Customer segmentation in practice

Grouping your actual customers by how they behave and what they are worth.

What it means

Behavioural segmentation groups existing customers by purchase frequency, value, product mix and tenure — commonly through RFM analysis of recency, frequency and monetary value.

Why it matters

It is segmentation done on data you already hold rather than on assumptions, and it usually reveals that a small group produces most of the profit.

What it looks like in practice

Rank customers by contribution and look at the top and bottom deciles. The characteristics of the top group are your targeting criteria.\n\nThe bottom group is a pricing or a service-model question, not necessarily one to lose.

What to watch out for

Segmenting by revenue rather than margin. The largest customer is frequently not the most profitable.

Where to get proper advice

Your own figures.