Page 283 - Introduction to Electronic Commerce and Social Commerce

9.3 Market Research for E-Commerce 269

posts from its 30 million members. For more information products, spyware applications are typically bundled together
about Xiaomi and its social media engagement, see thenext- with freeware that is downloaded onto users’ computers.
web.com/asia/2014/04/09/xiaomis-social-media-strategy- Many users do not realize that they are downloading spy-
drives-fan-loyalty-books- it-242m-in-sales-in-12-hours. ware with the freeware. The best defense against spyware is
to install anti-virus software, which should detect and
Observing Customers’ Movements Online remove any viruses or other harmful intrusions.

To avoid some of the problems of online surveys, espe- Web Analytics and Mining
cially the reporting bias that occurs when people give false
or biased information, some marketers choose to learn Web analytics deal with the monitoring, collecting, measur-
about customers by observing their behavior rather than by ing and evaluating, and reporting tasks related to Internet
asking them questions. Keeping track of consumer’s online data and activities (e.g., see Provost and Fawcett 2013). Web
behavior can be done by using transaction logs (log files) analytics help us understand and optimize Web usage. Such
or cookie files. This allows activity-based filtering to be analysis is done, for example, by retailers for market research.
done. For example, see IBM Coremetrics (ibm.com/software/
marketing-solutions/coremetrics; now part of IBM
Transaction Logs Enterprise Marketing Management). A company can also
A transaction log (for Web applications) is a user file that use Web analytics software to improve its website’s look and
records the user’s activities on a company’s website from the operation. Web analytics can provide quick feedback from
computer log. A transaction log can be further analyzed with customers to help marketers decide which products to pro-
log file analysis tools (e.g., from Oracle) to get a good idea mote. For tutorials on data, text, and Web mining, see mydat-
about online visitors’ activities, such as how often they visit amine.com, tutorialspoint.com/data_mining/index.htm,
the site. and the video “Introduction to Data Mining 1/3” at youtube.
Note that, as customers move from site to site, they estab- com/watch?v=EtFQv_B7YA8.
lish their clickstream behavior, a pattern of their move- For details and methods used, see Clifton (2012). A spe-
ments on the Internet, which can be seen in their transaction cial type of Web analytics is clickstream analysis, or just
logs (see upcoming discussion of clickstream analysis). click analysis.

Cookies and Web Bugs Clickstream Analysis
Cookies and Web bugs can be used to supplement transaction- Clickstream data are data that describe which websites users
log methods. Cookies allow a website to store data on the visit, in what order, and the time spent on each. This is done
user’s personal device. When the customer returns to a site by tracking the succession of “clicks” each visitor makes. By
visited previously, the website can find what the customer did analyzing clickstream data, which can be maintained in a spe-
in the past from the cookie. The customer can be greeted by cial database or data warehouse, a business can find out, for
his(her) name, or a targeted ad can be sent to her(him). For a example, which promotions are effective and which popula-
comprehensive description of cookies, including examples tion segments are interested in specific products.
and privacy concerns, see the Indiana University Knowledge Several companies offer tools that enable such an analysis.
Base (kb.iu.edu/d/agwm). Cookies are frequently combined For example, Analytics 10 from Webtrends, Inc. (analytics.
with Web bugs that are tiny (usually invisible) objects con- webtrends.com) features several advanced tools for analyz-
cealed in a Web page or in e-mail messages. Web bugs trans- ing clickstream data (webtrends.com/solutions/digital-mea-
mit information about the user and his or her movements to a surement/streams). Finally, clickstreamr.com configures
monitoring site (e.g., to find out if the user has viewed certain Google Analytics standards and can be used for such analysis.
content on the Web page). Many believe that cookies and
Web bugs are an invasion of a user’s privacy. Web Mining
Web mining refers to the use of data mining techniques for
Spyware both Web content and usage in Web documents in order to
Spyware is software that enters your computer like a virus discover patterns and hidden relationships. Web mining has
does, without your knowledge. It then enables an outsider the potential to change the way we access and use the infor-
to gather information about your browsing habits. Originally mation available on the Web. For mining the social web, see
designed to allow freeware authors to make money on their Russell (2013).