The power of the Clickstream to Produce Internet Metrics
· It can tell how many orders it has lost as a result of the out of-stock condition.
· It can tell that visitors must look through several pages to find the desired item. One way of ascertaining this is to see that they are following circuitous paths through the site instead of going straight to the product.
· It can capture customer names, one of whom can be a gift-giver and one of whom, based on current knowledge, can be a fly fisher.
Uses of Metrics
1. The data point to possible usability problems that need to be examined and corrected.
2. The data are used to generate site effectiveness measures that are essential for a number of marketing purposes:
a. Managing the site
b. Determining advertising rates for the site
c. Understanding the impact of marketing programs like e-mail and online advertising campaigns
3. They provide a mother lode of data that can be mined to develop an in-depth understanding of visitor behavior on the site, which will in turn help improve the site with new features and marketing programs.
4. Since internet users are becoming aware that these data exists, they are concerned about the loss of privacy, even though they may not understand the results of its use by Marketers
Multiple Perspectives on Effectiveness
· Usability/Usability testing—the way visitors look at the site, the way they gauge its ease of use and value to them.
· Traffic and audience measurement—a set of techniques used to provide effectiveness data vital to marketing management.
· Site performance—involves data that are needed by site technicians to gauge and improve site performance.
What is the Purpose of Usability Testing
· It is exclusively designed to see if the site works in a user-friendly fashion according to the expectations of members of the target market.
· Ensuirng that it is easy for visitors to navigate and, in general, to find desired content quickly and efficiently on a web site.
· Essentially qualitative, and they are performed by target site users, not by technicians.
Stages of Usability Testing
· Concept testing—the earliest stage and reflects none of the actual site programing. In testing at this stage, one or more concept boards are shown to respondents, such as focus groups, who critique it from the perspective of how logical they perceive it to be and how easy they think it would be to use.
· Prototype Testing—is the second level where the site design is complete and at least some parts of the site are fully functional. Testing a prototype affords an opportunity to get reactions to the appearance of the site and to get some information about the degree to which the site structure is consistent with customer expectations.
· Full Usability Testing—it indicates that they site has been uploaded to a server and is fully functioning, even though it is not accessible to the general public.
Enterprise Metrics: Site Performance
· IT specialists maintain and tune site
· Marketers must be aware of site performance > visitor experience, because the traffic and audience data information has direct relevance to marketing programs
Enterprise Metrics: Business Performance
· Traffic data describe activity on the site. These data include metrics such as number of visitors, sessions, and page views.
· Audience data describe both the behavior of people on the site—where they come from, what paths they take through the site, and whether they take desired actions—and the people themselves using both anonymous and identified profile data.
1. Hit Counters (Site Administered)
a. A small piece of software that can be added to a web site to provide a few basic metrics. A hit counter is a few lines of HTML code that are placed on each page in the site that the owner wants to measure.
2. Server Request Log Files (Purchased Service)
a. Are created by the server that houses the web site
b. Each time a browser requests a file in order to build a web page, it generates an entry in the server request log.
c. Each graphic on a web page is a serparate file, and there can be several content files on a single page.
i. If for example, a web page that has 3 graphics and one content blocking, each time that page is accessed, four hits are registered on the server log. A hit is counted for each file on the page.
d. Server Request Log Records:
i. The IP address of the requesting computer
ii. Date and time of request
iii. Code indicating whether request was successful or not
iv. Number of bytes of data transferred
v. Referring site
vi. Type and version of web browser making request
vii. Operating system of computer making request
3. Coded Pages—Tracking web site visitors (Purchased service)
a. Also known as data tag, pixel tag, transparent GIF, we bug, web beacon, and others.
b. It’s a new method of collecting data, on the “browser(user’s) side” of the information transaction instead of on the “server side.”
c. The typical file used to implement coded pages is 1 pixel by 1 pixel, and it is placed on the web page itself, where it usually is invisible
d. The tag is part of the page that is requested by and served to the user’s browser. The file then sends data back to the server.
e. Working with cookies, the tag can send back detailed information about the visitor’s activity on the page. On the surface, the data are the same as those collected by server log.
f. Advantages over server log data:
i. Data may be more accurate.
ii. Data collected by web bugs has smaller storage requirements so they only store relevant data.
iii. When data are collected in this manner, it permits real-time processing and viewing of site metrics.
4. Collecting Panel Usage Data
a. Similar to the panels in the general marketing research process
i. Size of the universe to be sampled
ii. Segments that are to be broken
b. Specialized software is downloaded onto the participant’s computer to record the clickstream data.
c. Benefits over server logs:
i. The source of the data is unambiguous, source is known, which it often is not with server logs.
1. International IP address can be difficult to identify
ii. A person who uses the internet from both home and work can be identified as same person through carefully planned and maintained panel data.
iii. The measurement firm can collect demographic and behavioral data from panel households
iv. Produces longitudinal data
5. Basic Metrics
a. Traffic—measures simply document site activity:
i. Hits—the number of files requested
ii. Impressions—the number of times an ad banner is requested by a browser
iii. Page views—the number of times a web page is requested
iv. Sessions—the amount of activity on a site during a specific period
v. Click-throughs—the number of times visitors come to the site by clicking on an ad
b. Audience—measures provide data about the people who visit the site:
i. Visitors—the number of people who visit a site
1. Total or unique during a specified time frame
2. Unidentified or identified (registered or customer)
ii. Unduplicated audience—the visitors that are unique to a web site
iii. Behavior on the site:
1. Number of page views
2. Session time
3. Path through the site
4. Shopping cart abandonment
5. Entry page (many visitors do not enter through the home page)
c. Campaign—measures provide data about the effectiveness of marketing efforts
i. By communication channels: e-mail, mail, online banners, etc.
ii. By offer:free shipping versus 25% off, for example
iii. Search effectiveness by keyword
d. Common metrics follow:
i. Average number of visits per day
ii. Number of page views
iii. Average visitor session length last month
iv. Number of hits for each hour of the day
v. Paid search results for the most recent 7-day period
vi. Etc.
6. Traffict, Audience, and Campaign Metrics
a. Report Dashboard
i. Type of presentation that allows marketers easy, real-time access to important reports directly from their desktop.
ii. The metrics report page visits, number of visits, the percentage of all visits, the average duration of the visits, the average number of pages viewed per visit, and the revenue produced by the set of visits
b. Marketing Campaign Reports
i. number of reports can be available in a single category:
1. calendar feature
a. allows the marketer to select any day, week, month, or other specific time period for the report
b. produce comparative reports by selecting two time frames for comparison of a particular reporting item
c. Browser-Based Reports
i. Report shown in browser format
ii. Data that are overlayed on the page represent the percentage of visitors who click through on each link on the page.
iii. Data shows number of visitors who see the page, how much time they spend, where they came from, and where they go when they exit the page.
iv. Another panel shows the specific sites from which the visitors come, search engines, affiliates, newsletters, and so forth
d. Path Analysis
i. Shows the basic steps that would result in successful task completion
ii. The center arrow shows the basic steps
iii. The data to the left show the entry points
iv. The data to the right show the exit points
v. People who are backtracking are experiencing navigational problems
vi. Visitors who exit may still be looking for the same product
e. Segmenting by Behavior and Channel
i. When segmenting visitors, by behavior pattern or by information provided by the visitor, is important to many types of marketing decisions
ii. When there is a difference between the pages most viewed by demographic groups—say men, women, and teenagers—the site will be able to charge more for advertising that is able to target a specific demographic.
iii. It can also show multichannel behavior, shown in the form of a “conversion funnel”
1. Shows the number of potential customers who begin the purchase process, the number who remain at each stage in the process, and the number who complete the purchase task.
7. Exposure versus Action
a. CPM (cost per thousand advertising impressions)
i. Primary pricing mechanism for many media, including the internet
b. CPA ( cost per action) or CPC (cost per click)
i. Cost per click-through
ii. Cost per order (synonymous with cost per sale and cost per conversion)
iii. Cost per lead
iv. Cost per qualified lead