Abstract:
There is a growing
reliance on the Internet and Web sites for promotion of tourism.
Evaluation of Web site effectiveness is necessary because of significant costs
for setup, advertising, and maintenance. Research suggests there is a great need
for evaluations to go beyond hits and page viewings. The objectives of a study
were to develop and apply a low-cost, automated, Internet-based survey
methodology to investigate effectiveness of a tourism promotional Web
site. Both online and email surveys were used to gather data from visitors to a
state Web site. Although responses were received from 833 persons, response
rates were low. A check for nonresponsive bias indicated that respondents were
different from nonrespondents. Respondents primarily learned about the Web site
through two sources and reported that the Web site significantly influenced
their travel plans.
Summary:
Cost:
Evaluation of Web site effectiveness is necessary because there are significant costs involved in setup and maintenance of Web sites ($180,000/year according to the Association of National Advertisers 1999)
Measurement of effectiveness:
Although
direct online e-commerce sales are one good indicator of effectiveness, many
tourism Web sites do not directly sell services and cannot use this as a gauge.
Tourism companies that sell complex products, such as tours, have noted that
visitors often only gather information from a Web site and then call or write to
request a brochure or make a reservation. In this case, the Web site may have
had a powerful role in the purchase decision, but the sale itself did not take
place over the Internet. Therefore, additional research is needed to fully
determine Web site effectiveness in promoting tourism services.
As the 1999 Tourism
Industries of America study showed, the most common statistic used to measure
Web site effectiveness is the number of hits to the site. The term hits is
commonly defined as the number of files that are downloaded from a Web server
(host computer). Tracking hits as a way of measuring traffic can be misleading.
The number of hits a site receives is usually much greater than the number of
actual visitors because a Web page usually contains more than one file. More
reliable is the statistic called "user sessions," defined as "a
session of activity (all hits) for one user of a Web site" (WebTrends,
2000). Web traffic software, such as WebTrends (see http://www.webtrends.com),
can provide information on site visitors' domain name and type, region or
country of request, organization type (com, net, org, etc.) ad viewings
(clicks), and time of view. However, this type of software cannot give critical
information about the user, such as income, reason for visiting the site,
satisfaction with the site, or actions taken because of viewing the Web site.
Clearly, to fully understand the impact of a Web site, more information than
just the number of hits or user sessions is needed.
This is an extremely important point, with software like this it could actually mean that you could derive useful information from your web-site as opposed to hits. This also begs the question of cost.
It is possible for host computers to use coded messages sent to and stored in the viewers browser, called "cookies," to identify if a browser has visited a Web site previously and track respondent preferences and page viewings. But cookies cannot identify users, and sophisticated users simply turn off the cookies option in their browser.
Survey technique:
Online (Email surveys) or Web-based data collection surveys
Use of incentives with Internet surveys has been shown to cause at least three methodological concerns: response bias, multiple entry, and unwanted entries. Probably one of the most serious concerns about Internet based surveys is nonresponse bias.
This was shown to be a major downfall of this research (the low 2.0% response ratel) and indeed a major problem for all organisations that wish to get feedback. This means that it is irrelevant about the fact that the internet can target so many people, and despite the 2% response rate, you can still get enough data, because the respondents differ so much from non respondents.
This is of course the danger with internet marketing, the costs that are involved (as outlined at the beginning of the article are significant when you consider to problems with feedback for this I am not including the programs which provide this due to the limited value of the information.
Online surveys have a potential advantage in the ability to acquire large sample sizes at a very low cost and in a short time. This is due to the automation of data gathering, and entry and compilation of data into descriptive statistics. Internet survey software is available and takes the written survey and converts it into email or Web-compatible formats, emails or administers online surveys, and then automatically collects responses, enters them into a database, and calculates descriptive statistics. Using this software on a busy Web site, it is possible to collect hundreds of responses in a short period.
Registration of visitors, wherein they must provide a minimum of data about themselves before they are able to proceed to the rest of the Web site, is one approach being used to gather demographic data. However, Web users are so concerned about privacy of their personal information that registration has been shown to discourage sites visitation (Neilson 1999).
Web
users are so concerned about privacy of their personal information that
registration has been shown to discourage sites visitation (Neilson 1999).
*
All text in red indicates quotes taken from the full
text article.