What Is a Segment?
A segment is a filter applied to your analytics data that isolates a specific subset of users, sessions, or events for examination. Rather than analysing your entire audience as one undifferentiated mass, segmentation lets you ask more precise questions: how do mobile users behave compared to desktop users? Do visitors from paid social convert at a different rate than those from organic search? What do users who completed a purchase have in common that users who abandoned the cart do not?
In Google Analytics 4, segments can be applied in the Explorations section of the platform. You can define a user segment based on attributes like device category, country, acquisition source, or whether a user has completed a specific event such as a form submission or video view. You can define a session segment based on what happened during a particular visit. You can compare multiple segments side by side in the same report, which is one of the most powerful analytical techniques available.
Segments are closely related to dimensions, but they serve a different purpose. A dimension lets you break down a table by a categorical attribute. A segment lets you apply a persistent filter across an entire report or exploration, sometimes combining multiple conditions to define a highly specific group. For example, a segment defined as "mobile users in Gauteng who visited the pricing page but did not convert" cannot easily be replicated with a single dimension alone.
Beyond analytics, the term segment is also used in paid advertising. In Google Ads and Meta Ads Manager, audiences are often referred to as audience segments, which include custom audiences (people who visited your website), lookalike audiences (people who share attributes with your customers), and interest-based or demographic segments. Understanding how these audience segments perform is essential for efficient paid media management, and it connects directly to your customer acquisition cost targets.
Segments In Practice
A Durban-based insurance broker running Google Ads and a blog-driven SEO programme used segments in GA4 to investigate a conversion rate problem. Overall site conversion rate was 1.8 percent. By creating two segments, one for organic traffic and one for paid traffic, they discovered that organic visitors converted at 2.9 percent while paid visitors converted at 0.7 percent. The paid traffic segment revealed that most of those visitors landed on blog pages rather than the service pages, indicating a landing page mismatch in the campaign settings.
Further segmentation by device type within the paid traffic subset showed that mobile users who clicked paid ads had a conversion rate of just 0.3 percent. The combination of segment analysis and metric review produced a clear action: redirect paid ads to the correct landing pages and improve mobile UX on those pages. Within six weeks, paid traffic conversion rate improved to 1.4 percent, reducing the cost per acquisition by 38 percent without increasing the advertising budget. Segmentation is the analytical lever that turns vague performance concerns into specific, fixable problems.
How segments work in analytics
A segment, in analytics, is a subset of your data isolated by defined conditions, so you can analyse a specific group of users or sessions rather than the whole. Where the full data set averages everyone together, a segment lets you look at, say, only mobile users, only visitors from paid search, only people who converted, or only new versus returning visitors, revealing patterns that the overall average hides. Segments are applied to reports to focus the analysis, and can often be compared side by side, so you can see how different groups behave differently. This matters because aggregate data misleads: a middling overall conversion rate might hide a strong desktop rate and a weak mobile one, a distinction only segmentation reveals. Segments turn a single blended picture into a comparative one, which is essential for understanding which audiences, sources or behaviours actually drive results and where problems lie.
Using segments for insight
Segments are one of the most powerful tools for turning analytics data into insight, because most useful findings come from comparing groups rather than reading totals. By segmenting, you can answer specific questions: does mobile convert worse than desktop, do visitors from one channel engage more than another, do returning visitors behave differently from new ones, where do converters differ from non-converters in their path. Each comparison can point to a concrete action, fixing a weak mobile experience, investing more in a high-converting channel, addressing where a valuable segment drops out. The distinction between a segment and a filter is worth noting: a segment isolates a subset for analysis without permanently changing the underlying data, whereas a filter, in some tools, permanently alters what data is collected or shown. Because segments are non-destructive and comparative, they are the everyday instrument for exploring data, and using them well, always asking which groups differ and why, is what separates genuine analysis from staring at averages.
FAQ
What types of segments can I create in Google Analytics 4?
GA4 supports three segment types: user segments (based on user attributes or lifetime behaviour), session segments (based on what happened in a specific session), and event segments (based on specific events or event parameters). South African businesses commonly use user segments for high-value visitors, session segments for campaign traffic analysis, and event segments to isolate video engagement or scroll behaviour.
How does a segment differ from a filter?
A filter permanently limits or restricts the data in a view or report, while a segment temporarily applies a subset to a report for comparison without altering the underlying data. Segments allow you to see two or more subsets side by side, which is more flexible for analysis. Filters are better for permanently excluding irrelevant traffic such as internal staff visits.
What types of segments can you create in GA4?
GA4 supports user segments (people meeting conditions, such as purchasers), session segments (sessions meeting conditions, such as those from a channel), and event segments (based on specific events). These let you isolate and compare groups by who they are, how they visited, or what they did, to find patterns the overall data hides.