Successful Strategies for yes.com.hk - A Case Study for Internet Advertising Business

R.C.T. Cheung [HREF20], Assistant Professor, Department of Computing [HREF21], Hong Kong Polytechnic University [HREF22], Hong Kong. csronnie@comp.polyu.edu.hk

P.P.F. Lo, S.I. Leong, Y.M. Law, C.M. Chau,  Hong Kong Cyber University, HKCyberU [HREF23], cyberu@comp.polyu.edu.hk

Abstract

yes.com.hk, a spin-off company of the publisher of the teen magazine "YES!", started its operation in mid 1996 as a vertical portal for teenagers in Hong Kong. The company has gone through several different business models from subscription-based to B2C and has decided to stay with the pure advertising-based model after the failures of previous business models. Nonetheless, the advertising model has suffered many setbacks after the recent dotcom crash and yes.com.hk's revenues have decreased since mid-2000. In this paper, we investigate several methods to improve revenue for yes.com.hk by considering the advertising strategies of yes.com.hk. Finally, we make several suggestions based on the results of our analysis.

Introduction

Traditional media companies-newspapers, magazines, TV and radio broadcasters-are looking for new ways to obtain more revenue out of the Internet. They are also particularly keen on developing Internet advertising business since their brands are already well known in local communities, and they have plenty of content that can be used on their Web pages to attract viewers. "Everyone in the broadcasting business is experimenting with the Internet to enhance their audience and explore new sources of improving their bottom lines," said Dennis Wharton, spokesman for the National Association of Broadcasters, which represents radio and television broadcasters [HREF18].

Nevertheless, one of the main challenges is finding the right approach with advertising. "Click-through" rates, or CTR in short, which measure the number of times viewers click onto a particular ad, have plummeted to an average of 0.30 percent from around 8 percent to 10 percent in 1996, according to industry sources [HREF18].

Background of yes.com.hk

In Hong Kong, the management of Yes Communication Ltd., the publisher of the popular teenager magazine "YES!", realized that the Internet was an attractive medium of information disclosure and a potential lucrative market on its own. For that reason, the company has launched in 1996 a subsidiary named yes.com.hk [HREF11], responsible for development of a companion web site of the magazine.

Teenagers - a profitable market

The success and popularity of the "YES!" magazine, which had a weekly circulation of about 92,000 in the first half of 2000, according to HKABC, is certainly a sound proof that the market dedicated to teenagers is real and profitable. 

Teens are often described as being heavy users of media and greatly influenced by images in the media, and thus comprise an appealing market to advertisers and marketers. However, selection of any particular medium by teens varies as a function of the limitations of each medium and the gratifications that are sought by youth. Both traditional and Internet advertising have its limitations of each medium. They make active choices about their media they use according to their personalities, socialization needs, and personal identification needs. In fact, teens already use the Internet as a conduit for social stimulation (Arnett, 2000).

The growing number of single parent families among our society leads to a side effect where teens tend to purchase household goods. In spite of this, special care must be taken to tackle this market because research has shown that teens are skeptical toward advertising and they are more apt at recognizing emotional advertising than previous generation (Manglebury, 2000). 

Teenagers on the Internet - previous research

The interactivity of the Internet provides teens an opportunity to communicate directly with advertisers. The Internet is able to combine several unique qualities of each medium (i.e. print, sound, and visual), which was not possible before. Chat rooms and newsgroups are replacing the traditional conversation among young people. 

The ability of the Internet to deliver and obtain information in a flexible, effective manner at relatively low costs is very attractive. However, not many research projects have focused on how teens use the Internet and the effects on the time spent with other media activities, as well as its ability to fulfill interpersonal communication needs. Advertisers are not sure about the impact of Internet and its relationship with teens. Logically, what are the most effective advertising strategies for advertiser to communicate with teens?

Previous business models of yes.com.hk

yes.com.hk's web site went online in August 1996 and began as a subscription-only content site. The response was not positive, with less than 1,000 subscribers. Therefore, the company abandoned this model in 1998 and provided the content free of charge to all Internet users, with some parts accessible only to registered users, which is also free.

The company has also tried B2C commerce in 1997 by selling online some of the magazines published by the parent company and artists-related items. As local people could buy these directly in the nearby newsstands, the customers were mainly from overseas, but the business was plagued by high delivery costs, usually 40-200% of the items being sold. Realizing that the online business was not likely to be profitable in the near future, in 1999, the management has decided to refocus its attention to the advertising business model and at the same time, keeping the spending under control because it simply makes no sense to spend millions and millions per month without any clear or confident way of making profits, whether it is e-business or not. The cash burn rate of yes.com.hk is less than half a million per month, which is relatively low when we consider the fact that, in February 2001, the average daily hit rate of the web site was 2 million and this figure is comparable to large portals like hongkong.com. 

Current pricing scheme

yes.com.hk currently charges advertisers a fixed monthly rate to display banner ads. The fee is negotiable, but usually it is related to the number of impressions. 

All the banners at yes.com.hk are random banners, independent of the section of the web site (ROS or run of site banners). Sometimes, yes.com.hk provides editorials that cover certain products at the advertiser's request. The company also offers a special package that combines both the offline print ads of "YES!" magazine with online banner ads, targeted for skeptics of Internet advertising.

Registered users

As of February 2001, the number of active registered users at yes.com.hk was about 288,000 with a male to female ratio of 55:45.

Year 2001 - the future of advertising model looks grim  - yes.com.hk, upon the decision of focusing mainly in Internet advertising, is facing many challenges ahead. The dotcom bubble outburst, combined with the slowdown of the global economy (Starling, 2000) [HREF15] has forced many companies to reduce their expenses and this includes ad spending. The average rate (of placing ROS banners) for yes.com.hk has decreased from HK$10,000 per month in early 2000 to HK$8,000 in 2001, a 20% decrease. 

The hope is not lost, however. Hong Kong was found to have the highest CTR in the world (Lai, 2001). According to a survey in 2000 (Lai, 2000), the industry average CTR in Hong Kong was 0.35% and the average CTR of vertical portals in Hong Kong: 0.79%. Comparing these figures to the average CTR of 1.23% of yes.com.hk in 2000, we can see that the company is way ahead of many portals in Hong Kong and the potential to earn profits is still very high.

Potential competitors

In the offline world, the magazine "YES!" has so far enjoyed a virtual monopoly in the market. Yes Communication Ltd. possesses a large library of copyrighted photos of famous artists and celebrities of Hong Kong, Taiwan and Japan that constitutes one of its major strengths and at the same time a tough barrier to entry of other potential players as well.

In the online world, as the cost of switching from one site to another is almost nil, any web site that appeals to teenagers in Hong Kong is a competitor. This fact has made the advertising business model not profitable up to now.

Research goals

Our research goal is very simple, yet difficult to have an immediate and definite answer: how to make Internet advertising business of yes.com.hk more attractive?

 

  Figure 1. Selected top 10 Banners

Research methodology

The adoption of successful advertising strategies will very likely help the company to reach profitability. We believe that a good strategy should lead to one of the following:

• Increase in the value of the advertising service, so that yes.com.hk can charge more for its services.
• Reduction of the seasonal factors that affect the revenues.
• Attract more people to the web site, making it more appealing to potential advertisers.

The methodology we describe here to increase the brand value of yes.com.hk is by studying the success factors of the advertising campaigns that the company has helped advertisers in the past. By pushing an ad strategy that encompasses these success factors to potential advertisers, it is likely to increase the probability of success of the campaign. The raise of success rate leads to a better brand name, allowing yes.com.hk to distinguish itself from other Internet websites in Hong Kong.

Table 1. Average click-through rates of top 10 banners

We have collected the top 10 banner ad campaigns (as shown in figure 1) of yes.com.hk from January 2000 to May 2001 and we have extracted the components of each promotion for analysis (as shown in table 1). The very definition of "success" varies from advertiser to advertiser and it is quite difficult to assess objectively whether a given component is a success factor or not. To amortize this problem, we have divided the measurement into two parts: a subjective measurement (Vs) by the advertiser and an objective measurement (Vo) obtained from the following formula to determine whether a campaign is successful or not:

                                                                    

where Cn denotes the normalized banner click-through rate and is calculated by:

                                                            

where Ct denotes the average click-through rate of day t, (t the ad exposure duration in days, Cy the general average click through rate of the year and St the seasonal compensation factor, to be explained in the later sections.

Both the values Vs, Vo are normalized to the scale of 0 to 10. The success degree V is then computed as follows:

We have decided to place more emphasis on the subjective measurement because we believe that customer satisfaction is more important than any other means of evaluation.

                                                                 
 

Why click-through rates?

Jakob Nielsen (Nielsen, 1997) believes that only loyal customers will provide revenue for the site and the majority of visitors (site tourists) will not contribute any revenue at all. Therefore, dotcoms should refuse to pay for eyeball measurements like Cost Per Thousand (CPM). 

Although many advertisers have found measuring the branding effects of banner ads difficult (Zaret, 2000) [HREF14], a banner click represents an active and positive acknowledgment of having viewed an ad and at least a momentary desire to further investigate the marketing message.

It is of our opinion that Click-Through Rates (CTR) is one step beyond simple eyeball measurements and can be easily traced. As such, CTR is used in this paper as an objective measurement of banner ad effectiveness.

Analysis of "success factors" 

yes.com.hk, being targeted mainly at teenagers, suffers from seasonal fluctuations in traffic due to exam periods and summer vacation. The measurement of success by direct comparison of the click-through rates in different periods would be unfair. As such, a seasonal compensation factor is introduced to make the comparisons more meaningful.

After calculating the success degree values, we have fed the values, along with chosen factors, to a data-mining tool called Clementine [HREF19] for analysis. By using the "Rule Induction" tool, many rules have been extracted. Most of them were irrelevant (confidence level too low) and the two most relevant rules are shown below:

Rule 1. Free gifts = Yes & Celebrities = No & Banner Design = Good => Success = Yes 
Rule 2. Game = Yes => Success = Yes 

The results (as shown in table 2) imply that the presence of a small online game is important to the success of the ad campaign. In other words, interactivity of a banner ad is a key factor #1. The success of the Wrigley's chewing gum promotion (Head, 2000) with the use of online games further highlights the importance of interactivity. The research firm KPE has coined the term advergaming for the mix of a small online game with ads (Olsen, 2001a).

Offering free gifts is the next important factor. However, to be effective, the free gifts must be of some value to users. Take the case of Red Earth, for example. Although it has offered 10% discount and free make-up products, the overall end-result was not satisfactory, because most customers have 20% discount with their Esprit cards and the free make-up products were not up to the most current trend. In other words, the need of vertical integration should not be ignored in marketing campaigns.

Banner design is also essential to attract visitors to click[HREF13], [HREF17]. Even when a user does not click on an attractive banner, anything that can capture attention contributes to the brand awareness.

The use of celebrities is not recommended due to the high costs and uncertainty in their effectiveness, unless the product being promoted is somehow related to the person/group (the case of Netvigator.com and Warner Music). yes.com.hk should, therefore, make recommendations to new advertisers based on the above-mentioned factors.

               Table 2. Source data used for rule extraction

Legend:

Game - Availability of an online game after banner click
Free gifts - Free gifts are offered along with the campaign
Celebrities - The use of stars, idols and famous celebrities in the ad
Banner Design - The quality of the banner ad design

Criteria for a good banner design :

Vivid colours
Attractive figures or graphics
Legible fonts 

Seasonal factors

According to Figure 2 and Table 3, we can see that the following periods have high click-through rates:

• February - Only the first half of the month got high CTR due to St. Valentine's Day.
• April - Easter holidays
• June to August - Summer holidays
• December - Christmas holidays

The low traffic periods are:

• January
• October to November
  
  Both these correspond to examination periods, when teenagers do not have enough spare time to spend with an entertainment site like yes.com.hk.
  
  Since unstable click-through rates will very likely affect the revenue of the company, yes.com.hk should find ways to fight against low click-through rates during the examination periods. Given that teenagers generally place higher priorities on studying over entertainment, yes.com.hk should consider adding new sections that provide exam tips, counseling, online research facilities that can help students to solve their problems in assignments and exams. In a sense, we suggest shifting between the usual emotional branding through idols that emphasizes fun and entertainment and rational branding that enforces a utilitarian view of the web site accordingly.
  

Table 3. Seasonal compensation factor

C_y = 1.233 (from Apr-00 to Mar-01)

Figure 2. Average click-through rate of yes.com.hk

Addressing the needs of teenagers

A study conducted in the U.S. by Ferle in fall 1997 (Ferle et al, 2000) has found that teenagers used the Internet to fulfill various needs shown in the following list (sorted by decreasing order of importance):

        1. Fun and games
        2. College/universities and homework
        3. Music
        4. Socialization (travel, making friends)
        5. Health
        6. Shopping

The "New GenerAsians" youth survey in 2000 by AcNielsen has found that ICQ was voted as the most popular tool by HK teenagers, which means that the socialization needs of teenagers in HK are ranked in a relatively high position. yes.com.hk until now addresses mostly the first point ("fun and games"). Internet music broadcasting was once attempted, but the high costs of licensing music have deterred the company from continuing this activity.

The web site has currently a section for making friends called "Love matching". However, the scope is rather limited.

There is still a large room for improvement to fill the gaps for points 2,4,5 and 6 above. Therefore, we recommend the addition of chat rooms and private e-mail boxes with online advisers that respond to questions of health and school related problems. As the Internet has a distinctive advantage of allowing anonymity that makes it easier for teens to ask embarrassing questions, we believe that incorporating such features will increase the attractiveness of the web site and the user loyalty.

Since the inclusion of shopping requires a new infrastructure that includes payment gateways, new pricing schemes and a redesign of the user interface, we suggest the postponement of this problem to a later stage.

yes.com.hk should also new types of banner ads, such as rich media banner ads. Yahoo!, the giant web portal, has recently pushed a new type of rich media ad that provides greater interactivity and better screen effects than simple animated GIFs (Olsen, 2001b). In addition, new banner ad sizes should be tested, according to the new rules that the Internet Advertising Bureau has recently issued to build a better "emotional" element into an Internet (Olsen, 2001c). 

Almost all the banner ads hosted by yes.com.hk are random. New packages that offer targeted ads based on the gender, sex or other preferences of registered members should be provided. By addressing the needs of different advertisers, yes.com.hk may be able to increase its customer base.

Conclusion

As stated in the beginning, the objective of this research was to identify the most effective Internet advertising strategy through banner advertisement for yes.com.hk. An effective strategy can proceed from different perspectives:

• Focus on the users' needs - diversify the content to improve the loyalty of users.
• Focus on the advertisers' needs - provide more flexibility and variety in advertising packages and introduce new pricing schemes.
• Minimize revenue fluctuations - different strategies should be employed for peak and non-peak seasons.
• Increase the success rate of ad campaigns at yes.com.hk

For the last point, several critical attributes have been identified which include Game, Free Gifts, Celebrities, Banner Design and they were carefully studied. These attributes were selected out from the top 10 banners from January 2000 until May 2001. Data extracted from the existing web log was processed by the data-mining tool Clementine for analysis.

The research was important because it sought to determine the important attributes that can increase the time spent by teenagers with the ads. The success of banner advertising will prove the legitimacy of the advertising model of yes.com.hk (as shown in table 4).

Based upon the findings, it was concluded that:

• Small interactive games and free gifts can provide the highest brand impression after normalization of the banner click-through rate.
• The revenue of yes.com.hk was influenced by the seasonal factors such as exam periods and summer vacation.
• Media rich banner design is important to attract teenagers to click on the banner. It has to include elements such as video, visual and graphic. 
  

Recommendations

Based on the conclusions, the following recommendations are made:

• Although the attributes that can attract teenagers to yes.com.hk have been identified, ongoing monitoring of user behavior on yes.com.hk is required to understand their navigational behaviors.
  
• The use of gaming and free gift is encouraged because they have been proven to be effective in terms of impression and click-through rates.
  
• The pricing scheme for yes.com.hk should shift from fixed monthly rate to adjusted monthly rate according to the level of customization, banner ad size, interactivity, and the season. For example, for non-exam periods, yes.com.hk should charge for a higher price due to a higher traffic volume.
  

Future Directions

The followings are the future directions of research relating to yes.com.hk :

• Research on the impact of co-branding on the impression and click-through rates. We have found that for the marketing campaign of Konica, co-branding had been used. Konica co-operated with Honey Bee which is a new brand developed by YES!. "Honey Bee" was welcomed by the teenagers and generated good sales volume. In this campaign, customers buying a package of Konica film could buy the new model Honey Bee handbag at special price. The response of this campaign was very good. At the third day of the promotion, nearly one thousand handbags were sold by the redemption banner. This shows that co-branding can attract more traffic.
  
• Study the conversion rate from Click to Sales. The lack of standards in measuring the banner ad to sales will jeopardize the revenue stream for dotcoms who heavily rely on advertising revenue for survival.

   Table 4.  Analysis of Top 10 Advertising Campaign and the Overall Result 

References

Arnett, J. (2000). "Adolescents' Use of Media for Self-Socalization." Journal of Youth and Adolescence 24, 5 (1995); pp. 519-32.

Chang, H.C.  et al.  (2000). The Impact of Banner Exposure and Clicking on Attitude Change, University of Nebraska-Lincoln April 2000 [HREF1]. 

Donatello, M. (1997). How Do I Click Thee? Let Me Count The Ways, NAA Marketscope, 1997 [HREF 24].

Ferle, C.L., Edwards, S.M., Lee, W.M. (2000). Teens' use of Traditional Media and the Internet, Journal of Advertising Research May-June 2000 

Head, O. (2000). ONLINE PROMOTION: Campaign of the month, Revolution Asia October 2000 [HREF2]. 

Lai, W. (2000). Vertical portals offer higher click-through rate, ComputerWorld June 7, 2000 [HREF3]. 

Lai, W. (2001). HK has highest banner ad click rate, ComputerWorld HK, April 13, 2001 [HREF4].

Manglebury, T.T.  and Bristol, T. (2000). "Socialization and Adolescents' Skepticism toward Advertising", Journal of Advertising (Volume 27, Number 3, Spring): pp. 11-20.

Nielsen, J. (1997).  "Why Advertising Doesn't Work on the Web ." , useit.com, September 1, 1997 [HREF5].

Olsen, S. (2001a). Major brands play for attention, CNET News.com May 3, 2001 [HREF6].

Olsen, S. (2001b). Net ad industry guns for new rules, CNET News.com May 7, 2001 [HREF7].

Olsen, S. (2001c). Group to issue standards for bigger Web ad, CNET News.com February 25, 2001 [HREF8].

Starling, A. (2000). Dotcom Bubble - "The Emperor Has No Clothes!", Web Developers Journal June 2000 [HREF9].

Zaret, E. (2000). The click-through conundrum, MSNBC September 25, 2000 [HREF10]. 

Hypertext References

HREF1

http://www.ciadvertising.org/studies/reports/info_process/aaa20Cho&Leckenby.html

HREF2

http://www.revolutionasia.com:8080/news/view.cfm?r=4&id=41728

HREF3

http://www.security-informer.com/english/crd_through_443115.html

HREF4
         http://www.idg.com.hk/cw/readstory.asp?aid=20010410004
HREF5

http://www.useit.com/alertbox/9709a.html

HREF6
          http://news.cnet.com/news/0-1005-200-5812147.html

HREF7

http://news.cnet.com/news/0-1005-200-5848747.html

HREF8

http://news.cnet.com/news/0-1005-200-4942655.html

HREF9

http://www.webdevelopersjournal.com/columns/ajs_bubble.html

HREF10

http://www.msnbc.com/news/467160.asp

HREF11

http://yes.com.hk

HREF12

http://www.tnbt.com/jsp/TNIP.jsp?x=3&y=3&p=TNNewsStory.jsp&a=68363

HREF13

http://www.marketers-reference.com/articles/BannerTenTips.html

HREF14

http://www.marketers-reference.com/articles/BannerTenTips.html

HREF15

http://www.financialexpress.com/fe/daily/20001219/efe19006.html

HREF16

http://www.mbinteractive.com/resources/reports/wired_rich_media.htm  

HREF17

http://www.bannertips.com/bannerdesign.shtml

HREF18

http://www.wilsonweb.com/webmarket/banner-design.htm  

HREF19

http://www.spss.com/clementine/  

HREF20

http://www.comp.polyu.edu.hk/~csronnie

HREF21

http://www.comp.polyu.edu.hk

HREF22

http://www.polyu.edu.hk

HREF23

http://hkcyberu.com

HREF24

http://www.naa.org/marketscope/research/newmedia/archive/Click1.html

Copyright

Andrew Treloar, © 2000. The authors assign to Southern Cross University and other educational and non-profit institutions a non-exclusive licence to use this document for personal use and in courses of instruction provided that the article is used in full and this copyright statement is reproduced. The authors also grant a non-exclusive licence to Southern Cross University to publish this document in full on the World Wide Web and on CD-ROM and in printed form with the conference papers and for the document to be published on mirrors on the World Wide Web.