Advertising Strategy for Blu Homes Makoto Kondo and Sinit Vitavasiri Blu Homes, an innovative housing company based in the Boston, is looking forward to having the opportunities to expand its customer base on the West Coast. The company sells durable, energy-efficient prefabrication homes and has several million dollars of revenue annually. Blu Homes has a concrete mission to build super-efficient prefabrication green homes by leveraging their leading-edge technology. Objectives The primary purpose of Blu Homes for this Internet advertising campaign is to learn about the market demographics on the West Coast, about in which specific areas in California most people are interested in buying their products, and about the effective advertising strategies that would attract potential customers’ attention, including what type of keywords that they will be attracted by. Blu Homes would specify how they should leverage their money as their investment strategy and how they implement their marketing strategy for the upcoming 6-12 months, based on the advertising campaign result. There are two important metrics for the advertising campaign. The success of the campaign depends on how much Blu Homes has learned about their potential customer base in California. The number of people who fill out the contact form the overall traffic of their website will also be taken into account. Advertising Platforms Why we choose Google and Yahoo! Since almost all internet users over the world use search engine to access information they need, we decide to spend all the assigned budget on the two top search engines, Google and Yahoo!. These top two search engines have the market share of nearly 80% in the United States [1]. Moreover, in the United States, the third-ranked search engine managed by Microsoft, Bing, adopts Yahoo!’s advertisement platform. This gives us a great opportunity to reach more than 90% of the internet user base in the United States. Both Google AdWords and Yahoo! Advertising are keyword-driven advertising services. The keyword-search feature enables us to directly reach the target group that potentially -1- has a specific interest in Blu Homes’ products. Advertising on Google and Yahoo! search pages also helps us collect statistics on keyword search in order to understand customers’ damand effectively. Moreover, both advertising platforms have great tracking system and analytics tools, such as Google Analytics, to help us identify statistics on impressions and clicks as well as visitors’ demographics. However, exposure of the ads depend on bidding strategy. We need to devise an effective bidding strategy such that our ad campaign effectively increase potential customers’ awareness of Blu Homes’ products. Why we do not choose Facebook Although social network websites, such as Facebook, are very popular such that an advertisement on Facebook can reach more than 500 million users, it is difficult to to identify people in our target group. An advertiser on Facebook could specify a target group by static information such as user’s current location, home town, education background, or age/sex demographics. However, a user’s profile on Facebook contains the location information only if the user enters his or her current location in the profile. [2] This factor makes it difficult for us to target potential home buyers in California. Moreover, advertising on Google and Yahoo! has an advatange over advertising on Facebook because on Google and Yahoo!, we could measure users’ explicit interest in Blu Homes’ products directly by their queries. On the other hand, it is difficult to evaluate each Facebook user’s interest in Blu Homes’ products directly. Facebook advertising platform also frequently shows ads which are not relevant to a user’s status or demographic information. Therefore, given a limited budget, we have decided to not run the advertising campaign on social network websites. http://siteanalytics.compete.com/google.com+yahoo.com+facebook.com/ Figure 1: Monthly traffic volume on Google, Yahoo!, and Facebook -2- Why we do not choose housing search engines We have also decided to not run the advertising campaign on housing search engines, such as zillow.com or trulia.com, because these websites mainly focus on ready-made houses for renting or selling. The target group for these housing search engines is thus different from Blu Homes’ potential customer base. The main advertising method on housing search engines is display advertising on the right-hand side of the home search result pages. An advertiser have to pay about $500 or above monthly to activate the campaign. Advertising on search engines is thus not feasible given our budget. Furthermore, traffic volume on housing search engines is significantly lower than that of Google and Yahoo!, as reported by Compete in Figures 2 and 3, respectively. http://siteanalytics.compete.com/trulia.com+zillow.com+google.com/ Figure 2: Monthly traffic volume on Google, trulia.com, and zillow.com http://siteanalytics.compete.com/trulia.com+zillow.com+yahoo.com/ Figure 3: Monthly traffic volume on Yahoo!, trulia.com, and zillow.com -3- Why we do not choose display ads Unlike sponsored search textual ads, display ads could achieve significantly lower performances in terms of CTR. Since display ads would potentially not optimize our advertisement strategy, we have decided not to use display ads in our advertising campaign. Also, with our tight budget, it would not be feasible to run a display advertising campaign for 6 weeks and achieve effective brand exposure and product awareness. Budget Allocation According to comScore’s statistics in October 2010 [1], Google has market share in the online advertising industry of 66.3%, while the coalition of Yahoo! and Bing has market share of 28%. This implies that competition in advertising on Google is abount 2.5 times as much as competition on Yahoo!. We would like to show ads in similar ad rank positions on Google and on Yahoo!. Our budget allocation for advertising on Google and on Yahoo! would thus be determined by the relative level of competition. Hence, the ratio of our budget on Google to budget on Yahoo! is 2.5 to 1. Target Group Blu Homes’ potential customers that we want to show our ads are mainly second-home or holiday-home buyers who are interested in environmental-friendly (‘green’) or energyefficient houses. Since Blu Homes is looking to expand their customer base in northern California, we would like to focus on potential customers in big cities such as San Francisco, Sacramento, Sillicon Valley, or Napa Valley because there is a strong demand for second or holiday homes in those cities. We decide to focus on potential home location in Napa Valley, Lake Tahoe, or along the pacific coast since those locations are not too far from big cities in northern California and, thus, would be attractive for building second or holiday homes that are environmental-friendly. Additionally, we would like to push Blu Homes’ products to potential customers who look for second or holiday homes but are not aware of or not currently interested in energy-efficient homes with innovative design. Therefore, our advertising campaign would not only target potential customers for ‘green’ home buyers, but would also create a new customer base for Blu Homes’ products by effectively leveraging our internet sponsored search strategy. -4- Keywords and Creatives As described in the section above, we started to focus on potential customers who look for ‘green’ homes. Our potential keywords would relate to the second- or holiday-home characteristics (e.g. second home, holiday home, weekend home) and the environmentalfriendly feature (e.g. green, eco-friendly, cost-saving, energy-efficient, efficient design). We have utilized Google AdWords keyword Tool [3] to identify related words and phrases that are relevant to our advertising campaign. This powerful tool provides us with the number of local monthly search and search trends. Our goal is to identify relevant keywords with medium competition so that we can show our advertisement among the top spots to effectively reach our target group with our tight budget. Moreover, we will leverage on Google Suggest (now also called Autocomplete System) [4] to come up with phrases that our potential customers may search for. We have prepared multiple creatives in order to identify key messages that could attract certain types of potential customers effectively. The four creatives used in the campaign are the following: 1. Green Homes in California Best designed & energy efficient. Save money & environment. Get info! 2. Energy-efficient Homes in California Best designed & energy efficient. Save money & environment. Get info! 3. Cost-saving Homes in California Best designed & energy efficient. Save money & environment. Get info! 4. Best-designed Homes in California Cost and energy efficient. Save money & environment. Get info! -5- Bidding Strategy As discussed above, we plan to spend around 60-70% of our budget on Google advertisement since Google has a bigger user base than Yahoo!. Moreover, we aim to show our ads among the top of ad positions at any search result page in order for our creatives to get highly exposed to our potential customers. Because of our limited budget, it is important that we learn bidding behavior of other bidders without exhausting our budget. We plan to start bidding in the medium range and then increase our bids so that each of our ads can be among top four ads. We need to monitor our bids by continuously adjusting our bids since other bidders may want to outbid our ads. This strategy is similar to a greedy strategy, where we try to solve for an optimal dynamic bidding strategy with limited information on other bidder’s valuation and tight budget constraint. Google AdWords Traffic Estimator [3] would help estimate ad position, number of daily clicks, and expected daily cost for each keyword. This tool helps us manage our budget effectively. Performance Metrics We run the advertising campaign on Google and Yahoo! for about 6 weeks (from the week of October 25th to the week of November 29th). We measure the click-through rate (CTR), average ad position each day, and number of impressions shown to analyze the performance of our marketing campaign. Both Google AdWords and Yahoo! Advertising have their own tracking tools to monitor and measure access volume and user demographics. We then analyze the performance data on booth Google and Yahoo! and provide recommendations for Blu Homes’ future online marketing campaigns. Overall Performances Table 1 reports performances of our advertising campaign in a 2-week period when creative 1 (which emphasizes the ‘green’ feature) is used. The click-through-rate and the average position are similar on both Google and Yahoo! campaigns due to our budget allocation strategy. The competition for keywords is more intense in Google than in Yahoo!. The average CPM on Google is thus higher than that on Yahoo!. -6- Table 1: Performances over a 2-week period Metrics Google Yahoo! 81 33 # clicks 19,420 9,987 # impressions 0.42% 0.33% CTR $1.79 $0.77 Average CPC $145.06 $25.38 Cost 4.9 4.64 Average position $7.47 $2.54 Average CPM Table 2 reports the competitive analysis of our advertising campaign over the 6-week period. The market share for impressions for our keywords are 10.59%, indicating that our campaign has significant market share for all keywords. The main reason that we could not achieve higher market share is because we did not increase the ad rank by bidding more. Increasing maximum bids on CPC will burn our budget per day too quickly. We will investigate further on this issue in the later section. Table 2: Competitive analysis of the advertising campaign Metrics Percentage 10.59% Impression share 5.15% Lost impression share due to budget limit 84.26% Lost impression share due to low ad rank Keyword Performances Keyword performances in Google and Yahoo! are reported in Tables 3 and 4, respectively. The key result from Table 3 is that “green home” is the best performing keyword in terms of number of clicks and CTR. Moreover, “holiday home” and “holiday houses” keywords have high number of impressions but receive almost no click. This implies that most people who are looking to buy holiday homes are not interested in the ‘green’ home. Therefore, Blu Homes should not focus on advertising the ‘green’ aspect to people who are looking for holiday homes in the sponsored search advertising campaign. Keyword performances on Yahoo! (Table 4) are in line with keyword performances on Google (Table 3). However, the number of clicks on Yahoo! are spread out across keywords. This implies that Yahoo! users have slightly different click behavior than Google users. -7- Table 3: Keyword performances in Google in a 2-week period Avg. # # Avg. Keyword CTR Cost Posi Clicks Impressions CPC -tion 62 6,321 0.98% $1.84 $113.95 4.7 Green Home 6 660 0.91% $1.62 $9.73 4.5 Eco Home 6 654 0.92% $1.93 $11.57 3.7 Eco House 4 554 0.72% $1.27 $5.09 6.1 Efficient Home 1 8,234 0.01% $1.30 $1.30 5.4 Holiday Home 1 337 0.30% $2.17 $2.17 4.1 Second Home 1 301 0.33% $1.25 $1.25 3.0 Second House 0 1,627 0.00% $0.00 $0.00 4.0 Holiday House 0 43 0.00% $0.00 $0.00 5.7 Energy efficient House 0 254 0.00% $0.00 $0.00 6.3 Energy efficient Home 0 179 0.00% $0.00 $0.00 4.5 Efficient House 0 14 0.00% $0.00 $0.00 5.0 Cost saving Home Total 81 19,420 0.42% $1.79 $145.06 4.9 Avg. CPM $18.03 $14.74 $17.69 $9.19 $0.16 $6.44 $4.15 $0.00 $0.00 $0.00 $0.00 $0.00 $7.47 Table 4: Keyword performances in Yahoo! in a 2-week period # # Avg. Avg. Keyword CTR Cost Clicks Impressions CPC Position 9 2,881 0.31% $0.91 $8.20 7.42 Green Home 1 281 0.36% $0.96 $0.96 4.43 Eco Home 1 475 0.21% $0.79 $0.79 3.40 Eco House 5 605 0.83% $0.66 $3.30 3.38 Efficient Home 2 850 0.24% $1.01 $2.01 4.35 Holiday Home 4 1,455 0.27% $0.59 $2.35 3.67 Second Home 1 487 0.21% $0.23 $0.23 2.50 Second House 3 995 0.30% $0.51 $1.54 2.56 Holiday House 0 84 0.00% $0.00 $0.00 4.33 Energy efficient House 4 278 1.44% $0.89 $3.56 4.25 Energy efficient Home 0 208 0.00% $0.00 $0.00 2.30 Efficient House Total 30 9,448 -8- 0.32% $0.76 $22.94 4.64 Click-through-rate Analysis Figure 4 plots CTR against average position on Google and Yahoo!. Many keywords have similar average position (in the range of 3 to 5) but have different CTR. Therefore, CTR is determined mainly by quality of the keyword. Selecting high-quality keyword is, therefore, important to run the advertising campaign optimally. 1.20% 1.00% CTR CTR 0.80% 0.60% 0.40% 0.20% 0.00% 0 2 4 6 1.60% 1.40% 1.20% 1.00% 0.80% 0.60% 0.40% 0.20% 0.00% 8 0 Average position 2 4 6 8 Average position Figure 4: CTR vs Average Position on Google (blue) and Yahoo! (green) Figure 5 plots CTR against average cost-per-click (CPC) on Google and Yahoo!. A higher CPC cannot guarantee a higher CTR. This result strengthens the claim that quality of the keyword is the main determinant of CTR. 1.20% 1.00% CTR CTR 0.80% 0.60% 0.40% 0.20% 0.00% 0 1 2 3 1.60% 1.40% 1.20% 1.00% 0.80% 0.60% 0.40% 0.20% 0.00% 0 Average CPC 0.5 1 1.5 Average CPC Figure 5: CTR vs Average CPC on Google (blue) and Yahoo! (green) Figure 6 plots CTR against number of impressions shown for each keyword on Google and Yahoo!. The CTR and the number of impressions have weak positive correlation. This feature is good for keywords with high number of impressions because those keywords can achieve high CTR and can also increase customer awareness of the product. -9- 1.20% 1.00% CTR CTR 0.80% 0.60% 0.40% 0.20% 0.00% 0 1.60% 1.40% 1.20% 1.00% 0.80% 0.60% 0.40% 0.20% 0.00% 2000 4000 6000 8000 10000 0 Number of Impressions 1000 2000 3000 4000 Number of Impressions Figure 6: CTR vs number of impressions on Google (blue) and Yahoo! (green) Day-of-week Analysis Performances of the overall campaign on Google and Yahoo! for each day of week are reported in Tables 5 and 6, respectively. According to Table 5, competition for keywords on Google increases on Fridays and weekends such that the average position on those days is worse than on other days. So if we have a limited budget to advertise on Google, we should avoid advertising on Fridays and weekends and focus more on advertising on Mondays to Thursdays to achieve high CTR. Figure 7 supports this recommendation. In Yahoo!, competition for keywords do not seem to increase on weekends since our average position does not significantly increase on weekends. Therefore, we should focus on advertising on Fridays and weekends since there is higher traffic volume on those days such that we could achieve high CTR without sacrificing on average position. Table 5: Day-of-week Analysis on Google Day Monday Tuesday Wednesday Thursday Friday Saturday Sunday # impressions # clicks CTR 1,704 2,368 790 1,153 1,127 1,045 1,524 5.0 6.5 7.5 6.0 5.0 5.0 5.5 0.29% 0.27% 0.95% 0.52% 0.44% 0.48% 0.36% - 10 - Avg. CPC $2.08 $1.82 $1.61 $1.69 $1.76 $1.76 $1.91 Cost $10.38 $11.86 $12.07 $10.12 $8.82 $8.82 $10.48 Avg. Position 4.78 4.65 4.76 5.19 5.30 5.07 5.07 Table 6: Day-of-week Analysis on Yahoo! Day # impressions # clicks 975 818 812 499 692 409 791 3.0 2.5 2.0 0.5 2.5 4.0 2.0 Avg. CPC 0.31% 0.31% 0.25% 0.10% 0.36% 0.98% 0.25% 1.00% 0.90% 0.80% 0.70% 0.60% 0.50% 0.40% 0.30% 0.20% 0.10% 0.00% CTR CTR Monday Tuesday Wednesday Thursday Friday Saturday Sunday CTR $0.63 $0.63 $1.06 $0.06 $0.80 $0.77 $0.99 Cost $1.90 $1.57 $2.13 $0.03 $1.99 $3.09 $1.99 Avg. Position 4.40 4.41 4.58 6.06 4.46 4.82 4.41 1.00% 0.90% 0.80% 0.70% 0.60% 0.50% 0.40% 0.30% 0.20% 0.10% 0.00% Figure 7: Average CTR different days of week on Google (blue) and Yahoo! (green) Analysis on bids We experiment with changing bids and measure performances. Increasing max imum CPC bid decreases average position and increases CTR for a keyword. This is because we show impressions less often and users are more likely to click the ads since it is in a higher position. On the other hand, decreasing maximum CPC bid increases average position and decreases CTR for a keyword. We would like to show our ads on the average position of less than 5 because we think that increasing bid would have a negative marketing impact on product and brand awareness, since we would then show ads less frequently with higher bids subject to daily budget limit. Decreasing bids would also have negative impact on product awareness since our ads will be shown in low-rank positions with poor exposure. - 11 - 1.30% 5.5 1.20% 1.10% 5 CTR Average Position 6 4.5 4 1.00% 0.90% 0.80% 3.5 0.70% 3 0.60% Normal Increasing Decreasing condition max CPC by max CPC by 25% 25% Normal Increasing Decreasing condition max CPC by max CPC by 25% 25% Figure 8: Average position for “green home” in different bidding conditions Analysis on Creatives We also experiment with changing creatives and measure performances of the campaign. We run campaign on the four creatives that emphasizes four different aspects of Blu Homes products: ‘green’, ‘energy efficient’, ‘cost saving’, and ‘best design’. Click distributions for each creative on Google and Yahoo! are similar, according to Tables 7 and 8. The main take-away from the results is that different keyword attracts different type of customers. If Blu Homes want to target people who are interested in the ‘green’ feature rather than people who look to buy second or holiday homes in general, advertising with creative 1 would draw most attention for this target group. Moreover, holiday-home buyers mostly look for the cost-saving and the design aspects rather than the ‘green’ aspect. Creative 2, which highlights the energy-efficient feature, draws more clicks on the energy-efficient related keywords than other creatives do. - 12 - Table 7: Number of clicks over 4 days with different creatives on Google Keyword Creative 1 Creative 2 Creative 3 Creative 4 ‘energy ‘cost ‘best ‘green’ Key message of creative efficient’ saving’ design’ 4 16 14 Green Home 2 Eco Home 2 1 2 Eco House 1 1 Efficient Home 1 Holiday Home Second Home Second House 5 Holiday House 1 Energy efficient House 1 Energy efficient Home Efficient House Cost saving Home Total 14 18 19 Table 8: Number of clicks over 4 days with different creatives on Yahoo! Keyword Creative 1 Creative 2 Creative 3 Creative 4 ‘energy ‘cost ‘best ‘green’ Key message of creative efficient’ saving’ design’ 6 Green Home 1 Eco Home 1 1 Eco House 1 1 1 Efficient Home 1 2 1 Holiday Home 2 3 1 Second Home 2 1 Second House Holiday House Energy efficient House Energy efficient Home 1 Efficient House Cost saving Home Total 7 8 - 13 - 5 6 Recommendations From running the sponsored search advertising campaingn in both Google and Yahoo! for 6 weeks, we have learned some important features in advertising Blu Homes’ product. We, thus, would like to recommend the following for Blu Homes’ future advertising campaign: 1. Optimize keyword selection since the keyword quality is the main determinant of CTR . 2. If budget is limited, show ads on Mondays to Thursdays in Google, and on Fridays and weekends in Yahoo! to achieve high CTR. 3. Use creative that advertises on the ‘green’ feature to achieve high CTR from our target group (i.e. people in California who is interested in buying ‘green’ homes). This is because different aspects on the ads would attract different customer groups. 4. If budget is limited, bid moderately such that ads are among the top four or five positions to achieve good CTR while having good exposure to potential customers to increase product awareness. 5. Understand the nature of websites that Blu Homes advertises on because people who use different websites may have different purposes of using it; this leads to different click behavior for each sponsored search website. Future work If Google Analytics for the landing page is accessible, we would like to calculate conversion ratio and CPA for each keyword. Moreover, we would like to use demographic data from Google Analytics on the landing page to identify the potential customer base in California for Blu Homes. Summary Sponsored search advertising campaign on Google and Yahoo! provides good performances in terms of click-through-rate and in terms of increasing product awareness for potential customers. The Google AdWords and the Yahoo! Advertising platforms facilitate the implementation of our bidding strategy and the data collection process. We hope that the analysis on our online advertising campaign would be useful for Blu Homes’ future advertising campaign. - 14 - References [1] comScore Releases October 2010 U.S. Search Engine Rankings http://www.comscore.com/Press_Events/Press_Releases/2010/11/comScore_Releases_Octobe r_2010_U.S._Search_Engine_Rankings [2] Reach and Targeting : Facebook Ads http://www.facebook.com/adsmarketing/index.php?sk=targeting [3] Google AdWords Keyword Tool / Google AdWords Traffic Estimator https://adwords.google.com/o/Targeting/Explorer [4] Features: Autocomplete (Google Suggest) http://www.google.com/support/websearch/bin/answer.py?hl=en&answer=106230 - 15 -