CHAPTER 6 EVALUATION
6.4 D ISCUSSION OF F INDINGS
6.4.1 Propositions
6.4 Discussion of Findings
The purpose of this research is to design a mechanism to stimulate customer engagement behavior through empower SMEs to Plan+ Value co-create + Do+
Value co-create+ Act+ Value co-create(See Figure 5.1.3) on search engine. In section 6.2 and 6.3, we exhibited the experiments and evaluated our proposed propositions as well as uncovered some possible explanations. In this section, We discuss in further detail of these findings below.
6.4.1 Propositions
Figure 6.4.1 propositions
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
98
(1) Proposition 1-1, 1-2 is accepted:
In proposition 1-1 and 1-2, we examined the long tail keyword services and inlink service can increase the web visibility.
First, engagement site’s long tail keyword service enable SME, during the plan period, to value-co-create with search engine to choose long tail keyword to revise the campaign description. As we see the result from the campaign 1(See Table 6.2.5), the keywords that provide by engagement site’s long tail keyword module have significant higher ranking over SME-original-choose keywords rankings. It implicates that the keywords suggestion from engagement site indeed can help SMEs to value co-create with search engine and enable SMEs to have more power to revise the campaign as they want to acquire their targeted customers.
After SMEs revise the campaign and go through the suggested process in our engagement site, we examine the web visibility change as Table 6.2.5.
There is dramatically improvement on keywords ranking and we about double the numbers of keywords appear in google search engine. In practical, we also prove that through our long tail keyword, SMEs can increase their web visibility significantly.
Second, engagement site’s inlink service enable SME, during the do period, to value-co-create with inlink owner to adjust campaign together. In this way, the value bring to inlink owner is the content of campaign that increase the value of their website, and SMEs can take the value of getting feedback of their campaign. If the inlink owners are willing to introduce and put inlink to the campaign, then the increase of web visibility help SMEs to acquire more new customers from different sources.
(2) Proposition 2-1 is accepted, but 2-2 is not accepted:
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
99
Search targetability saves customer’s valuable time on shorten the search process. Thus, it’s important to prove that with inlink and long tail keyword service we can increase search targetability.
First is the SME-revised long tail keyword which from customer’s angle has been strongly agree to have higher search targetability. Here we also present a graph to show that indeed our mechanism for choosing higher fitness keywords in long tail keyword service will makes the search targetability higher (See Figure 6.4.2). Thus it’s quite reasonable that SMEs can follow our advice to pick up keywords to optimize for. We then prove that our service can increase the search targetability.
Figure 6.4.2 Max fitness keywords search targetability comparison Second is inlink service can’t show significant difference in bring search targetability to user. We further analyze for the reason to influence it.
We further analyze the four campaigns if there are difference in image-related vs image-unrelated promotion inlink site score. We found that the average score over image-related vs image-unrelated promotion inlink site search targetability on campaign 1, 2, 3, and 4 are 4.1, 4.5, 5.22, and 3.375.
We can see that campaign 4 image-related link has the worst score. And we
1 2 3 4 5 6 7
max vs min fitness keyword search targetability
max fitness vs random fitness keyword search targetability Liker 7 point scale of agree degree on max fitness keywords have higher search targetability
Search Targetability
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
100
look into the detail of campaign 4. Campaign 4 is a natural campaign (體驗有 機養生藥膳，手作原住民飾品). The natural image campaign are related to the MOMIGO website which we recommended as inlink for the SMEs due to its natural image (left side in the figure 6.4.3).
Subject 26 gives us feedback,” There are many other perspectives to consider rather than image when they view a website.” Subject 22 also feedback, “Image is only one part of aspect to view from, I myself only focus on things that I have experienced or interesting.” Thus we conclude that this proposition is not supported though there are slightly better score for average search targetability for image-related inlink site over image-unrelated inlink site, but there may be more variables we should control of like UI, UX, interesting of subjects or topics…etc.
Figure 6.4.3 Image-unrelated inlink site (left) vs image-related inlink site (right) (3) Proposition 3 is accepted but need further explanation:
Though the inlink service can actually bring higher campaign and partner image consistency but doesn’t seem to be very significant. Thus, we do further graph to compare groups that have high image distinguish ability (Top 50% in subjects) and low image distinguish ability (Bottom 50% in subjects) over the image-related or image-unrelated link consistency. As Data shows (See Figure 6.4.4, 6.4.5), we can see that image-related link consistency can tell from high
MOMIGO
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
101
image distinguishing ability people which is good for SMEs to use our link-building service to help focus on customers that can distinguish the image.
Figure 6.4.4 High Image Distinguish Ability Customers Judgement Over Image-related and Image-unrelated Link Consistency
Figure 6.4.5 Low Image Distinguish Ability Customers Judgement Over Image-related and Image-unrelated Link Consistency
(4) Proposition 4-1, 4-2 is accepted, but 4-3, 5-2 is not accepted:
For proposition 4-1, we proved that both inlink service and keyword service can influence on engagement level from external data from engagement site. It suggests that the web visibility indeed is one of the most important which we
33%
67%
High Image Distinguish Ability Customers judgement over related and
image-unrelated link consitency
Image-realted inlink has higher consistency
Image-unrealted inlink has lower consistency
50%
50%
Low Image Distinguish Ability Customers judgement over related and
image-unrelated link consitency
Image-realted inlink has higher consistency
Image-unrealted inlink has lower consistency
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
102
also tell from the interview that the SMEs all very focus on how to acquire customer. And the most important things in new media to acquire customers are to increase the web visibility. Thus we prove that web visibility is actually cannot be neglected and should put effort on.
For proposition 4-2, both high search targetability keyword and inlink can bring significant higher engagement behavior. It implicates that instead of buying the popular keywords’ advertisement, SMEs should focus on high search targetability long tail keywords to create higher engagement behavior.
It also implicates that SMEs should not pay lots of money to popular blogger or website owner to promote themselves, but can focus on high search targetability inlink site to acquire and exchange with free experience of campaign or so to trade for promotion and reach high search targetability and stimulate the higher engagement behavior.
For proposition 4-3, 5-2, we further analyze on 4-3 but not 5-2 since it is obvious not supported. To further analyze this, we divide high and low group of campaign and partner image consistency to find if there are difference between these two groups in different level of engagement behavior (See Figure 6.4.6, 6.4.7). But we still find little difference between these two groups no matter in which engagement level. Thus, we conclude that campaign image consistency will influence little on engagement level and do not have an inverted-U shape relationship with engagement level in a campaign situation.
‧
Figure 6.4.6 High campaign and partner image consistency inlink site engagement level
Figure 6.4.7 Low campaign and partner image consistency inlink site engagement level
Though proposition 4-3 and 5-2 are not accepted, but these are actually good things to SME, since they can pay less attention on image consistency in the campaign situation. But we also find that there are 31 out of 33 subjects Customer's 7 point Likert scale of aggree to behavior
High campaign and partner image consistency inlink site engagement level
Customer's 7 point Likert scale of aggree to behaviorLow campaign and partner image consistency inlink site
engagement level
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
104
consistency, but still need to focus on their own image to attract their own targeted customers.
(5) Proposition 5-1 is accepted:
Proposition 5-1 says that the engagement site’s engagement level has positive relationship with target site’s engagement level. It’s very intuitive and has been proved very significantly, thus we suggest that under the assumption of engagement site is a popular site than SMEs’ target site (our engagement site actually has average 651 clicks per month for the pass years) which is just in the beginning of creating their own target site. Thus it’s a plus for SMEs to engage their customers also on the engagement site and lead the traffic to their target site
(6) Proposition 6-1.1 and 6-1.2 is accepted:
Proposition 6-1 says that the long tail keyword service can help SMEs through revise the keyword and content to make the engagement level increase. As we can notice from the data, SME-revised keywords tend to be high search targetability keywords. Thus it can increase the engagement level of customers.
What is interesting is what we find in Figure 6.3.30, 6.3.31, 6.3.33, and 6.3.34, customers have higher involvement level on engagement site and higher interaction and intimacy in target site if customers use SME-revised keyword as a search query. It implicates that through click into the campaign of engagement site, customers (with higher potential to engage) have higher interaction and intimacy toward the SME. We believe that SMEs thus can acquire their loyal customer through applying long tail keyword service.
(7) Proposition 6.2.1 and proposition 6.2.2 is not accepted:
In this proposition, image-related inlink that is suggested by the inlink building service should be better than the image-unrelated inlink. However
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
105
this proposition is not statistically accepted. As we stated before inlink targetability may increase the engagement level, but the image-related link and image-unrelated link are not the promise of high inlilnk search targetability (stated at discussion of proposition 2-2). Thus we argue that there are more relative factors should be included for example categorize topics and only recommend the relevant topics inlink to the SME. Though the proposition is not accepected, the average involvement, interaction, intimacy, and influence level is actually higher than image-unrelated link. With more data and add more relent data, we still believe the inlink building start at a right track to arouse more engagement level of customers. As two out of four SMEs told us that inlink bring them quite amount of traffic to their official website and is a chance for higher conversion to purchase than social media. Thus we conclude that the inlink service can be further implement with relevant factors and arouse higher engagement level for SM.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
106