Segmentation research needs to serve our strategy, not be the strategy. And we should only do it if it’s the best way to understand our audience, if we keep our communications rooted in our values, and if we can actually target the segments we find.
On the face of it, audience segmentation is a no brainer for campaigners.
Sorting people into groups makes sense because there are real and important differences between us—like our beliefs about what causes poverty, or whether the UK should have left the EU. We know that some of these differences are significant in how people respond to communications. And campaigners need to communicate in ways that are relatable and compelling for different audiences.
If the golden rule of communication is to understand who you are communicating with, then segmentation surely helps you do that?
It definitely can. But without the right tools and strategy to hand, it can be used in a crude and damaging way.
What is ‘audience segmentation’?Audience segmentation means putting people into groups based on demographics (like age, gender, region) and/ or psychographics (like values and attitudes), to understand how each group thinks and feels. For instance, More In Common’s project Britain’s Choice finds seven distinct groups in the British public—such as ‘Progressive Activists’ and ‘Disengaged Traditionalists’. Targeted audience segmentation means tailoring communications in order to specifically appeal to a particular segment.
As civil society organisations increasingly embrace targeted audience segmentation, I interviewed ten people* who have pioneered, practised or critiqued the approach, to understand what makes segmentation helpful rather than harmful.
The most important thing I learned is:
Research into how people think now doesn’t tell you where they can go, or how to get there.
You can have a sophisticated portrait of the world view of a ‘Disengaged Traditionalist’, but this doesn’t tell you how the views and behaviours of a person in that segment can shift, or which messages they will respond to. That’s because segmentation is descriptive not prescriptive. The biggest problem with segmentation projects is that people think that the research is the communication strategy: that it tells you what to do.
Talking to these ten experts I gained deeper insight into when segmentation is helpful, and when it isn’t.
When is segmentation helpful?
Segmenting people can give us useful insight into how people think and feel.
A good segmentation project, characterising people’s views on a range of topics, will capture complexity and nuance that is missed in other approaches—like focusing on simple demographics like age and gender, or a single attitude like political affiliation. But it’s not just about the quality of the insight we get from the research, it’s also about how we plan to apply segmentation to communications.
Before beginning a project, these questions can help guide our decision:
|☐||Does segmentation serve your strategy?||Clarify how this research will support your theory of change. You might be trying to bring everyone along with you in one universal message, or to move your passive supporters to be more active, or to fire up one audience in order to carry the message to another.|
|☐||Are you focused on where people can go, rather than where they are now?||As part of your strategy, check your own assumptions about the possibility of changing the mindset. Campaigners are trying to progress social, cultural and political change, rather than reinforce the status quo. That means seeing people as dynamic, with potential to change their thinking and behaviour.|
|☐||Is it the best way to understand your audience?||Ask whether this is the best use of time and money to get what you need. If your work is to organise and mobilise in a community, you might be better off getting to know the lay of the land through being in regular direct contact with people, through relationships, and partnerships.|
|☐||Can you actually target these groups?||Assess your capacity to reach audiences, and reach them cleanly. Larger, well resourced organisations might be able to do sophisticated social media targeting, but many just don’t have the capacity to go much beyond basic metrics like gender and age.|
|☐||Will you stay rooted in your values and follow good comms practice?||Identify your core values, just iterate these so they resonate for different groups of people. There are many great resources on tried and tested communications principles, like the Opportunity Agenda’s toolkit on ‘Vision, Values and Voice’; PIRC’s work on ‘The Narratives We Need’; Anat Shenker-Osorio’s Messaging this Moment and We Make the Future, and Ralph Underhill’s communications traps. The combined wisdom of these tells us:- Lead with shared values- Don’t make people feel ashamed and guilty- Point people towards clear solutions to problems- Tell individual stories but in a systemic context- Don’t parrot the language of the opposition|
Segmentation has been put to great use in many projects:
The Race-Class Narrative Project
What they did: The Race-Class narrative project is a collaboration between Ian Haney López (author of Dog Whistle Politics), Anat Shenker-Osorio (ASO Communications), and the public policy organisation Demos. They seek to engage people by talking about race and class in ways that strengthen social solidarity, reduce division and scapegoating. Through many rounds of research, messaging guides and local grassroots organising, they helped progressive candidates win the 2018 Minnesotan elections on a race-class ticket.
Why it worked: The strength of this work lay in having a clear strategy to begin with: To mobilise the base in order to persuade the persuadables. The segmentation research was in service of this strategy. It also worked because it combined research-led communications with effective, long term local organising.
Britain Talks Climate
What they did: Climate Outreach, using the segments from More In Common’s project Britain’s Choice, identified where these seven groups stood on climate and how to talk to them. One of the main impacts of their work has been to show that climate concern is not on the fringe anymore—it’s a universal issue, but with some variations in how people think and feel about it. Their toolkit covers both in-depth audience insights (including common ground, as well as differences between groups), and also framing recommendations.
Why it worked: One key strength of this project is that they combined rigorous new research with Climate Outreach’s wealth of existing expertise on good climate communication. Another strength was that they designed the project with implementation in mind. They worked with many organisations from the start, and continue to support campaigners on how to apply the segmentation through Climate Outreach’s Engagement Lab.
The Yes Equality Campaign
What they did: In the run up to the 2015 referendum on marriage equality, the Yes Equality campaign in Ireland used segmentation to great effect to identify different messengers for their audiences. For instance, they found that men over 45 were most likely to go from a soft yes to a no vote—that their main block was not outright homophobia, but doubt about two men raising children—and that they responded best to footballers, rugby players and certain radio presenters (‘journeymen’), who could share their stories about voting yes,
Why it worked: In this campaign, they didn’t deviate from their playbook of sharing personal stories and talking about the values of love, commitment, kindness and inclusion. They fitted segmentation firmly into their values-based approach.
We need a strategy to help us apply segmentation to communications.
If we start a segmentation project without a clear strategy, the risk is we fall back on default assumptions like: ‘these groups have fundamental, unbridgeable divides and we need to meet each of them where they are.’
That’s because segmentation insight tends to lead people into a few traps.
Focusing on differences between groups, rather than similarity across them.
If you are given segments, it is much more tempting to caricature, stereotype and pull those people apart in your mind than to focus on what unites them. More in Common, who produced the work behind Britain Talks Climate above, find seven distinct segments in the British public, but show that there is much more common ground between these groups than is often assumed. Their approach to segmentation is that we must understand difference in order to foster unity. That is an admirable approach, but it takes discipline to follow because people love putting other people in boxes.
It’s also easy to use short-hand and labels to describe people, but this in itself can be divisive. We should guard against a label like ‘Red Wall’ or ‘Anxious Middle’ becoming used as codes for race and class.
Assuming people have a neat set of views and values.
IMIX, in their segmentation work on migration, caution that people can rarely be neatly categorised, and in any case don’t like to be. Their rules of thumb are that good segmentation is non-reductive and non-judgemental.
Most people are complex, contradictory or vague in their opinions, multiple in their values, and responsive to different frames. The Common Cause Foundation has shown that even people with strong ‘extrinsic’ values (of power, money, social status and prestige) can still be effectively engaged on their ‘intrinsic’ values (of acceptance and open mindedness). So instead of assuming we should appeal to a group’s existing value priorities, we can and should appeal to them on the basis of their intrinsic values.
Meeting people ‘where they are’.
If we’re trying to make the world a less racist, more just, greener place, we should be focused on the potential for people to change. Anat Shenker-Osorio, founder of ASO Communications, has a clear line on this: “where people are now sucks”. If the role of audience research is to ‘take the temperature’, then the role of progressive messaging is to change the temperature. So after finding out where people are, she will develop and test messages that actively fire up the base, persuade the persuadables and alienate the opposition.
George Lakey’s Spectrum of Allies approach is another way to think about audiences in terms of direction of travel—for instance, from being a ‘passive ally’ to being an ‘active ally’. If we pair this with segmentation—say, to describe the characteristics of a passive ally—we are less likely to think that people belong in boxes.
Thinking like a marketeer.
Your strategy will be different if the goal is to sell a product, or the goal is to shift culture.
Segmentation had its genesis in marketing, rather than campaigning and activism. If you’re selling products, you want the quickest, easiest path to a specific behaviour. You can segment the over 50s luxury cruise market, find out that one segment wants hedonistic abandonment, the other looks for reassurance and safety, and pitch your holidays accordingly.
That may work when you’re selling a cruise, or an electric car. But if you’re campaigning, you have to think of the wider implications of your communications on mindset and culture.
An environmental campaigner might be trying to motivate people to install heat pumps, but not at any cost. If the campaign sells heat pumps but does nothing to foster concern and action on climate change, then it’s unhelpful. If the campaign moves people on climate change but dog whistles classism and xenophobia, then it’s worse than unhelpful.
If we think like marketeers and work in silos then our communication strategies will be too narrow.
Not testing our comms.
When we have commissioned a big research project to give us insight into audience segments, it’s tempting to think that this will tell us how to communicate. But because of the traps above, we can still make poor decisions about communications. We should test our comms with our audiences, and we might be surprised by what works. Testing doesn’t need to be expensive (although, with a complex segmentation it probably will be). The PIRC and ILGA-Europe resource How to Test Your Comms outlines the core principles and process for testing messages, on any budget.
We need to support people to implement segmentation
Nearly everyone I interviewed talked about how researchers need to help organisations use segmentation work and not just dump the data on them. There are tools to help with this. As above, the Britain Talks Climate team have done an excellent job of helping organisations apply the segments to their work, and sharing resources, like the training and tools to find the segments in their own lists.
But often there is a mismatch between the desire for audience insight, and the ability to reach those audiences. As one of my interviewees said, “If you can’t actually target the segments on the back end then why are you doing it on the front end?” It’s usually only the larger organisations that can do the clean and sophisticated social media targeting. And there is an unfortunate trade off, because the better the segmentation work (the more detailed, more multifaceted etc), the harder those segments will be to reach.
So if an organisation or funder is going to commission a segmentation project, it needs to consider whether the results can realistically be used to target those segments.
So, target with care
As a tool for understanding, segmentation can be invaluable. It can help us make more sense out of messy human complexity, and find common ground as well as divides. It can help us reflect on who we are communicating with and check our assumptions. It can even help us understand ourselves, our family divisions, and dynamics at work.
But if you want to target audiences, the bottom line is that the segmentation insight won’t tell you how to do this. You have to decide who you are talking to and why, based on your strategy and your theory of change. And you have to decide what you say to people based not only on how they think, but also on your own values, beliefs and understanding of good communication.
*Huge thanks to: Adam Corner, Alice Sachradja, Dora Meade, Emma Harrison, Gráinne Healy, Leo Barasi, Míriam Juan-Torres, Nat Kendall-Taylor, Ruth Taylor and Steve Ballinger. None of the opinions in this piece should be attributed directly to any of these interviewees, or the organisations they represent. Responsibility for all opinions and mistakes lies with PIRC.
Leave a reflection or ask a question below…