What is a segment?
A segment is a set of visits that are similar to some extent, e.g., they have the same traffic source, the same page visited, or have a basket value higher than X. If you are familiar with Google Analytics, you should already know that. There is however something unique to segmentation in UseItBetter, as it allows you to find users who interacted with any element on the page: clicked something (even if it’s not clickable), edited a form field, or hovered over a help icon.
That allows you to create a very precise behavioral segment for your analysis.
Segmentation by Section Name
Segmentation by Section Name is the most basic type of segmentation. A section is like a page and usually is based on a page URL. You can read more about this here.
Start your analysis by creating a segment of visits in which the user visited a section (page) with your forms.
If you want to analyze forms on:
create a segment with
to make sure that within all the visits in your report, users actually went to this page.
When you (1) select the created segment and (2) open the Forms Report, the system will try to load 5000 visits matching its criteria. Looking at (3), the sample details, you can determine if it is a full 100% of captured traffic or just a sample, and what is the exact time frame of the analyzed data. The default sample size should be enough to get started, but you can (4) increase the sample size if needed.
Improving The Sample Quality
Increasing the sample size isn’t always the best option. With increased size, the loading and processing time will increase too, but the quality of the sample may not improve..
Before you start increasing the sample size, you should try to find a segment that is more suitable for your analysis.
Some basic approaches include:
Split traffic by device, interface and screen width
Use the (A) screenWidth attribute to analyze forms on different screens:
If you have a responsive site with multiple templates for different screen width ranges, you can use:
You can combine the (A) screenWidth attribute with the (A) interface attribute to differentiate between tablets and computers:
(A) screenWidth>1000 (A) interface=touch
There is also a (A) device attribute that allows you to segment visits from recognized types of devices including smartphones, tablets, and also TVs, fridges, and other devices falling under the category of the Internet of Things. For example:
Split traffic by referrer, source, affiliate partner
The information about where users came from is typically present in one of two attributes:
(A) referrer=http://www.linkedin.com/abc (A) referrer=http://www.linkedin.com/xyz
which you can use for segmentation, typically using wildcards:
If you are tracking only a part of your website (i.e. checkout) and you want to analyze visits based on a traffic source that you acquired from pages earlier in the process (i.e. product pages) , you would need to keep the traffic source information in your data layer and send it to UseItBetter when a user enters the checkout process:
Segmenting by Progress
When you analyze forms using a segment based on visitor’s profile or by a section (page) they visited, you will notice that there are plenty of visits in which users dropped out (1) without editing any of the form fields. While this information is very important, you might be interested in analyzing just the visits of users who actually started filling in the forms. That leaves you with a reduced sample, which is then further reduced when you start analyzing (2) the visits of users who dropped out from the funnel.
If your form consist of multiple fields, you will notice in the (1) “Changed in visits” column that you lose some visitors after each of them. That again reduces your sample size if you want to analyze the visits of users who progressed further down the funnel.
In such a case, you can (2) hover over a form field and use it to create a segment of users who changed the field value. When you click the field to create a segment, you will see a model with a new segment definition.
Add an easy to understand (1) label to it, make sure that it’s added as a child of your (2) main segment, and click (3) apply to create the segment. Once your reload the Forms report for that segment, all the visits in the default sample will have that label.
Of course, if your funnel has multiple steps and each of these steps have a separate section, you can use them to create segments:
- Section=Step 1
- Section=Step 2
- Section=Step 3
Exact (non-sampled) metrics
The Forms report was designed to explore data, so that you can identify potential problems and opportunities related to your forms.
If your objective is to measure how many times a certain field was edited or how your funnel metrics change over time, you should create, within your main segment of (1) visits in which users went to a page with forms , subsegments (2) based on form field change events (as described above) or other events (i.e. Section, Action, clicks…). You should then use the (3) Segments report.