AdWords Post-Campaign Report: London Children’s Museum

Executive Summary The London Children’s Museum is a non-profit organization determined to engage and educate children through interactive learning and play. Over three weeks, we ran four campaigns (General Brand, Birthday Parties, Camps, and Memberships) to raise awareness for the museum and increase the number of camp registrations and birthday party bookings, all of which were targeted to London and a 40km radius to reach smaller communities. Throughout the duration of our AdWords campaign, we hoped to see a decline in CPC and an increase in our overall CTR. We also tracked conversions by measuring: views to the hours & admissions page, time on site greater than 2 minutes, and visits with more than 3 pageviews which were site traffic averages. We continuously monitored our account and took action to optimize various areas by pausing or modifying ad groups, campaigns, and keywords, conducting experiments, implementing ad schedules, and changing our bidding strategies to achieve our goals. The campaign began on March 13th and ended March 31st with a total cost of $250.49. Initially, we had set out to achieve: 12,500 impressions, a 4% CTR with 500 clicks, and an average CPC of $0.50. At the end of our campaign, we surpassed all of these goals, achieving a CTR of 14.75%, 657 clicks, and an average CPC of $0.30. As well, we achieved a conversion rate of 62.46% on our General Brand campaign. These statistics suggest that we were able to reach a more qualified audience than initially predicted and boost overall awareness of the London Children’s Museum brand. With this knowledge, it is reasonable to suggest that the client should continue implementing a General Brand campaign to increase brand awareness and drive qualified traffic to their website. We recommend that the museum also further utilize the Birthday Parties and Camps campaigns after the launch of their new website and with an increased budget to ensure an improved page position for highly competitive keywords. Once this site is created with an ecommerce feature, we recommend tracking conversions related to bookings. It would be also important for the client to improve the quality of their landing pages to increase the likelihood of conversions in the future, including prominent calls to action, relevant information related to the topic of keywords and ads, and relevant offers to increase interest. Finally, due to a lack of search volume, we do not recommend that the client continue using a Membership campaign. 1

Industry Component Campaign Overview: Prior to our efforts, the London Children’s Museum was not using AdWords advertising, so we were eager to take on this challenge. During our campaign which ran from March 13th to March 31st, we spent a total of $250.49 by running 4 campaigns with a total of 10 ad groups and 273 keywords. We had hoped to allocate our budget equally across all 4 campaigns, but soon realized some ad groups were more responsive than others and rearranged our strategy to accommodate. We hoped to achieve 12,500 impressions, 500 clicks, a CTR of 4%, and an average CPC of $0.50 while running our campaigns solely on the search network. We also sought to maintain quality scores of 5/10 or higher on all keywords using a 30 km radius around London. To monitor these results, we executed daily search query reports, downloaded reports, checked the opportunities tab, and kept an eye on quality scores. We planned to achieve our goals by performing ad testing, pausing unsuccessful campaigns and keywords, and managing our bidding to achieve the highest page position with a low CPC. The chart below demonstrates the stability of our low average CPC across all campaigns:

Evolution of Campaign Strategy: Week One: Although we had developed what we believed to be an exhaustive list of keywords, we decided to only load those that would drive the most qualified traffic to start such as “book summer camps”, “london childrens museum” and “kids parties in london”. We also began each ad group with an A/B ad test. As soon as enough data was accumulated, we determined the winning ad and replaced the losing ad with another creative. Throughout the week, we monitored metrics regarding keywords, quality scores, and CTR as people began to click the ads. We initially developed a rule to pause any keywords with a QS of 4 or below, but soon changed it to a QS of 3 or below. For those keywords with a QS of 4, we tried tighter match types in an attempt to improve their score. In some cases, adding either a phrase match or exact match revived the keyword and improved overall results. We also added negative keywords such as “campground” and “cakes”, and good keywords including variations to pre-existing keywords in all campaigns from the search query report. Near the end of the week, we switched from an automatic to manual bidding strategy, as this allowed us to have more control over individual 2

keyword bids, and devote more time towards optimizing problem areas. For instance, bids were inflated for keywords with low page positions in an attempt to boost their performance, or deflated for those keywords that were not driving any clicks or had low search volume. Week Two: We added more keywords and ads into each campaign to maintain high QS, drive more traffic, and prepare for upcoming bid experiments. We also developed a mobile ad for each ad group and modified low-performing ads to include the Dynamic Keyword Insertion (DKI) tool. By doing so, we gained more qualified traffic and further decreased our CPC. To make our changes easier to track, we added labels for branded keywords and those with boosted CPC bids. We noticed a positive correlation between times of day and CTR in our Camps campaign, so we created an ad schedule which focused on delivering 10% more spend to times that performed well and 30% less to those with low CTR. We also started bid experiments on keywords with impressions but no clicks, with the goal of improving page position. Due to limited budget, we paused the experiment after seeing minimal results. In an attempt to optimize our location reach, we changed our radius from 30 to 40 km based on traffic data from Google Analytics indicating qualified traffic from other areas outside of London. We also decided to set the daily budget to $3.13 for all campaigns, but soon decided that funding needed to be reallocated to higher performing campaigns. As a result, we increased the General and Camps campaigns to $3.63 per day and reduced Birthday Parties and Memberships to $2.63 a day. Week Three: This week was heavily focused on maintaining our CPC and CTR, as well as experimenting with different functions of the platform to further our results. Each day we monitored our spend in comparison to the amount we had left in the competition and re-adjusted our budgets. We added a flexible bid strategy to the Birthday Parties campaign to increase clicks and drive more qualified traffic, but did not notice any significant gains. By monitoring locations, we realized some postal codes in the London area were performing better than others, so we opted to inflate and deflate bids to lift CTR. In the final few days, we inflated our budgets based on recommended levels to increase impressions. By doing so, we saw an increase in clicks within 24 hours totalling $60 in spend. Throughout the campaign, one of conversions we tracked involved navigation to the hours and admissions page from the landing page of an ad. In comparison to the three weeks prior to the campaign, views to this page increased by 76% with a decrease in bounce rate and 2,134 new visitors. These results began to populate during the first week of the campaign and dramatically increased towards the end. 3

Key Campaign Results: The table on the

Metric

Pre-Campaign Goal

Actual Outcome

Impressions

12,500

4,400

Clicks

500

657

CTR

4%

14.75%

Average CPC

$0.50

$0.30

right illustrates a comparison between our pre-campaign goals and our actual outcomes. We surpassed all of our primary goals. Google Analytics data confirms that 89% of all AdWords traffic during the three weeks was attributed to our General Brand campaign, with our Membership campaign having the poorest performance, accounting for under 1% of total paid search traffic. Throughout the course of the campaign, we found it was necessary to pause various ad groups with poor performance. For instance, our “Kids Activities” ad group under the General Brand campaign received no clicks or impressions after one active week. In order to continue improving the performance, we paused the ad group allowing for more of the budget already allocated to the campaign to focus on improving the page position for better performing ad groups. While we did not meet our impression goal, we were able to achieve a higher click-through rate and number of clicks overall. This demonstrates that we reached more qualified traffic than was originally predicted with an average page position of 1.4. As a result, we achieved a final conversion rate of 62.46% and average conversion cost of $0.23 by measuring: views to the hours & admissions page, time on site greater than 2 minutes, and visits with more than 3 pageviews which were site traffic averages. During the last few days of the competition, we developed a strategy focused specifically on high performance areas. In the end, we only ran the General Brand campaign, which allowed us to allocate the remaining funds to achieve the lowest number of impressions on the last day within a tight budget, but also the highest CTR of the entire campaign at 26%. By allocating the budget in this way, we reached qualified traffic and boosted our account in the final days of the competition. The below chart outlines the conversions that were tracked for the General Brand campaign and their overall results:

4

Mobile: We received an overall mobile CTR of 17% with 59 conversions. The average CPC across these devices was in line with our account-wide average at $0.35. In total, 5 out of 11 of our mobile ads generated stats, while the others were paused due to low traffic. All devices had bounce rates below 45%, but the best performing device was the computer at 22.5% and an average of 4.3 pages viewed per sessions. Geographic & Bid Adjustments: London drove the highest traffic to the website, accounting for 97% of total clicks. The remaining 3% were a result of traffic from four surrounding communities which each achieved a CTR above 2%. As a result, expanding our geotarget to encompass a 40 km radius around the city did very little to improve our reach. In the last week of the competition, we looked closely at the locations of our clicks and inflated or deflated by 10% based on postal codes and cities. For instance, we noticed that London was receiving the highest traffic so it was inflated for our General Brand campaign to increase reach. We did not make any bid adjustments to devices, as each one was performing well. We also experimented with ad schedule bid adjustments, but we did not see positive results and paused them after a few days. Extensions: When callout extensions were displayed with the ads in any campaign, the overall CTR increased to 17.6%. Our top performing extensions included: “3 floors of exhibits”, “interactive play” and “day camps” at approximately 34% CTR. The use of callout extensions therefore dramatically improved the results of our campaigns by taking up more real estate on the search results page and attracting more qualified traffic. The use of sitelink, call, and location extensions also dramatically improved our CTR when present in the ads. Below are examples of sitelinks with the highest performance:

Rules & Experiments: We experimented with the keyword pause rule, which specified that keywords with a QS lower than 3 be paused automatically. As well, we specified that reports related to keywords, ads and ad groups be sent directly to our email account daily. These rules were useful for aggregating data into one central location. A bid experiment was also conducted on the Camps campaign. Here, we inflated bids on keywords that had impressions, no clicks, and

5

a page position lower than the first page. In doing so, we noticed improved page position of a few keywords, but opted to stop the experiment after a week due to budget restrictions. Bid Strategies: In the last week of competition, we opted to try a flexible bid strategy to increase clicks on our Birthday Parties campaign. The use of this strategy for 4 days led to a decrease in impressions and clicks, but an increase in CTR. This suggests we were able to reach more qualified traffic and achieve a better page position through a higher bid. Keyword Quality Scores: In the pre-campaign report we had set a goal of maintaining a keyword quality score of 5 out of 10 on most keywords. We were able to reach this goal, achieving 229 out of 273 keywords with quality scores above this threshold. Our best performing keywords were those in the General Brand campaign, with an average CTR of 10% and page position of 1.0. Broad match and phrase match keywords performed the best, while modified broad match and exact match performed the poorest with less than 5 clicks per keyword in any campaign. The two keywords that performed the best over the entire account included: london children’s museum and “london childrens museum”. Ads: On average, ads using the Dynamic Keyword Insertion tool were shown more often, received more clicks, more conversions, and a higher CTR. The ad on the right performed the best of all ads in the account, with 1,054 impressions, a 33% CTR and 266 conversions. We noticed that performance was lower on weekends and increased during the week. By further optimizing at the end of the campaign, we were able to achieve a spike in CTR and a lower average CPC. This change can be attributed to the ending of the flexible bid strategy and the pause of the Camps campaign to allocate more budget to the General Brand campaign which was performing well. Conclusions: Overall, we feel the success of the General Brand campaign is related primarily to the need for brand awareness. It is possible that the community may not be as familiar with the Children’s Museum as the client may think, therefore the ads caught the audience’s attention. By having the opportunity to experiment with various optimization techniques and the use of trial and error, our group achieved success that worked with our client’s marketing objectives. By achieving a low CPC, high page position and CTR, and an outstanding conversion rate, we have demonstrated just how effective paid search advertising can be for the Children’s Museum in the future. Our successes and failures have helped to formulate strategic recommendations that we can provide the client for campaign development moving forward. 6

Future Recommendations: Our recommendations focus on content and site improvements. Though the client’s site does a great job of displaying key information above the fold and featuring useful content, there are key aspects that should be improved. The Children’s Museum offers a variety of camps throughout the year, therefore they could benefit from improving related pages. A direct result was seen with our Camps campaign, as we achieved lower quality scores because of the landing pages. Including ecommerce would help conversion tracking and could increase the number of bookings. Regarding AdWords content, we found the General Brand campaign was most successful, therefore we recommend continuing with a campaign related to the brand. Another key recommendation is related to geotargeting. Even though we had a radius target that encompassed many of the communities around London, we found it had very little advantage. It is recommended that target spend be focused on the city of London entirely and only include those smaller communities that are in close proximity. In the future, we also suggest the client apply for Google Grants as a non-profit in order to receive $10,000 in inkind advertising to promote their mission and services to the London community. Learning Component Learning Objectives & Outcomes: By challenging ourselves to achieve results under budget restrictions and client expectations, we hoped to learn about the advanced optimization features on the AdWords platform including the search query report, ad schedules, flexible bid strategies, and bid adjustments. We wanted to gain real world experience working with a client who was looking to achieve tangible marketing objectives. Through this process, we came to understand how to run an effective campaign on a limited budget, while analyzing key metrics and optimizing the account regularly to improve results. We realized it is also important to monitor what competitors are doing. Throughout the campaign, we used the Ad Preview tool to monitor companies ranking for similar keywords and evaluate how we could improve our ads. One of the key challenges was working with a client who did not have strong landing pages that were relevant to the ads we created. We learned that in this situation, it is important to optimize the factors we could control including ad relevance and tight audience-relevant keywords. By doing so, we are able to improve our QS for core keywords even with poor landing page experience. A key unexpected outcome was that we had to pause more ad groups than we initially predicted to improve our account statistics due to low search volumes, high CPCs and high competition in some ad groups. We were surprised that our General Brand campaign was the 7

most successful, since we assumed most people in London were familiar with the organization and would rather learn about activities they offer. We realized that it is important to run a General Brand campaign when developing a strategy for a client, especially if one of their primary goals is to increase foot traffic. It became evident that it is crucial to understand the culture of the community you are marketing to and specifically what they are searching for. For instance, running a campaign related to Summer Camps might have been more successful in April or May when parents are thinking about registering their children for the summer. Group Dynamics: It was difficult to coordinate schedules and stay cohesive as a team while ensuring everyone was aware of next steps. To overcome this obstacle, we developed a shared Google Doc with a checklist of daily account optimization tasks and a list where each person could explain the changes they made each day. As well, some members had strengths in various areas, but not everyone was equally familiar with each. To capitalize on each other’s strengths, we designated weekly tasks such as budgeting, ad writing, and keyword development. Client Dynamics: When we began working with our client, we were informed a new website was in the process of being created; however, it was delayed and ultimately impacted our ability to provide relevant landing pages in our ads. To overcome this obstacle, we chose landing pages on the current website that best suited the keywords and ads. As well, we were informed shortly before our launch that the museum’s March Break camp was fully booked. This impacted the structure of our campaigns, and therefore caused us to reallocate the budget. Overall, our client was helpful, enthusiastic about the use of AdWords, and would like to continue using it as a marketing vehicle in the future. Future Recommendations: One way we could have worked better as a team is if we had scheduled frequent weekly meetings rather than just meeting for major changes and checkpoints. In doing so, each group member would have had a better opportunity to actively participate in campaign updates. If we had the opportunity to re-do the campaign, we would have tried to better manage the budget to reduce the risk of overspending. After originally pausing the account at $248, we noticed the total spend continued to increase shortly after as a result of over delivery. In the future, we would monitor this more closely and adjust our budgets. In order to better manage our client relationship, we would have liked to communicate with them more frequently throughout the campaign to better monitor increases in foot traffic and bookings since the website did not include an ecommerce tracking feature for these metrics. 8

Social Impact Statement London Children’s Museum www.londonchildrensmuseum.ca 21 Wharncliffe Rd S, London, Ontario N6J 4G5 Jill Osborne – [email protected] Community Impact of the London Children’s Museum: Since 1975, the London Children’s Museum has been enriching the lives of children and families in London and surrounding communities. With nine permanent exhibits that provide guests with an interactive learning experience, children can indulge in a stimulating learning experience focused on the areas of: history and heritage, science, social relationships, art, and culture. Through on-going educational programs like community groups, camps, and educational field trips, the London Children’s Museum continues to make an impact on the London community and provide a unique experience that cannot be replicated. Designing AdWords Campaign to Meet Client’s Goals: At the start of the campaign, the Children’s Museum specified that they would like to focus on increasing March Break and summer camp sign-ups, increase the number of birthday party bookings, encourage people to purchase family membership, and increase foot traffic. As a group, we designed the AdWords campaign to meet these goals, keeping in mind the ways we could drive greater site traffic and increase overall awareness in the community. To do this, we developed campaigns for each objective: Birthday Parties, Camps, Membership, and General Brand. Within each of these campaigns, we researched the types of phrases people use to search and refer to each topic and implemented them as ad groups and keywords. In developing our strategy, we also tracked conversions in the General Brand campaign, specifically: views to the hours & admissions page, time on site greater than 2 minutes, and visits with more than 3 pageviews. By achieving a very high conversion rate, we were able to drive qualified traffic to the site and encourage people to take actions which were in line with our client’s objectives.

Campaign Summary in Less Than 100 Words: We were able to surpass all of our primary goals, achieving 4,400 impressions, 657 clicks, a CTR of 14.75% across all campaigns, and an 9

average CPC of $0.30. Google Analytics data suggests that 89% of all AdWords traffic during the 3 weeks was attributed to the General Brand campaign, with Membership having the poorest performance, accounting for under 1% of total traffic. Our General Brand campaign was responsible for all conversions (62.46% CTR). As a result of poor landing pages and low search volumes, our Camps and Birthday Parties campaigns did not perform as well as we had hoped. How AdWords Impacted the Client’s Mission: After analyzing the effectiveness of the General Brand campaign, we discovered that there was an increased interest in the museum as a family destination. This was seen primarily through users’ time on the site and high clickthrough and conversion rates. By showcasing the unique offerings of the museum, the AdWords account was able to generate an increase in foot traffic and persuade people to take action to book parties and enroll their children in camps. This aligns with the museum’s mission of providing interactive and educational experiences for children. By increasing brand awareness and driving greater engagement, the museum is able to stay true to their mission and offer unique experiences to more families.

Future Recommendations for the Client: In the future, we recommend that the London Children’s Museum continue to invest in and make use of AdWords campaigns. Through analyzing the General Brand campaign, our group was able to determine that brand awareness should in fact be a focus for the museum. It will also be important to advertise specific areas of the organization, including birthday parties and camps, but these should be run strategically throughout the year rather than on an on-going basis due to high costs. We recommend the client complete their new website and design landing pages that can be used alongside AdWords. These pages should include specific content, calls to action, and offer a positive user experience. After evaluating the success of the campaign over the past three weeks, we recommend that the museum continue to develop AdWords campaigns and establish conversion tracking for bookings once the new site is launched. In the future, we also suggest the client apply for Google Grants as a non-profit in order to receive $10,000 in in-kind advertising to promote their mission and services to the London community.

10