This article illustrates how to use, modify, and create new segments to analyze your data.
In this article:
Compare Converters and Non-Converters
System segments are designed to cover a broad range of common use cases. In this example, we’ll use two system segments, Converters and Non-Converters, to compare the two things every site or app owner cares about: users who convert (complete goals and/or transactions) and users who don’t.
Understanding the users who convert helps you refine the successful aspects of your marketing, and shows you where you can improve your efforts to reach users who demonstrate untapped potential.
Developing insights into why users aren’t converting lets you address the weak spots in how you approach them.
For this first example, let’s apply the two system segments Converters and Non-Converters to the Audience Overview report to see how those segments provide a new look at the data.
You can see that there are fewer users who convert (20,042 vs 54,212), and those users conduct fewer sessions (27,788 vs 59,080).
However, even though Converters account for less than half the traffic to the site than Non-Converters, they account for much more of the activity on the site:
- Over 4 times the pageviews
- Over 7 times the pages per session
- Over 10 times the average session duration
- About ⅙ the bounce rate
Not necessarily a surprise that users who convert are more engaged, but the data indicate that once you’ve gotten converting users onto your site, they become highly engaged. And they come back: a little over a third of all Converters are returning users.
When you apply segments, they remain in effect for all your reports, which makes it easy to evaluate the data in any number of different contexts.
Open the Demographics > Age report.
Notice that the ratio of conversions to non-conversions decreases steadily with age.
The 25-34 group accounts for the most sessions with conversions (6312 or 42.65%), and is the group most evenly split between sessions with conversions and sessions without conversions (6312 vs. 6886).
It could be that users 25-34 are more naturally inclined to purchase online, or that your products or services have a particular appeal to this group. Or maybe your marketing focuses more on younger users. These statistics could also indicate that your site is geared to a younger, more technically adept audience. You can’t tell exactly what’s at play from just this information, but there’s evidence so far that the youngest demographic is especially valuable to you, and that the other demographics may be less valuable as age increases.
The two age groups 25-34 and 35-44 together conduct over half of all sessions (56.9%), and account for an even higher percentage of conversions (64%).
Notice that the 65+ age group accounts for only 5.1% of all sessions, and only 3.5% of all conversions, which would appear to reinforce the value to you of the younger demographic.
However, while the ages 25-44 account for 64% of conversions, users 65+ have a higher conversion rate: 0.77% versus a combined rate of 0.63% for users 25-44. Fewer conversions from 65+, but there seems to be some strong, untapped potential in that group.
Open the Demographics > Gender report, and set the Conversions metrics to All Goals, and see whether gender groups convert differently.
Male users account for 2.5 times as many sessions with conversions as female users (12,011 vs. 4,756).
However, if you look at Goal Conversion Rate, you see that while fewer numbers of females convert, their Goal Conversion Rate is a little higher than men (84.13% vs 83.56%).
With the simple application of two system segments, you can navigate through a few reports and start to see patterns emerge:
- Younger users account for more total conversions, but older users have a higher conversion rate.
- Male users account for more total conversions, but females have a higher conversion rate.
While this initial investigation may not provide sufficient justification for changes in how you allocate resources, it does provide direction for further investigation.
For example, create segments for each age group and gender, and apply them to your Campaigns reports to see whether your marketing is appealing primarily to one group. If your marketing has an unintentional, narrow appeal, you can create additional campaigns and ads geared toward those groups who show potential but who aren’t responding well to your current marketing (for example, women or users 65+).
Apply those same segments to the Geo Location report to see whether there are locations in which you are not running campaigns, but that have higher ratios of those users who represent a lot of potential.
Apply those segments to the Interests Overview report to see how widely interests vary among groups, and whether you need to develop more specialized audiences for your programmatic ad buys.
Once you’ve made that initial discovery of meaningful data (for example, the groups that represent a potential source of conversions), you can create the corresponding segments, apply them to your reports, and conduct a thorough analysis to see what kinds of new efforts and allocations of resources you can make to take advantage of that insight.
Analyze Sessions with Conversions from a specific geographic area
In this example, we’ll copy and modify the system segment Sessions with Conversions.
Start with the Audience Overview report, and apply the Sessions with Conversions segment.
Remove the All Sessions segment so you can focus on just sessions in which users completed conversions.
With that single system segment applied, you can look through your reports to see if there are subsets of that data that might be interesting, for example, geographic regions that have a relatively high number of sessions with conversions. Open the Geo > Location report.
In this case, the United States has more than 10 times the number of sessions with conversions than the next most successful country.
You can copy and modify that original segment to add additional filters so that you can examine subsets of that data (for example, Sessions with Conversions from United States). With that more narrow segment applied, you can navigate through your reports with a focus on just that subset.
At the top of the report, open the menu for Sessions with Conversions, and click Copy.
The original segment definition opens in the segment builder.
Click + Add Filter to add an additional condition filter that limits the segment to just sessions with conversions that originated in the United States.
Give the new segment a name that means something to you (e.g., Sessions with Conversions - United States).
Click Save.
Remove the Sessions with Conversions segment so you can focus on just conversions in the United States.
From here, you can open any of your other reports and examine just this specific subset of your data.
Given the high rate of conversions, it would be helpful to understand what kinds of users comprise this audience.
Open the Demographics > Overview report.
You can see immediately that most of the users who convert are 25-34 and male.
Open the Interests > Affinity Categories report.
While there’s a fairly even distribution among the top-10 interest categories, Technophiles, Movie Lovers, and TV Lovers are the most popular.
Open the Technology > Browser & OS report.
For these users, Chrome is by far the most popular browser.
With a minimum of configuration and just a few clicks, you can focus on meaningful aspects of your data, and start to develop a real understanding of who are the most valuable users in that segment. In this example, it is Males 25-34, with interests in Technology, Movies, and TV, who use the Chrome browser to initiate sessions from the United States. Armed with this kind of information, it’s easy to build audiences for your marketing efforts that are targeted to your most responsive users.
Create a high-value-user segment
In addition to using system segments in their default configurations, or making modifications to them, you can also create your own custom segments to focus on any data in which you’re interested.
The most valuable insight into your users is who among them are high-value in terms of your business: the ones who have interacted with your content or purchased recently, who interact or purchase frequently, and who engage in high-value conversions.
You can create a Recency-Frequency-Monetary Value (RFM) segment that identifies those users.
Recency: Users who have interacted with your content or purchased recently (for example, within the last two days or last week) are more likely to interact or purchase again.
Frequency: Users who interact or purchase frequently (for example, every week or month), as well as recently, are more likely to interact or purchase again.
Monetary Value: Users who engage in the most valuable conversions, along with having converted recently and frequently, are more likely to convert again.
You need to specify the RFM thresholds that identify your high-value users.
To create an RFM segment, base it on filters like the following:
Behavior
Days Since Last Session < 5 (recency)
Sessions > 5 (frequency)
Ecommerce
Revenue per user > 100 (monetary value)
Conditions > Filter Users
Goal Completions per user > 10 (monetary value)
Goal Value per user > 10 (monetary value)
As in the previous examples, you can build this kind of segment and then navigate through your reports to see which users are included (for example, which countries/states/cities, which demographics, which technologies, which channels), and then develop your audiences and marketing around that data.
Create a cohort segment
You can build segments to identify cohorts, for example, new users to your site on a specific date or during a specific date range who arrived as the result of a specific campaign. Use filters like the following:
Date of First Session: the date range of your campaign
Traffic Sources: Campaign exactly matches name of your campaign
With cohorts, you can follow the behavior of the same set of users over time. For example, you can create cohorts based on campaigns, and follow those users over a period of weeks or months to see how quickly and to what extent those users converted, and how long they continued to convert. If you notice that the lift from your campaigns is sustained longer than you expect, you might decide to run fewer campaigns. If there’s a regularity to the lift and drop off, you can use that information to start your new campaigns as the effects of the previous ones start to subside. You can also make direct comparisons of campaigns to see which of them are more effective in terms of overall conversions and revenue, and which of them have the most sustained effects.
Create a potential-purchaser segment
One group of users you want to identify and reach out to again via remarketing are the ones who started down the purchase funnel but didn’t complete the process, for example, users who added items to their carts but never completed their purchases.
To identify these users, create a segment with Conditions filters like the following:
- Users > Include
Page contains ProductDetails - Users > Include
Event Action exactly matches AddToCart - Users > Exclude
Page exactly matches ThankYou.html
This segment matches users who viewed product-detail pages, clicked Add To Cart, but never viewed the order-confirmation page that is always displayed at the end of an order, indicating that their orders were never completed. Given that these users have indicated a strong interest in purchasing, they’re the perfect audience to approach again with a remarketing campaign.
Next steps
After you create these segments of your users, you can create audiences from the segments to use in your marketing efforts.
