Custom segments

Why averages are not enough

Segments show where investment pays off, because they reveal how visitor groups differ. Averages hide these differences. An average conversion rate can mean that one group converts very well and another hardly at all.

Types of segments

• By source: search engine, campaign, newsletter, referral

• By behaviour: topics visited, number of visits, features used

• By attributes: device, region, new or existing customer

• By likelihood: users who are likely to buy or to churn

Which segments make sense depends on your goals. We start with the groups that make the biggest difference to revenue or enquiries.

Set-up in GA4

In Google Analytics 4 we create segments in explorations and define audiences that can also be used for ad campaigns. GA4 offers predictive audiences, for example for purchase probability, provided enough data is available.

Comparisons in the standard reports make segments visible to teams without analytics experience too.

For deeper analysis, we export the raw data to BigQuery and combine it with data from your CRM or shop. This allows segments by revenue, contract type or customer value as well.

From segment to action

A segment is only useful if a decision follows from it. We therefore define a question and a possible action for every segment: adapted content, dedicated campaigns, changed budgets. We then check the effect in the same segment.

A few clearly defined segments are more valuable than many that nobody looks at regularly.

Data protection

Segments are based on analytics data that is collected only with consent. We make sure segments do not become so small that individuals can be identified.

Audiences passed on to advertising platforms require appropriate consent. We check this before any audience is shared.