How do I interpret the Site Effectiveness standard report?

Environment:

Analytics, Site Effectiveness

Resolution:

The sections of the Site Effectiveness report are described below, including some tips on how you can use this report to determine how many of your end users are finding their own answers.

The Site Effectiveness report compares the activity of users who viewed answers on your site to those who did not. This report contains three sections: Visit Activity, Question Submittal, and Visit Comparison reports.

Visit Activity: This section indicates how many of your end-users are using the self-service portions of your site, including viewing published answers, performing searches on the answer page, and submitting questions. The percentage listed as the Self Service Rate is the percentage of visits that did not include an Ask a Question submittal.

Question Submittal: This section provides statistics that are relevant to the customer visits that included an Ask a Question submittal. This section displays the number of questions submitted and the number of questions with a confirmed submittal, and tracks how many of those performed each one of these four actions.

When an end-user clicks the Submit Question button on the Ask a Question page, the number in the Submitted column of the report increases by one. Then, when the end-user clicks the Finish Submitting Question button, the number in the Confirmed column increases by 1. The statistics in each of these columns reports on user activity prior to submitting those incidents.

By looking at the difference between the Submitted and the Confirmed numbers, you can see how many users did not continue to submit their incident for any reason. Reasons may include that they found their answer in a Smart Assistant Suggested solution or that they closed their browser without finishing their request.

The Resolved column shows how many end-users were deterred from asking a question. This number is calculated by subtracting the number of confirmed questions in each visit type from the number of submittals in each visit type. This number is then divided by the number of submittals in each visit type. For example, if 10 of the visits that viewed an answer also clicked the Submit Question button, but only 8 of the visits actually submitted an incident, the percent deterred would be 20 percent (10 - 8 = 2, 2/10 = 0.2).

Separate statistics are listed for the type of actions performed in the visit as listed below:

	Question Only: The end-user visited the Ask a Question page without searching or viewing an answer.
	Answer Only: The end-user submitted a question after viewing at least one answer, but did not perform a search.
	Search Only: The end-user submitted a question after performing at least one search, but did not view an answer.
	Search and Answer: The end-user submitted a question after performing at least one search and viewing at least one answer.

Visit Comparison: This section compares the action of end-users who did not submit an incident (the Self Service column) and those end-users who did submit an incident (Non-Self Service column).

For each type of visit (self-service and non self-service), the report lists the percentage of each type of visit that included a visit to each page. For example, if 10 visits that did not submit an incident occurred during the specified time period, and 5 of those visits visited the Support Home page, the percentage of self-service visits that visited the Support Home page would be 50 percent.

Note: Single-page visits that only visit the Support Home page are not included in the statistics reported in the Self Service column of this report.

This report also calculates the percent difference between each of the percentages. The percentage is calculated by subtracting the non self-service number from the self-service number and dividing by the non-self service percentage. For example, if 70 percent of self-service visits visited the answers list and 50 percent of non self-service visits also visited the answers list, the difference would be 40 percent (70-50/40).

What the percentage difference means:

1. If the number is a big positive number (e.g. something close to 100%), the trend is that people visiting that page tend to not ask questions (self-service).  This is expected this for the Answer Detail page, assuming the people asking questions do not look at answers.
2. If the number is a small negative number (e.g. something close to -100%), the trend is that people visiting this page tend to ask questions (non-self service).  This would be expected for the Ask a Question page.
3. If the number is right around 0, either positive or negative, there is not much difference between the two visit types (those that do vs. do not ask questions visit the page about the same rate).

This data can help you quickly determine which pages are viewed more often by self-service customers than non self-service. For example, you can see how often non self-service end-users are viewing an answer in comparison to how many self-service users. An effective site would show that a high percentage of end-users are viewing an answer without submitting an incident.