CHAPTER 6 EVALUATION
6.4 Evaluation Results
There are two businesses are willing to hold a joint campaign through our system service.
They are Vocals and Ming Shi Fang.
Table 6.4 Case study business information Business Information
Vocals Product and service: A music platform for no famous people.
Official website: http://www.vocals.tw
Facebook fan page: https://www.facebook.com/vocals.tw Ming Shi Fang
(茗熙坊)
Product and service: Herbal tea.
Official website: http://mingshifang.com.tw/
Facebook fan page: https://www.facebook.com/MingShiFang After discussion, the following are the details of joint campaign:
1. Campaign name: Solicit the slogan for new joint product.
2. Main marketing online channel: Facebook fan page.
3. Campaign type: Gamification.
4. Campaign details: They will post a campaign information article, and encourage fans to response the article with “Slogan: Design slogan. Tag one friend.” Activities for a week, whose slogan get the most "Like" at the end can get prizes.
And we collect the data to prove our mechanism is useful and effective.
6.4 Evaluation Results
In this chapter, we will display our result to support our propositions we mentioned in previous chapters through the interview data and statistic test. After analysis of the questionnaires, we first selected out opinion leaders (Figure 6.9). Through analysis of the results of the questionnaire on the hypothesis that we can make some conclusions.
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
76
Figure 6.9 Introvert leaders’ analysis through our questionnaire
In x-axis, positive value of x reflects “being not influenced by neighbors”, and negative value of x reflects “being influenced by neighbors.” In y-axis, positive value of y reflects
“inward-interaction preference” and negative value of y reflects “outward-interaction.” The integral calculation already mentioned in chapter 6.3.2 (See Figure 6.2, 6-3, 6-4 and 6-5). As Figure 6.4 shows, introvert leaders are subject 11, subject 17 and subject 31. We classified the grouping of introvert leaders and non-introvert leaders, and then do a cross analysis for the content of the article (See Figure 6.6, 6-7and 6-8). Results of the analysis are sufficient proof to our propositions.
Proposition 1: In a brand community, different types of articles will have different degree of engagement for introvert leaders with fan page.
‧
Table 6.5 Descriptive statics of different content to introvert leader
Table 6.6 Descriptive statics of different content and article type to introvert leader
In Table 6.5, we can know that the discount information article gets the highest mean score, but it also gets the largest standard deviation. In the interview, subjects agreed that this type of
article dependents on the goods. Even discount articles, if the goods are not the subject's needs, subjects will not read the article. As the second-highest mean score content is informative type, and it’s only 0.577 standard deviation. Introvert leaders in these types of articles have a more consistent degree of preference. In contrast, entertaining articles have lower mean score consistency. In addition, in Table 6.6, we can know that whether informative or imagery content, introvert leaders prefer “text + image” more than “text” and “text + video”. And subjects also mentioned that compared to image, film is a more risky factor. Film has a lot of factors that will affect the attitude of the reader, such as movement, sound, story, etc. Compared with watching image, readers have the higher expectation when they desired to watch the film.
Based on these reasons, subjects are more willing to watch the image than film. We summarize
Descriptive Statistics
N Min. Max. Mean Std. Deviation
Informative - Text
Informative - Text + Image Informative - Text + Video Imagery - Text
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
78
the conclusions of this part, introvert leaders prefer informative articles, and text and image type of articles more attractive to them. According to the above findings, we will validate that the informative article with text and image can engage introvert leaders in case study.
Figure 6.10 The behavior on Vocals Facebook fan page during campaign
Figure 6.11 The behavior during June and July
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
79
Figure 6.12 The behavior of the post on 7/9
Figure 6.13 The behavior of the campaign post
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
80
Our joint campaign (See appendix A) begin on 7/5, and we can see that the behavior (likes, share, response) of fans on Vocals Facebook fan page is increasing obviously (See Figure 6.11), especially in “sharing” on 7/9. However, the article we posted on 7/9 didn’t get any log of
“sharing” (See Figure 6.12). After a cross-comparison and we found that through the spread of article on 7/9, let the campaign indirect exposure. The "share" behavior in Figure 6.13 can find that most of the “share” behavior belongs to "inwards interaction." It means user share the article through the personal message box or forward article through the url but not the “share”
feature. We presume that we had attracted introvert leaders, and they shared the article through inward interaction with their neighbors. In addition to "share" behavior increased significantly, interactive behavioral fan page also increased a lot (See Table 6.6).
As the evaluation method we mentioned in chapter3, we can calculate eWOM performance through subtracting the data this week and last week’s data. We use “The total number of people seeing your Page post” as our data (Table 6.6.) The meaning of this data is the “post’s reach.”
P = 𝑒𝑊𝑂𝑀𝑡− 𝑒𝑊𝑂𝑀(𝑡−1)
In this equation, the t means time period. We need to calculate the average of eWOM of a time period. In our data (Table 6.6), 𝑒𝑊𝑂𝑀𝑡 is the performance of article during 7/5~7/11, and 𝑒𝑊𝑂𝑀(𝑡−1) we use the history data during 4/29~7/1. After calculating, we get the P = 1.077. We have improved the eWOM performance in 7.7%.
‧
Table 6.7 Data from Vocals Facebook fan page
In Table 6.7, the data is logged from Vocals Facebook fan page insight manager. The number 10 article is the first article of our campaign. The data is the behavior on Facebook of unique user. We can see that since the launch of our designed article, interactive user there is an increasing trend. We calculate the interaction rate as the number of clicks (The number of people who clicked anywhere within your post) divided by the number of reach (The total number of people see your Page post). The increasing rate of interaction rate can be more clearly seen in Figure 6.14.
Number 1 2 3 4 5 6 7 8 9 10 11 12
Date 4/29 5/1 5/2 5/3 6/27 6/30 7/1 7/2 7/5 7/8 7/9 7/11 The total number of people see
your Page post 525 268 601 1222 2227 273 656 1297 840 581 1213 1173 The number of people who
clicked anywhere within your post
12 9 16 74 96 10 35 100 162 47 118 166
The number of people who saw your Page post because they've
liked your Page
348 88 391 808 860 199 579 679 206 502 1026 994
The number of people who have liked your Page and clicked anywhere within your posts
20 11 28 85 83 13 39 69 36 49 121 160
Interaction rate 0.06 0.13 0.07 0.11 0.10 0.07 0.07 0.10 0.17 0.10 0.12 0.16
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
82
Figure 6.14 The trend of interaction rate in Vocals Facebook fan page
Proposition 2: The articles from cooperation stakeholders with different types of articles will have different degree of engagement for introvert leaders with fan page.
In the process of cooperation, Ming Shi Fang said they do not tend to forward articles form Vocals, and they gave some experience sharing (Table 6.8).
Table 6.8 Interview of Ming Shi Fang
Subject Interview
Ming Shi Fang 茗熙坊
1. Fans prefer the original articles from fan page more than forwarded articles.
2. They have scheduled their posting process, adjusting it will cause confusion to fans.
3. Ming Shi Fang thinks the campaign is held on Vocals fan page, posting on their Facebook fan page cannot get much effect. They prefer focus posting on Vocals Facebook fan page and tag Ming Shi Fang on the article, it will also get the exposure on the walls of Ming Shi Fang’s fans.
4. As a Facebook fan page manager, Ming Shi Fang considered
‧
that the introvert leader belonged to Vocals, they will be engaged by Vocals but not Ming Shi Fang.
When the campaign started, Ming Shi Fang forwarded the campaign information to their fan page, but only got 17 likes, lower than their history average performance. According to their observation, their history average performance of posts got 20~60 likes generally. Ming Shi Fang considered that the article style is too different with their style to make response.
However, in Table 6.9 we can see that subjects are willing to do the further understanding on the cooperating business. It supports that businesses joint with each other still can get some acquisition on potential customers.
Table 6.9 Descriptive statics of stakeholder cooperation’s user engagement
Descriptive Statistics
Image that you are the loyalty customer of store A. Now A and B announced joint campaign…
N Min. Max. Mean. Std. Deviation
I’m willing to read carefully 32 1 5 3.78 1.070
I’m willing to browse B fan page 32 1 5 3.72 0.991
I’m willing to search B’s information on
search engine 32 1 5 3.47 1.295
I’m willing to search B’s information on recommender platform
32 1 5 2.94 1.134
If I have a chance, I’m still willing to participate the campaign even I was not familiar with B.
32 1 5 3.41 1.103
Proposition 3: The different kinds of behavior from introvert leaders will drive different degree of engagement for fans on the fan page.
Through our questionnaire (See Table 6.10), we have found that subjects have the lowest engagement degree when they obtained the information from fan page directly. There are not too much different between the sources that are delivered from friend or famous person, but if they (no matter friends or famous people) add their own comments will enhance subjects’
‧
engagement degree. In other words, if fan page manager can engage the introvert leader and stimulate them to pass the information, introvert leader’s influence power to his neighbors may not be lower than a famous person.
Table 6.10 Descriptive statics of user engagement from the article source
Based on our case study result, too much interaction that belong to “inward interaction.”
It causes the difficulty on crawling and analyzing data. However, the exposure of article caused by the inward interaction is very clear. After the “inward sharing” being happened, the behavior on the article were increased (See Figure 6.11).