literature review and presentation of results from 53 case studies, Behav. Inf. Technol 27

(5) (2008) 355–373.[21] S.Y. Jhuang, Tuekcell launched a new mobile advertising service model, greatly improve the click rate, Market Intelligence & Consulting Institute (MIC), Taiwan,

2009, http://www.find.org.tw/find/home.aspx?page=news&id=5591. Retrieval date: September 23, 2009.[22] A. Scharl, A. Dickinger, J. Murphy, Diffusion and success factors of mobile marketing, Electron. Commer. Res. Appl. 4 (2) (2005) 159–173.[23] E. Zoller, V.L. Housen, J. Matthews, Wireless Internet business models — global perspective, regional focus, OVUM. 2001 Report, 2001, pp. 1–64.[24] M.M. Tsang, S.C. Ho, T.P. Liang, Consumer attitudes toward mobile advertising: an empirical study, Int. J. Electron. Commer. 8 (3) (2004) 65–78.[25] U. Varshney, Issues, requirements and support for location-intensive mobile commerce applications, Int. J. Mob. Commun. 1 (3) (2003) 247–263.[26] S.T. Yuan, Y.W. Tsao, A recommendation mechanism for contextualized mobile advertising, Expert Syst. Appl 24 (4) (2003) 399–414.[27] M. Leppaniemi, H. Karjaluoto, Factors influencing consumers' willingness to accept mobile advertising: a conceptual model, Int. J. Mob. Commun. 3 (3)

(2005) 197–213.[28] D. Drossos, G.M. Giaglis, G. Lekakos, F. Kokkinaki, M.G. Stavraki, Determinants of effective SMS advertising: an experimental study, J. Interact. Advert. 7 (2)

(2007), http://www.jiad.org/article90. Retrieval date: July 12, 2009.[29] Q.H. Mahmoud, L. Yu, Havana agents for comparison shopping and location-aware advertising in wireless mobile environments, Electron. Commer. Res.

Appl. 5 (3) (2006) 220–228.[30] H.S. Lee, C.H. Lee, G.H. Lee, Y.H. Kim, G.B. Lee, Analysis of the actual response rates in mobile advertising, Innov. Inf. Technol 1 (2006) 1–5.[31] S.A. Nasco, G.C. Bruner II, Perceptions & recall of advertising content presented on mobile hanged devices, J. Interact. Advert. 7 (2) (2007), http://www.jiad.

org/article93. Retrieval date: September 21, 2009.[32] M. Merisavo, S. Kajalo, H. Karjaluoto, V. Virtanen, S. Salmenkivi, M. Raulas, M. Leppniemi, An empirical study of the drivers of consumer acceptance of mobile

advertising, J. Interact. Advert. 7 (2) (2007)http://www.jiad.org/article92. Retrieval date: September 23, 2009.[33] B.C. Evelyne, Regulating mobile advertising in the European Union, Comput. Law Secur. Rep. 24 (5) (2008) 421–436.[34] D.J. Xu, S.S. Liao, Q. Li, Combining empirical experimentation andmodeling techniques: a design research approach for personalizedmobile advertising applications,

Decis. Support. Syst. 44 (3) (2008) 710–724.[35] S. Soroa-Koury, K.C.C. Yang, Factors affecting consumers' responses to mobile advertising from a social norm theoretical perspective, Telematics Inform. 27

(1) (2010) 103–113.[36] Enpocket, 2006, http://www.enpocket.com/news/press-releases/researchshows. Retrieval date: October 09, 2008.[37] C.F. Surprenant, M.R. Solomon, Predictability and personalization in the service encounter, J. Mark. 51 (April 1987) 86–96.[38] C. Allen, Personalization vs. customization, 1999, http://www.searchz.com/Articles/0713991.shtml. Retrieval date: September 10, 2008.[39] ARC Group, Factors of restricting mobile advertising development, 2001, http://www.linkedin.com/pub/mauricio-kurcgant/6/188/30b. Retrieval date:

September 10, 2008.[40] D. Zhang, S. Shijagurumayum, Personalized content delivery to mobile devices, Syst. Man. Cybern. 3 (2003) 2533–2538.[41] D.L. Hoffman, T.P. Novak, Marketing in hypermedia computer-mediated environments: conceptual foundations, J. Mark. 60 (July 1996) 50–68.[42] P. Barwise, C. Strong, Permission-based mobile advertising, J. Interact. Mark. 16 (1) (2002) 14–24.[43] G.E. Belch, M.A. Belch, Advertising and Promotion: An Integrated Marketing Communications Perspective, McGraw-Hill, Singapore, 1999.[44] T. Saari, N. Ravaja, L. Jari, M. Turpeinen, K. Kallinen, Psychologically targeted persuasive advertising and product information in e-commerce, Int. Conf. Elec-

tron. Commer, ACM press, Delft, the Netherlands, 2004, pp. 25–27.[45] A.K. Dey, D. Salber, G.D. Abowd, A conceptual framework and a toolkit for supporting the rapid prototyping of context-aware applications, Hum. Comput.

Interact. J. 16 (2–4) (2001) 99–116.

556 P.-T. Chen, H.-P. Hsieh / Technological Forecasting & Social Change 79 (2012) 543–557

[46] A. Schmidt, M. Beigl, H.W. Gellersen, There is more to context than location, Comput. Graph. 23 (6) (1999) 893–901.[47] V. Venkatesh, V. Ramesh, A.P. Massey, Understanding usability in mobile commerce, Commun. ACM 46 (12) (2003) 53–56.[48] U. Varshney, R. Vetter, A framework for the emerging mobile commerce applications, Syst. Sci, Proceedings of the 34th Annual Hawaii International Confer-

ence, 2001.[49] R. Vatanparast, Piercing the fog of mobile advertising, Six Int. Conf. Manage. Bus (ICMB 2007), 2007, p. 19.[50] D. Zhang, Delivery of personalized & adaptive content to mobile devices: a framework and enabling technology, Commun. AIS 12 (13) (2003) 183–202.[51] K. Cheverst, K. Mitchell, N. Davies, Design of an object model for a context sensitive tourist guide, Int. Workshop. Interact. Appl. Mob. Comput (IMC98), 1998,

Rostock, Germany.[52] X.Y. Li, The analysis of mobile advertising, Market Intelligence & Consulting Institute (MIC), Taiwan, 2007.[53] M.M. Lankhorst, H.V. Kranenburg, A. Salden, A.J.H. Peddemors, Enabling technology for personalizing mobile services, Proceedings of the Hawaii, Int. Conf.

Syst. Sci, 2002, pp. 1464–1471.[54] J.L. Naccarato, K.A. Neuendorf, Content analysis as a predictive methodology: recall readership and evaluations of business-to business print advertising, J.

Advert. Res. 38 (3) (1998) 19–33.[55] Y.H. Mohamed, G.J. Zeyu, S. Shim, Wireless advertising's challenges and opportunities, IEEE. Comput. Soc. 36 (5) (2003) 30–37.[56] B. Rao, L. Minakakis, Evolution of mobile location-based services, Commun. ACM 46 (12) (2003) 61–65.[57] D.T. Hawkins, Electronic advertising on online information systems, Online 18 (2) (1994) 26–39.[58] R. Zeff, B. Aronson, Advertising on the Internet, John Wiely and Sons, New York, NY, 1997.[59] J. Hagel III, A.G. Armstrong, Expanding Markets through Virtual Communities, Harvard Business School Press, Boston, 1997.[60] N.C. Dalkey, O. Helmer, An experimental application of the Delphi method to the use of experts, Manage. Sci. 9 (3) (1963) 458–467.[61] P.C. Chang, Y.W. Wang, Fuzzy Delphi and back-propagation model for sales forecasting in PCB industry, Expert. Syst. Appl. 30 (4) (2006) 715–726.[62] T.J. Murry, L.L. Pipino, J.P. Gigchvan, A pilot study of fuzzy set modification of Delphi, Hum. Syst. Manage. 5 (1) (1985) 76–80.[63] A. Kaufmann, M.M. Gupta, Fuzzy Mathematical Models in Engineering and Management Science, North-Holland, NewYork, NY, 1988.[64] Y.C. Shen, S.H. Chang, G. Lin, H.C. Yu, A hybrid selection model for emerging technology, Technol. Forecast Soc. Change 77 (1) (2010) 151–166.[65] A. Lee, W.M. Wang, T.Y. Lin, An evaluation framework for technology transfer of new equipment in high technology industry, Technol. Forecast Soc. Change

77 (1) (2010) 135–150.[66] P.T. Chen, Y.S. Chang, S.C. Chiou, H.C. Yu, Assessing ITRI OpenLab's assistance to incubate startups, Int. J. Innov. Incubation 1 (1) (2004) 117–136.[67] M.S. Lee, The study of key success factors for competitive advantages as well as strategies in promotion of agricultural biotechnology industry in Taiwan, J.

Agric. Econ. 82 (2007) 107–156.[68] Y.L. Hsu, C.H. Lee, V.B. Kreng, The application of Fuzzy Delphi Method and Fuzzy AHP in lubricant regenerative technology selection, Expert. Syst. Appl. 37 (1)

(2010) 419–425.[69] J.L. Funk, The future of mobile shopping: the interaction between lead users and technological trajectories in the Japanese market, Technol. Forecast Soc.

Change 74 (3) (2007) 341–356.

Professor Peng-Ting Chen received her B.S. degree in Chemical Engineering from Tunghai University, and Ph.D. degree in Technology Management from Univer-sity of National Chiao-Tung University, Taiwan, in 2001 and 2006, respectively. She is an assistant professor in the Department of Business Administration, I-ShouUniversity, Taiwan. Her current research interests include technology and business planning, strategies, and policies.

Hsin-Pei Hsieh received her B.S. degree from I-Shou University in Business Administration and MBA degree in the Postgraduate Programs in Management fromUniversity of I-Shou University in Taiwan in 1998 and 2000, respectively. Her major research interests fall in the areas of e-commerce.

557P.-T. Chen, H.-P. Hsieh / Technological Forecasting & Social Change 79 (2012) 543–557