Metric calculation

The Wakoopa dashboard tracks web usage of participants and gives you insight into the gathered data through a set of metrics. This page tells you how the Wakoopa dashboard gathers its data and how this data is made available to you.

Channels and Adresses

In the Wakoopa dashboard you can define channels and addresses to analyze (parts of) websites. For the examples on this page we have defined two channels:

Pageviews

The source of the data that the Wakoopa dashboard collects are the pageviews that are collected by the trackers. A participant will open a URL at a specific time and views the URL for a number of seconds before navigating to the next URL.

These 9 pageviews will be used as the data to explain the following concepts.

Visits

Our definition of a visit is:

• a series of pageviews
• within a group of URLs
• from the same participant
• with a time of no more than 30 minutes between each pageview

Calculating the visits will result in a list of visits with each visit having a:

• participant
• group of urls
• time of visit
• total number of seconds
• total number of pageviews

An important part of the definition is the “group of URLs”. These groups are defined by a domain, address or channel. Naturally this results in domain-visits, address-visits and channel-visits.

Domain Visits

Domain visits are calculated by default in the dashboard so they can quickly be compared in the Profile-feature. Applying the visits-definition on our example of pageviews we get the following domain-visits:

Address visits

These visits are calculated on demand when creating a new channel.

Channel visits

Metrics

Metrics are calculated using the visits data. The set of visits that is used to calculate a metric is determined by the filter and the period.

Filtering only selects visits from participants that match certain variables. For instance, only visits from males between 20 and 30 years old.

Selecting a period will only select visits for which the time of visit falls into the selected period. For instance, only visits in december 2010.

Based on this set of visits the following metrics are calculated:

• Unique visitors
  Unique number of participants

• Time on site
  Sum of seconds

• Total visits
  Number of visits

• Total pageviews
  Sum of pageviews

• Average depth of visit
  Sum of pageviews / Number of visits

• Average length of visit
  Sum of seconds / Number of visits

• Bouncerate
  Percentage of visits that have exactly 1 pageview

Example

As an example here are the calculated metrics for all the domain visits:

Sites before and after

Sites before and after is an overview of all the sites preceding a specific domain.
(Support for addresses and channels will be rolled out in Q1 2012.)

The metrics are calculated by looking at pageviews and their preceding or following pageviews.

A collection of pageviews by the same user is called a session. If a session times out, a new session will start. The threshold for a session timeout is currently 5 minutes.

For instance, if a pageview of cnn.com was preceded by a pageview on yahoo.com, cnn.com will get a yahoo.com as a “site before” entry. Yahoo.com will get cnn.com a “site after” entry. If there is no preceding or following pageview, there will be an “direct” (site before) or “unknown” (site after) entry.

An entry in the list can be one of three things:

• Domain (e.g. cnn.com)
  The pageview that preceded or followed a pageview on the selected domain was on a different domain.

• (Direct) or (Unknown)
  If no pageview is preceding a pageview on the selected domain, the entry point for that pageview is considered “direct”. If no pageview is following a pageview on the selected domain, then that pageview is considered “unknown”.

• Internal
  The pageview that preceded or followed a pageview on the selected domain was on the same selected domain.

Example

As an example here are the calculated sessions for the all the example data:

And the sites before and after table for store.com and news.com:

store.com

news.com