INTRODUCTION

Voted capital of the Middle East by The Independent, the emirate of Dubai made a superlative entrance into the twenty-first century (Macneil, 2007). The luxury capital of tourism, major events sponsor, architectural giant, and regional business, media, and shipping hub, Dubai has become a world contender and media magnet in many arenas. As a result, it has built what Govers and Go (2009) described as a ‘de facto place brand’, and what Bagaeen (2007) described as an ‘instantly recognizable’ cityscape around the world.

Along with these achievements came less desirable attention. Labor disputes, sex scandals, terrorism concerns and real estate busts have also grabbed headlines as Dubai sought to build its global reputation. Dubai also garnered seemingly paradoxical descriptions of its media presence. Lee and Jain (2009) described ‘an enormous amount of global exposure on television and in the press’ (p. 243), while also considering ‘limited media coverage’ (p. 236) a major threat.

To help clarify these issues, this case study explicates Dubai's rise to media prominence and assesses awareness of Dubai media images among three diverse stakeholders: professionals from New York, London and Singapore. It begins by tracking Dubai's headline projection growth in New York, London and Singapore since 1992, and then examines the stakeholders’ exposure to various information sources. Because information-source data hold implications for marketing budget allocations, the 2009 data collected in this study offered an updated alternate approach to the 2004 data from Govers et al (2007). The subsequent comparison of professionals’ perceptions with sample headlines sought to move research beyond the traditional emphasis on perceived image over projected image (Pike, 2002).

For the purpose of this article, perception refers to held images, whereas projection denotes conveyed images (for example, media mentions). Although media projections may arguably reflect media's own perceptions, such inquiry is beyond this study's aims.

LITERATURE REVIEW

Reviewed literature on projection and perception as it pertains to place branding reveals a scarce amount of research and hard data on Dubai branding, an excessive focus on local brand identity, projection and image, and an emphasis on generic North American and Western European populations.

Research in projection and perception

Historically, place image research has been set within a tourist destination image context. Whereas place branding has moved beyond mere tourism, destination image research still provides a useful starting point for exploring projection and perception research. Destination image studies have favored perceived image over projected image (Pike, 2002), and multiple studies have sought to tie perception to information sources (Govers et al, 2007), marketing strategy (Andreu et al, 2000), tourism operators (Grosspietsch, 2006), marketing strategy and news media (Saraniemi, 2009) and popular media (Mercille, 2005).

Unfortunately, the projection–perception link has not been operationalized clearly, wrongly equating strategy or tourism operator perception with actual projection (Andreu et al, 2000; Grosspietsch, 2006) or providing data and not tying it to perceived image (Mercille, 2005; Saraniemi, 2009). The one exception, Govers et al (2007) limited the connection to source exposure, rather than actual image projections.

Asserting a critical relationship between projection and perception, Andreu et al (2000) compared tourism strategy with potential tourists’ perceptions. The study reviewed Spain's tourism communication strategies, comparing them with image perceptions of 120 visitors to the UK's Spanish Tourism Office. Noting improved message coordination and alignment with tourists’ preferred communication channel from the 1980s to the 1990s, the subsequent interviews did little to tie perception to the actual projections in question.

Also focused on developing effective marketing strategy, Grosspietsch (2006) sought to compare tourist perceptions and tour operator projections of Rwanda. Both parties completed the same survey on their perceptions of Rwanda revealing that tourists generally possessed a better image of Rwanda than did tour operators. The study did not explore actual operator projections, implying synonymy between them.

Looking at Finland, Saraniemi (2009) described the country's branding efforts, brand projection, and brand perception through qualitative data analysis of internal documents and interviews, press articles and photos, and interviews of a random sample of European nationals. The substantial data yielded less than two pages of qualitative description that did not connect projection and perception.

Utilizing media effects theory, Mercille (2005) investigated first-time visitors’ exposure to two films, a magazine, and a guidebook compared with their perceptions of Tibet before and after their visit. Mercille argued that these sources provided either a strong or a negotiated influence on perception depending on socio-cultural background. However, media familiarity ranged from 10 to 43 per cent per source, and the study identified neither how many were unfamiliar with any of these sources, nor their familiarity with other sources. In addition, the study omitted the post-arrival and socio-cultural data needed to justify the conclusions.

Exploring media effects to assess marketing communication budgets, Govers et al (2007) investigated tourism promotion's effect on destination image. The study utilized Gartner's (1994) framework to analyze 1100 global participants’ information sources and perceptions of seven locations. Covert and autonomous agents (news and popular media) played a primary role in image formation, followed by organic agents (personal, friend and relative experience). The weaker presence of Internet and overt agents (advertising) called into question their relative importance in advertising budgets and marketing theories. However, the study did not indicate the extent of advertising presence in the participants’ diverse geographic and demographic markets.

Brand Dubai projected image

Two studies and one index have examined Dubai's publically projected image, addressing Internet depictions (Govers and Go, 2005), regional media coverage (Walters, 2006) and global media coverage (East West Communications, n.d.).

Govers and Go (2005) conducted online photo and descriptive text content analysis across 20 websites. The hundreds of images and nearly ten thousand words revealed a tension in Dubai's projected image, as public entities’ cultural identity projections contrasted with the private sector's commerce portrayals. Overall, image projection leaned considerably toward consumption rather than experience, possibly leading potential tourists to misperceive and subsequently misrepresent the experience.

Focused exclusively on local and regional English media coverage, Walters (2006) described Dubai's government-controlled image projection as hyperbolic superlative cheerleading printed directly from press releases. Known threats such as overstretched infrastructure and inflationary pressures were almost entirely absent in reporting.

Providing media ranking of 200 countries and territories since April 2008, the Nation Brand Perception Index (NBPI) has examined volume and favorability of coverage in 38 newspapers, magazines and television news broadcasts (East West Communications, n.d.). While named a perception index, the NBPI actually ranks projection among English-language media and English-language translations among predominantly US and UK media sources. With Dubai acting as the primary headline grabber, the UAE ranked third to 17th for the six quarters between April 2008 and September 2009, shifting dramatically to 194th in the final quarter of 2009.

Brand Dubai perceived image

The least studied of the three topic areas, perceived image of Dubai, has received the attention of only one academic study (Govers et al, 2007). Investigating tourism promotions’ role in destination image formation, Govers et al utilized a discursive approach and neural network content analysis to assess perceived images and their origins. The 598 highly educated and mobile global survey respondents most frequently identified television, previous travel and friends as Dubai information sources, with no source cited by more than one-quarter of participants. Many respondents had difficulty describing Dubai, and common images included negative regional stereotypes. Dubai-specific images highlighted ‘a modern wealthy city providing luxurious facilities as a sea, sun and sand destination that combines leisure with extensive shopping facilities’ (p. 251). A revealing and innovative approach, the study, however, did not quantify the occurrence of those who provided Dubai-specific descriptions.

In sum, apart from Govers and Go (2009), published material on Brand Dubai projections and perceptions has lacked hard data and focused excessively on local brand identity, projection and image, presuming a parallel with global perception. In addition, the heavy index emphasis on North American and Western European populations neglected Asia's importance in the Dubai market, whether for tourism, investment or immigration. Therefore, the current study will seek to add an updated layer to the analysis of global projections and perceptions of Dubai's competitive identity by answering the following research questions:

RQ1:

  • How did media coverage of Dubai change between 1992 and 2009?

RQ2:

  • What are Dubai's information sources?

RQ3:

  • Using Gartner's framework, what are Dubai's information sources?

RQ4:

  • Using Gartner's framework, how do Dubai's information sources differ by city?

RQ5:

  • How do Dubai's autonomous news sources differ by city?

RQ6:

  • How do Dubai's overt sources differ by city?

RQ7:

  • What are number of Dubai's overt sources across the three cities?

RQ8:

  • What is the relationship between Dubai's overt sources and organic experience?

RQ9:

  • What are Dubai's primary knowledge sources overall and by city?

RQ10:

  • How consistent are the perceptions of participants with Dubai's media portrayal overall?

RQ11:

  • How consistent are the perceptions of participants from each city with Dubai's media portrayal?

METHODOLOGY

To addresses both projected and perceived images, this case study utilized multi-method quantitative research, including headline analysis and questionnaires aiming for more accurate analysis than the single factor analysis historically employed by destination image researchers (Prebensen, 2007). Focused on potential stakeholders, it utilized a specific target population in three target contexts (Schensul et al, 1999; Anholt, 2002). In a two-step process, the research first employed quantitative analysis of newspaper headlines in these contexts, followed by survey data collection among a sample of the target populations (Brunt, 1998; Saraniemi, 2009). Although place image surveys have been common, this dynamic projection–perception model fills an important methodological gap (Pike, 2002; Rindell, 2007).

The study took place in New York, London and Singapore, three global cities serving as models, trade partners, allies and talent pools for Dubai (Anholt, 2006). The New York Times, The Times of London and The Straits Times of Singapore provided sample headlines published 12 months before the survey. Because of Dubai's competitive identity as a luxury business and tourism destination with a substantial white-collar expatriate workforce, professionals in the three cities served as the potential stakeholder target population.

Headline analysis

Temporal mass media portrayals affect distant locations’ perceived brand images (Anholt, 2004; Rindell, 2007; Avraham and Ketter, 2008; Saraniemi, 2009). Within these media portrayals, headlines establish significance and provide optimal relevance in their target audiences’ perception development, even if the audience hardly recalls the actual projection (Beasley and Danesi, 2002; Danesi, 2002; Dor, 2003). Consequently, this study examined recent headlines for the three target locations. Lexis-Nexis Academic enabled identification of Dubai headlines in each city back to 1992, allowing the study of historical headline growth. In addition, the study utilized a sample of 12 headlines to assess participants’ headline familiarity, drawing from the 12-month period of 24 February 2008 to 23 February 2009.

Surveys

As the most commonly used method of place image inquiry, a survey of potential stakeholders offered examination of information sources and headline familiarity (Brunt, 1998). The self-administered questionnaire provided insight into personal background, Dubai information sources and headline image familiarity via 48 questions (43 closed-ended and five open-ended). The questionnaire inquired about gender, age, citizenship and international travel frequency and destination. In addition, it utilized options that fit within Gartner's (1994) communication channel categories to address participants’ sources of information on Dubai. It also assessed headline image familiarity utilizing the 12 headlines mentioned in the preceding paragraph.

To provide context, the authors adjusted six of the original headlines. For example, the headline ‘Dubai Firm Buys Auction House’, which came from the sports section, was changed to ‘Dubai Firm Buys Racehorse Auction House’, so participants would not interpret it as a different type of auction house. The six adjusted headlines included three text insertions, two deletions and one substitution. Participants indentified whether media projections matched their Dubai perceptions, indicating whether the content of each of the 12 headlines was consistent or inconsistent with their overall impression of Dubai.

Participants

The study purposively sampled professionals from London, New York and Singapore, identifying their importance as prospective Dubai brand stakeholders (Warwick, 1975). As a target sample, this group's intentionally narrow and geographically specific target fits well with Dubai's luxury business and pleasure tourism and immigration profile (Watters and Biernacki, 1989, p. 430; Schensul et al, 1999). While the results are not generalizable to the population at large, they may be generalized to the target population.

Overall, 121 professionals participated in the study, including 56 participants from 14 Toastmasters International clubs and 65 participants identified through the authors’ LinkedIn network. Toastmasters clubs included businesses in consulting, finance and banking, whereas LinkedIn contacts consisted primarily of doctors, bankers, professors, business persons and engineers. Four questionnaires were excluded due to unusable data. In total, the study utilized 117 surveys from Singapore (44), New York (40) and London (33).

Of the 117 participants, 38 per cent (44) were male, 48 per cent (56) female and 15 per cent (17) did not specify gender. Seventy per cent of the professionals were aged 40 or younger, whereas 27 per cent fell between the ages of 41 and 60. Expatriates totaled 43 participants (37 per cent). All but three of the participants had traveled abroad since 1999. Since 2004, 92 per cent had vacationed abroad at least once, and 47 per cent had done so at least six times. Fifty-six per cent had taken an international business trip since 2004, while 18 per cent had visited Dubai since 1999.

Singapore participants had the highest percentage of males (43 per cent) and expatriates (52 per cent) and made up the oldest group (39 per cent over 40). They traveled internationally the most frequently for business, with 25 per cent doing so at least 10 times since 2004. In addition, 95 per cent had travelled internationally for vacation since 2004, and 14 per cent had been to Dubai since 1999. New York participants had the most females (58 per cent) and nationals (88 per cent) and were the youngest (78 per cent under 40). They were the least frequent international travelers, whether for business or leisure, with only 8 per cent having visited Dubai. London participants fell demographically in between the other two cities, with the notable exception that 36 per cent had visited Dubai.

Process

Utilized for hard-to-reach populations, this research followed Watters and Biernacki's (1989) four-step target sample acquisition process. The study sought participants in locations with a high presence of globally oriented professionals, such as Manhattan in New York and London's Canary Wharf. After contacting 52 corporate-based and professional Toastmasters International clubs, 14 clubs returned 56 completed surveys. Utilizing what Biernacki and Waldorf (1981) refer to as chain referral sampling, the authors also received 14 connections from nine LinkedIn contacts, yielding 65 participants. Surveys were collected between 29 May and 1 November 2009, 5 days before international media began extensive coverage of Dubai's economic troubles.

Analysis covered overall survey results, as well as comparison between cities, using frequency and cross-tabulation inquiries, in addition to a Cramer's V correlation test and a χ2 significance test. Following Govers et al (2007), the study coded information sources according to Gartner's (1994) framework of covert (sponsored media), autonomous (independent media), organic (personal, friend and relative experience) and overt (advertising) agents. The authors expressed response volume using percentages, with all results described as statistically significant holding P values of <0.05 (P⩽0.05).

RESULTS

RQ1:

  • Media Coverage of Dubai Between 1992 and 2009.

Media coverage of Dubai grew significantly between 1992 and 2009 in all three studied newspapers (Table 1). During the period of 1992 and 1994, Dubai received no headlines in the New York Times, three headlines in the Straits Times and 15 headlines in the Times (London). In contrast, the coverage increased during the period of 2007 and 2009 to 57 headlines in the Straits Times, 104 in the New York Times, and 271 in the Times. Percentagewise, the coverage increased between the periods of 1995–1997 and 2007–2009 by 2500 per cent in the New York Times, 613 per cent in the Straits Times and 868 per cent in the Times. In just 1 year – between 24 February 2008, and 23 February 2009, Dubai picked up 113 headlines in the three newspapers: The Straits Times (14), the New York Times (23) and the Times (76).

RQ2:

  • Dubai Information Sources

Table 1 Media coverage of Dubai (1992–2009)

The main Dubai information sources participants cited were friends, relatives or coworkers (71 per cent), newspaper articles (67 per cent) and TV news (60 per cent) (Table 2). Less than half of the participants identified other categories as a source of information on Dubai. These included Internet news (40 per cent), newspaper ads (40 per cent), magazine ads (33 per cent) travel TV programs (31 per cent), TV ads (30 per cent), travel magazine articles (29 per cent), magazine news (28 per cent), Internet ads (23 per cent), geography lessons (22 per cent), travel websites (16 per cent), movies (16 per cent) or personal experience (19 per cent). The least common sources cited were guidebooks (9 per cent), blogs (8 per cent) and radio ads (4 per cent). Eighteen per cent of participants wrote Internet or Google as an open-ended response, while 13 per cent identified a varied collection of other sources.

RQ3:

  • Dubai Information Sources Using Gartner's Framework

Table 2 Information sources (all cities)

Table 3 categorizes the information sources according to Gartner's (1994) framework. It shows that autonomous news sources (76 per cent) surpassed organic (un)solicited sources (71 per cent) as the most common Dubai information sources, while a majority cited overt sources (52 per cent). Covert sources (44 per cent) and autonomous (other) sources (36 per cent) followed, and organic sources (19 per cent) were the least common.

RQ4:

  • Dubai Information Sources by City Using Gartner's Framework

Table 3 Information sources (all cities) grouped by Gartner's framework

When comparing across the three cities, New York (93 per cent) and Singapore (80 per cent) participants cited autonomous news sources the most, whereas London (88 per cent) participants cited organic (un)solicited sources the most (Table 4). For London participants, autonomous news sources came in second (82 per cent), while overt (67 per cent) and covert (58 per cent) sources followed. The second most common sources for Singapore participants were overt (77 per cent) followed by organic (un)solicited (59 per cent). As for New York participants, organic (un)solicited (70 per cent) came in second, followed by overt and covert sources, both at 43 per cent. The starkest difference between locations occurred in the overt category with the New York responses (25 per cent) contrasting dramatically with those of London (67 per cent) and Singapore (77 per cent).

RQ5:

  • Dubai Autonomous News Sources by City

Table 4 Information sources by Gartner framework (by city)

Within the autonomous news category, New York participants cited almost all news sources more frequently than did Singapore and London respondents (Table 5). Significantly more New York participants (53 per cent) selected Internet news compare with participants from Singapore (36 per cent) and London (30 per cent). In addition, London professionals (18 per cent) were less likely than Singapore (32 per cent) or New York (33 per cent) participants to cite magazine news.

RQ6:

  • Dubai Overt Sources by City

Table 5 Autonomous news sources (by city)

As for overt information sources, London and Singapore participants selected these sources relatively equally, whereas New York professionals were much less likely to select any of them (Table 6). For instance, only 20 per cent from New York cited newspaper ads, whereas 48 per cent from Singapore and 55 per cent from London did so. In addition, whereas 23 per cent of Singapore participants and 39 per cent of London respondents cited Internet ads, only 10 per cent of New York participants did the same.

RQ7:

  • Number of Dubai Overt Sources Across the Three Cities

Table 6 Overt sources (by city)

When analyzing the number of overt sources for each city, participants from London had the most number of sources compared with those from Singapore and New York (Table 7). While 68 per cent of the London participants had between three to five sources, only 27 per cent from Singapore and 36 per cent from New York had the same. On the other hand, 74 per cent of Singapore participants and 64 per cent of New York participants had only one or two sources, while 32 per cent from London had the same.

RQ8:

  • Relation Between Dubai Overt sources and Organic Experience

Table 7 Number of overt sources (by city)

While information sources varied by location, overt sources correlated with organic sources, independent of location (Table 8). Correlating personal experience to the number of ads resulted in a Cramer's V value of 0.471 (P<0.001), representing a moderate association between the two variables (Field, 2005, p. 693). In addition, among participants who had not been to Dubai, 48 per cent cited no exposure to ads (overt sources) and 26 per cent cited exposure to one advertisement source. By comparison, 60 per cent of those who had visited Dubai cited exposure to three to five ad sources. Thus, London participants’ tendency to have a higher amount of advertising exposure is not necessarily the result of location, but rather the result of being nearly three times as likely as Singapore participants to have visited Dubai.

RQ9:

  • Dubai Primary Knowledge Source by City

Table 8 Correlation of number of overt sources with organic experience

In addition to identifying all sources of brand Dubai image influence, participants revealed their primary source of information (Table 9). Overall, autonomous sources ranked the highest with 44 per cent of all participants selecting at least one source in this category. This was followed by organic (un)solicited sources at 27 per cent, organic sources at 11 per cent, overt sources at 11 per cent, covert sources at 6 per cent and autonomous (other) sources at 2 per cent.

Table 9 Primary knowledge source (by city)

When comparing across cities, autonomous news sources were the most important for both New York (67 per cent) and Singapore (44 per cent) professionals, while primary for only a few London participants (17 per cent). London participants (47 per cent) noted organic (un)solicited sources as their most important, significantly more than did those from Singapore (18 per cent) and New York (19 per cent), though it was New York's second most important source. London professionals also identified the highest percentage of organic sources (20 per cent), whereas Singapore respondents’ second most important were overt sources (21 per cent). Cited by no more than 7 per cent of all respondents, Covert and autonomous (other) sources lacked importance in all three cities.

RQ10:

  • Perceptions of Dubai's Media Portrayal

Surveyed potential stakeholders viewed Dubai's media portrayal as largely consistent with their impressions. The majority of participants viewed 11 out of 12 headlines as consistent, with the highest percentages pertaining to the most positive and glamorous headlines (Table 10). Over 80 per cent of participants identified glamorous lifestyle headlines such as shiny new malls (82 per cent), big is best (86 per cent) and rich and famous (89 per cent) as consistent with their Dubai image. Dubai's architectural identity was familiar to 76 per cent of respondents. Headlines of Dubai's investment in the east (62 per cent), its plans to build a theme park (61 per cent), its attraction of big sports names (59 per cent) were slightly less familiar, though still consistent with a majority of participants. So were headlines that showed businesses moving to Dubai (58 per cent) and a Dubai firm buying a racehorse auction house (55 per cent). Simultaneously, two somewhat negative headlines, one that portrayed foreign workers fleeing Dubai due to an economic downturn, and another about a Briton facing jail time for adultery, garnered substantial familiarity at 57 and 51 per cent, respectively. However, only 30 per cent of participants regarded as consistent a headline about Dubai barring entrance to an Israeli tennis player.

RQ11:

  • Difference in Perceptions of Dubai's Media Portrayal by City

Table 10 Volume of perceived headline consistency (all cities)

Professionals across the three cities cited minimal variation in consistency (Table 11). Using χ2 as an indicator of significant difference between the countries, only two headlines returned a P value <0.01, the headline pertaining to Dubai's investment looking east and the headline about the city luring big sports names. For the first headline, New York participants ranked highest in consistency at 68 per cent, followed by London at 61 per cent and Singapore at 59 per cent. As for the latter headline, New Yorkers registered the lowest consistency rate at 35 per cent, whereas Singapore ranked the highest at 73 per cent, followed by London at 70 per cent. For both headlines, participants from Singapore and London registered close results and differed significantly from their New York peers.

Table 11 Variation of headline consistency (by city)

DISCUSSION

Coverage

Studies to date have assumed that because Dubai's branding efforts and internal identity provide a strong de facto place brand, they automatically match the Emirate's global projections and perceptions. However, Lee and Jain (2009) viewed limited media coverage as a major threat to the city's brand, contrasting with NBPI's high ranking of Dubai media coverage for most of 2008 and 2009. In addition, Govers et al (2007) noted that as of 2004, the global travel community often held non-specific views of Dubai that additionally contained inaccurate regional stereotypes. The present study described brand Dubai's robust growing coverage in three global media outlets – The New York Times, The Times of London and The Straits Times of Singapore – and the role of this media coverage in building brand image among a sample of professionals in Singapore, New York and London.

Sources

This study found that personal relations, newspaper articles and TV news were the three most prevalent information sources about Dubai among the surveyed participants, while the majority of participants also reported familiarity with at least one form of news and one form of advertising. These findings represent higher response rates than those collected by Govers et al (2007), for whom only 24 per cent of participants cited friends (compared with 71 per cent in this study), 5 per cent cited newspapers (compared with 67 per cent) and 6 per cent cited all forms of advertisement (compared with 52 per cent). However, this study's more targeted sample population and its closed-ended questions may account for some of the differences – compared with the open-ended responses collected by Govers et al. Nonetheless, the differences paralleled both media coverage growth between 2004 and 2009 and a 43 per cent compound annual growth rate of foreign tourists between 2003 and 2008 (Euromonitor International, 2009).

This difference in findings and approach offered two caveats to Govers et al's (2007) assertion that the Internet and advertising may have received more marketing consideration than merited. Like Govers et al, this study found that non-sponsored media, such as news and personal relations, proved to be the most common information sources when assessing all sources for combined populations. However, participant location, relative source importance and variety of source exposure warranted notes of caution in utilizing aggregated global results for marketing and advertising strategy decisions in local markets.

First, location highly influenced the relative importance of source type. Matching the findings of Govers et al (2007), organic sources mattered most to London participants, befitting Britain's position as Dubai's number one inbound leisure and business tourism market (Euromonitor International, 2009). The more distant New York participants counted news as their top information source, as did a much smaller number of Singapore participants. However, the majority of New York participants cited the Internet as a news source, and a substantial minority of those from Singapore identified advertisements as their primary source. Thus, location-specific importance of sources merits attention. In addition, connecting these top information sources, such as placing Dubai banner ads on a newspaper website, may result in powerful, image-building synergy. While less prominent than other sources, use of the Internet may indicate those actively searching for information. Rather than indicating less importance, its presence may confirm its value as an easily accessible tool at the information search stage of decision making (Peter and Donnelly, 2007).

Second, participants demonstrated a moderate association between visiting Dubai and advertisement awareness that was independent of location. While the study did not establish causality, these ads likely held importance in either attracting people to Dubai, affirming the worth of their previous Dubai experience, or a mixture of both. Thus, advertising budgets could be justified by either directly gaining tourists – new or repeat – or by persuading previous tourists to become information sources for their friends, relatives and coworkers.

Finally, budget assessments should consider both the breadth and depth of advertising penetration. By breadth, Singapore respondents (77 per cent) were more familiar with Dubai advertising than were those in London (67 per cent) (see Table 4). By depth, however, London participants were more familiar with any given category of advertising (see Table 6). In addition, among those with advertising exposure, 68 per cent of London participants had been exposed to three to five sources, compared with only 27 per cent for Singapore (see Table 7). While experience in Dubai impacts this finding as discussed above, considering both breadth and depth remains important.

Consistency

Concerning headline responses, the majority of participants considered all but one media depiction of Dubai as consistent with their overall impressions. Among these, lifestyle-oriented images of architecture, shopping malls, beautiful desert, wealth, luxury and modernity fit the impressions of over 75 per cent of participants. These results generally matched the de facto image assumed in previous studies (Bagaeen, 2007; Balakrishnan, 2008; Govers and Go, 2009; Lee and Jain, 2009). Although it garnered high levels of consistency among this study's participants, architecture was not as ubiquitously perceived as Bagaeen presumed, with a notably smaller majority aware of it in Singapore. Considering the targeted nature of the current study, this example and others should temper assumptions that Dubai image familiarity is universal (Kotler and Gertner, 2004; Olins, 2004), although the opening of the Burj Khalifa tower after this study was conducted also merits investigation into its role in promoting Dubai's architecture and other images of opulence and grandeur.

Tourism image familiarity contrasted with the less familiar but more frequently projected business image of Dubai. The lower awareness is unsurprising given participants’ much more frequent international holiday travel compared with business trips. In particular, London participants, who make up the largest Dubai expatriate population among the three compared cities, expressed the greatest business image familiarity.

Overall, Dubai's efforts serve as a model to other relatively unknown places seeking to build global awareness and benefit from increased global engagement. Dubai's increasing news media projections were well known among global professionals in Singapore, New York and London. Fortunately for Dubai, negative images were largely absent from the public consciousness. Given the highly adverse fourth quarter media coverage in 2009, however, further research could shed a helpful light on whether such repeated headlines may have had an enduring effect, or merely served as a temporary inconvenience.

In addition, future research could explore the connection between Dubai media coverage, information sources, and image consistency in additional locations. Important and emerging markets such as India, Germany, Russia, France, China, Pakistan, Turkey and Japan would provide a broader understanding, as well as analysis in places where the English press holds less importance. Assessing more peripheral markets such as Mexico, Brazil and Spain would also allow assessment in places where the brand may be in relative infancy.

CONCLUSION

This study established evidence of robust growing media coverage of Dubai in the past two decades. It found that the lifestyle-oriented images of magnificent opulence, luxury and modernity that predominated the media coverage of Dubai resonated with the Dubai images held by the majority of participants. In addition, the study identified personal relations, newspaper articles and TV news as the three most prevalent Dubai information sources for the aggregate sample. However, source type differed by participants’ locations, as did breadth and depth of advertising familiarity. Furthermore, participants who had visited Dubai in the past had a stronger awareness of advertisement sources than those who had never visited the city.

These findings suggest significant implications on future Dubai branding efforts and place branding efforts, in general. First, the findings discourage the use of aggregated global data for marketing decisions and justifying advertising budgets. Instead, tying specific media channels to the location of the target audience may offer better outcomes. In the case of Dubai, whereas aiming for organic sources (personal, friends and relative experience) will be more effective for London audiences, targeting news channels will make more sense for those in New York and Singapore. Second, the evidence suggests that former visitors of a place are effective information sources for others. The higher level of advertising awareness among Dubai visitors compared with non-visitors indicates that advertising campaigns should target previous visitors, persuading them both to revisit the city and to become brand ambassadors among their circles and contacts. Third, the relative variance in image consistency across the targeted population tempers assumed universality of Dubai's image and most probably that of other place brands. This implies investing further in building a more consistent global image for brand attributes that exhibit universally shared appeal, while building a target market specific image for areas where the appeal is uniquely local. Finally, further and continued research to determine the effects of the major recent events, such as the opening of Burj Khalifa, the recent economic crisis that plagued the small emirate, and the popular uprisings in the region, should offer more insight into developing Dubai brand images, whether negative or positive.