Type 2: Page-Based
Visit-based segmentation creates groups based on something specific to the visit. Sometimes we want to know about the visits where people saw a certain page or performed a certain function. We may want to know about people who arrived at our site via a marketing vehicle. These are all things that can change from visit to visit.


  • Campaign-spawned visits.
  • Visits where they were/were not logged in on the site.
  • Visits where they converted (purchased, signed up, etc).
  • Visits where they saw key content (such as a specific promotion page).
Type 3: Page-Based
For many people, visitor-based segmentation is the Holy Grail of Web analytics. The better you understand your audience, the more equipped you’ll be to appeal to them and win their loyalty. These segments are typically characteristic of the visitors who don’t change from visit to visit but who help us understand the visitor and his or her needs.

  • Registered vs Non-registered.
  • Age.
  • Gender.
  • Occupation.
  • Custom Loyalty.
  • Geographic Location.
Segmentation Example: Splitting Up the Traffic
Let’s say theta we have a site that had 10,000 Page Views yesterday. That’s great to know, but as we have stated, it is not the end goal. We want to know to whom and to what we can attribute those page views. Which pages are getting that traffic? Who or what is responsible for these page views?

1. Kinds of Visits.
2. Group of Pages.
3. Types of Visitors.

My site = 10,000 Page Views.

Logged in: 6000 Page Views.
Not Logged in: 4000 Page Views.

Store Pages: 7000 Page Views.
Info Pages: 3000 Page Views.

Customers: 2500 Page Views.
Public Visitors: 5000 Page Views.
Trial Users: 2500 Page Views.

As you can see in this example, these 10,000 Page Views can be split up many ways.

Visit-based info: If we have our site coded to determine whether a visitor is or not logged in on their visit, then we can determine how many page views came from people who were or were not logged in on our site. In this case, we have 6,000 Page Views from people who were logged in, and the remainder from those who weren’t logged in.

Page-base info: In addition to Page Names, which are segments themselves (very granular segment), we may want to put these pages into groups in order to see how much traffic comes through different areas of our site. In this case, we had 70% of our traffic to the Store Pages, and the rest on Info pages. This could indicate a lot of revenue.

Visitor-based info: Last, but not least, we have different kinds of visitors on our site. We could let people try out service before they sign up as a customer. Doing this allows us to see how much traffic comes from customers, trial users and the general public.