Behavioral segmentation
Also known as: Behavioural segmentation
Dividing a population by their behaviors, motivations, and barriers so interventions can be tailored to each group.
What it means
Behavioral segmentation is the practice of dividing a population into groups defined by what people do — and why they do it — rather than by demographics alone, so that an intervention can be matched to each group's motivations, barriers, and readiness to change. Instead of treating everyone the same or tailoring only to age and income, it sorts people by behavior, determinants, and attitudes, whether through theory-driven typologies decided in advance or empirical clusters drawn from survey data. A segment is only useful if it is measurable, reachable, and actionable. It matters because targeting the right behavioral lever to the right group is intended to separate an intervention that works on average from one that works for the people who most need it — though the evidence that segmentation itself lifts outcomes is thinner than its popularity suggests.
The bases you can segment on
Segmentation theory distinguishes four bases: sociodemographic (age, income), behavioural (what people currently do), psychographic (values, attitudes, identity), and epidemiological or risk-status. Behavioural segmentation privileges the last three over the first, because two people the same age can sit at opposite ends of motivation. Segments are built in one of two ways: a priori theory-driven typologies, where you decide the cutting variables in advance, such as stage of change or risk-perception crossed with self-efficacy; or empirical clustering, which lets patterns in the data define the groups. Whichever route, a segment is only useful if it is measurable, reachable and actionable. A psychologically real group you cannot deliver anything to is a description, not a target.
How segments are built
Empirical approaches run cluster analysis — k-means, two-step clustering, or latent class and latent profile analysis — over survey items covering behaviour, determinants and attitudes, then name and profile the resulting groups. There is no single correct number of clusters; analysts trade parsimony against fit and pick a solution that is interpretable and distinct. Good practice validates segments by showing they predict the target behaviour and respond differently to messages, not merely that they differ on the input variables. Slater and Flora's 1991 lifestyle segments, for example, differed on concurrent practices such as seatbelt and vitamin C use, and predicted different behaviours — eating less salt, losing weight, exercising — at a two-year follow-up. Without that predictive check, clusters can be statistical artefacts that carve the population into groups no intervention can actually move.
What the evidence shows
The case for segmentation is more assumed than demonstrated. An umbrella review of social marketing interventions found only a minority reported segmenting at all, and where they did the effect is hard to isolate — segmentation is a vehicle for building messages and offers, not a treatment you can switch on and off in a trial. The interventions that worked best tended to adapt the offer itself, not just reword a leaflet for each group, and to combine segmentation with several other design principles. A 2023 systematic review of risk communication reached a similar verdict: tailoring to heterogeneous audiences helps, but the evidence base is thin — fifteen studies — and rarely tests segmented against unsegmented delivery head to head.
Where it breaks down
Segments are probabilistic, not personal: membership describes a tendency, and treating everyone in a cluster as identical reintroduces the stereotyping segmentation was meant to escape. People also move between segments — someone contemplating change this month may act next month — so a static scheme dates quickly. More segments multiply cost without guaranteeing payoff, and a segment defined by an attitude invisible from any available channel is unreachable in practice. There is an equity edge too: the same tools can direct effort toward those least served, or quietly cream off the easiest conversions. The discipline is to split only where groups genuinely need different levers, and to stop when an extra segment stops changing what you would actually do.
Examples
A vaccination drive segments the hesitant: the access-constrained get mobile clinics, the misinformed get trusted-messenger facts, and the confident get a simple reminder — not one identical leaflet.
A bank splits savers by barrier rather than age: those with nothing spare get a round-up tool, the disorganised get an automatic transfer, and the already-saving are left alone.
A recycling drive treats flat-dwellers with no bin space, confused sorters, and committed sorters differently — an identical leaflet through every door reaches only the people already sorting.
A utility cutting peak demand sorts households by why they overuse: the unaware get a usage comparison, the well-meaning but forgetful get an automatic schedule, and price-sensitive renters get a tariff switch.
A tax authority splits late filers by cause: the confused get a simplified form, the disorganised get an earlier reminder, and the deliberately evasive get an audit warning rather than a gentle nudge.
First described in Social marketing practice.
Key references
- Bartolucci, A., Aquilino, M. C., Bril, L., Duncan, J., & van Steen, T. (2023). Effectiveness of audience segmentation in instructional risk communication: A systematic literature review. International Journal of Disaster Risk Reduction, 95, 103872. doi.org/10.1016/j.ijdrr.2023.103872
- Kitunen, A., Rundle-Thiele, S., Kadir, M., Badejo, A., Zdanowicz, G., & Price, M. (2019). Learning what our target audiences think and do: extending segmentation to all four bases. BMC Public Health, 19, 382. doi.org/10.1186/s12889-019-6696-2
- Kubacki, K., Rundle-Thiele, S., Pang, B., Carins, J., Parkinson, J., Fujihira, H., & Ronto, R. (2017). An umbrella review of the use of segmentation in social marketing interventions. In T. Dietrich, S. Rundle-Thiele, & K. Kubacki (Eds.), Segmentation in Social Marketing (pp. 9-23). Springer. doi.org/10.1007/978-981-10-1835-0_2
- Slater, M. D. (1996). Theory and method in health audience segmentation. Journal of Health Communication, 1(3), 267-283. doi.org/10.1080/108107396128059
- Slater, M. D., & Flora, J. A. (1991). Health lifestyles: audience segmentation analysis for public health interventions. Health Education Quarterly, 18(2), 221-233. doi.org/10.1177/109019819101800207