CHAPTER 2 LITERATURE REVIEW
2.3 S EARCH E NGINE O PTIMIZATION
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
14
Table 2.1 the brand partner relationship table
Partner relationship Definition
Co-advertising
Co-advertising is defined as the partnership that firms jointly promote their products together through advertising for achieving brand awareness and knowledge. The duration of the relationship normally last 3-4 months.(d’Astous et al., 2007)
Cause-related marketing
Cause-related marketing is defined as the partnership that improves a company’s corporate social reputation with a cause through partnering to introduce a new (or existing) product (Dickinson & Barker, 2007). For example, Apple accompanies with Motorola hit a success with Product Red through funding hundreds of millions to AIDs globally.
Co-branding
Long term alliances between two or more brands to create new product or service for entering an existing or new market. The identity of the co-branded brands is
communicated through the inclusion of the brand names on the product or service (Walchli, 2007).
2.3 Search Engine Optimization
To help the firm start the brand partnership and engage customers with search engine, we reviewed the existed search engine optimization methods. As we described in section 2.3, search engine optimization have these four methods, including keyword research, indexing, on-page optimization, and off-page optimization (Solihin, 2013). In this study, we will focus on keyword research
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
15
and off-page optimization (link optimization), because these two are the core elements for optimization that have many discussed papers of methods.
In our study, Link optimization is an optimization problem on PageRank.
PageRank (Page et al., 1999) is one of the most popular search engine ranking algorithms through link structure which is originally used by Google. PageRank can be divided into two parts to simulate a random web surfer, including personalization part and surfing through link part. Personalization part determined from individual, mimicking the behavior that user enter an url directly. For surfing through link part, determined from the behavior that user surfs the web by clicking through link in the site. For optimization purpose, these two parts should both manipulate and optimize through building links to relevant content and forming a better link structure.
Keyword research in our study is also an optimization problem that we want to provide service to give keyword suggestion for focal firm and its brand partners so that target customers can easily find the brand alliance-based campaign content on engagement site and co-create CEB value. We review different techniques of keyword research, including meta-tag spiders, proximity-based technique, query-log mining, advertiser-log mining, TermsNet, and using GA method to optimize query (See Table 2.2). We found that except GA method the keyword generation methods need a seed term to do the generations of keywords. However, most of the time marketers may not sure which keyword seed term to use. On the other hand, GA method can accept user-defined description as input and suggest the keywords based on the search results which will change through time. Additionally, GA method provides the optimization of the query so that through the method can generate ranked query which fits our optimization in search engine perspective. There is one setback of
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
16
the GA method which generates query that only a few people using these kind of query, since the GA method is originally developed to generate diverse and long query to search for topical documents. For example yahoo site wants to add the documents of different categories, they can use this method. To solve this setback, we will limit the query length.
Table 2.2 the brand partner relationship table Keyword generation
techniques
Definition
meta-tag spider
A meta-tag spider queries search engine for seed keyword and extracts meta-tag words from these highly ranked webpages.
Proximity-based technique
Proximity-based tools issue queries to a search engine to get highly ranked webpages for the seed keyword and expand the seed with words found in its proximity.
query log mining
The Google Adwords Tool (https://www.google.com/adwords/) relies on query log mining for keyword generation. Pick up the query that has the seed term and generate the query that include the seed term.
advertiser-log mining
To avoid only generating the query that includes the seed term.
Google Adwords (https://www.google.com/adwords/) also consider mining through the advertiser search logs to add more variety.
TermsNet
Including the relationship between words through top search result snippets’ word connection and then output the ranked keywords (Joshi & Motwani, 2006).
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
17 GA method
A technique through user defined description to select good query terms through web search result by genetic algorithm. By crossover and mutation to generate query diversity and
optimization through iterations of matching similarity between user-defined description and search result documents (Cecchini et al., 2007).
As reviewed through the CEB (section 2.1), we know that the five dimensions of CEB which will further being used in development of the CEB measurement matrix (section 4.4), how the CEB could be stimulated and form the virtual positive CEB cycle in the New media framework which will be further developed in our conceptual framework.
After that we further extended the virtual cycle through CEB co-creation in the ways of brand partnership (inspired from the notion in new media framework’s brand attitude). Reviewing the different levels of brand partnership and derive insight from a patent’s framework to heat up the bigger cycle with brand partners.
At the search engine optimization section, we dig down with existing method of search engine optimization. With the understanding of the keyword research and link optimization, we can then develop the practical implementation on the engagement site to build the CEB positive cycle in the perspectives of search engine.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
18