Shattering stereotypes

x Sponsored content on Research Live and in Impact magazine is editorially independent.
Find out more about advertising and sponsorship.

Jessica Anderson, a nurse in the acute admissions unit of the Royal London Hospital, completed the 2019 London Marathon in her uniform in three hours, eight minutes and 22 seconds, raising £2,399.72 for charity in the process. Guinness World Records (GWR) denied her the record for the fastest woman running the distance in a nurse’s uniform, however. The reason? Its rules demanded a ‘white or blue dress, pinafore apron and white cap’. 

A week later, having taken a kicking on social media, GWR changed its mind, saying its guidelines were ‘outdated, incorrect and reflected a stereotype we do not in any way wish to perpetuate’. GWR’s oversight may be because few people in its daily operations are focused on whether its rules perpetuate stereotypes – though one would assume that has now changed. But even large, research-driven organisations fall flat on their faces.

One of the most infamous examples came on International Women’s Day in 2012, when Bic launched an ad with the excruciating copy ‘Look like a girl, act like a lady, think like a man, work like a boss’. The product outdid the advertising: it was a new line of pens… for women. The difference between the new pens and plain old ‘male’ Bic pens was that these ones were pink. Hilarity ensued as women posted pictures of themselves holding signs written using their new Bic pens. ‘It’s a miracle!’ read one. ‘I wrote this all by myself,’ said another.

While both brands took criticism for their stereotypical views, any post-mortem has to ask – how did no-one notice? The uncomfortable truth is that stereotypical thinking has been pervasive in business for decades, and continues to drive much of sales and marketing.

It’s bad for the people who are treated as stereotypes, it’s bad for society, and it’s bad for businesses in every way if it means that they are developing inadequate products and services – or none at all – for entire groups, failing to spot opportunities to innovate, or offending their customers.

“Using stereotypes reveals a failure of the insight-generation process,” says Sam Knowles, founder of Insight Agents, a corporate and brand storytelling consultancy. “Insight, to me, is an expression or manifestation of empathy. That profound understanding allows us to get inside the mindset of those we’re looking to influence.”

“Of course I’m going to say that stereotyping is wrong, but there are specific ways in which it’s wrong,” says Dr Marie-Claude Gervais, co-founder and research and strategy director at Versiti, a consultancy specialising in strategic research around diversity and inclusion. “First, people treated in this way feel misrepresented, misunderstood – sometimes even offended. Much like women don’t like being portrayed as exclusively sexual objects, people from diverse ethnic backgrounds also resent being typecast in a narrow set of roles. But it’s also a problem for customers, because they might not find brands or products that meet their needs.”

An example is the work that Versiti has done for Estée Lauder, examining the make-up choices available to black, Asian and Chinese women, and whether the cosmetics are too focused on a stereotypical, light-skinned customer. The research found large gaps in the range of colour foundations and helped make Double Wear one of the bestselling foundations in the world.

“It is also a big problem for society more generally if some groups of people feel disenfranchised,” Gervais says. “It leads to less community cohesion and, possibly, more extremist views. Within organisations, it leads to loss of talent.”

Versiti is currently helping a luxury car brand to understand the perceptions, needs and preferences of ethnic minority consumers, to challenge stereotypical internal views of who its customers are (middle-aged, white males), and investigate how that manifests itself in the way it communicates.

“I am doing escorted visits to high-end car dealerships to observe, and speaking to ethnic minority customers and dealers who are from ethnic minority backgrounds. They confirm what I’m observing and what customers are feeling. With the best intentions, a receptionist in a dealership took a look at young Asian car buyers and thought they were not on brand because they were wearing tracksuit bottoms and designer trainers. They are 15 years younger than they expect, so they don’t even bother to greet them. That customer will never come back and – because word of mouth is powerful – they will spread the news,” she says.

It’s a powerful reminder that brand image is mediated and communicated by customers and by those potential customers who feel excluded. This is a failure to connect with potential and future customers that has its origin in a narrow, often out-of-date view of what the brand represents. The staff who are the custodians of a luxury brand’s values may implicitly (or explicitly) jump to the conclusion that marketing to certain groups would ‘cheapen the brand’, for example.

Dr Gervais’s experience is that it takes a challenge from external experts – using research and insight – to effect change to this type of thinking. Undermining stereotyped thinking about customers, however, is ethical, leads to better recruitment and internal communication, and – used well – is a creative force in product and service design.

How does the research respond? Ben Skelton, CEO of Quadrangle, says: “One of the main functions of research is to embrace the human story in all its wonderful complexity, but also to help the users of research access it in simple and enabling ways. Focusing on this means that, theoretically, you step beyond seeing people as types or expecting them to follow patterns based on predefined behaviours or cultural assumptions.”

If the challenge to brands is to break out of stereotyped thinking, the challenge to researchers is to make research truly inclusive, Skelton says.
“We have to think about changing techniques. These include coding questions in ways that target recipients will respond to best, or ensuring sample recruitment is properly inclusive. We also need to make sure we’re not working off assumptions made because there are not enough diverse voices in the room. For example, when there are focus groups that clients want to attend, they are often arranged in London, so they end up with a more extreme urban bias. As an industry, we need to get better about addressing issues like this.”

Of course, it is possible for brands and researchers to have the best intentions, but still fail to understand the needs and experiences of groups, because they come from a dominant culture. “I think it’s less that they don’t engage fully and more that they can be unaware of the ways in which they’re not engaging fully, and where they’re getting it wrong,” Skelton says. “This is in no way to diminish the issue or onus of responsibility. It is totally on us as an industry to get better at this.”

Georgina Lee, co-founder of The Age of No Retirement (TAONR) wants brands to get much better at designing products and services for seniors. As she points out, with luck it’s a group we will all be part of at some point. The trouble is, we might not like it much when we’re in it. “There’s this popular image of people turning 50 waiting around for dementia. When, in reality, today’s 60-year-olds could still have a career ahead of them,” Lee says.

Trying to undermine this narrative, TAONR commissioned research into 2,000 people of all ages, to understand what unites and divides the generations. “We had an instinct that we have more in common across the ages than separates us. What we found was beyond our wildest hopes – across every group, 84% of us do not want to be defined by age,” Lee says. The results (see ‘How different are young and old?’ below) show remarkable commonality across all ages in a variety of questions that influence design, leading to the report’s 10 ‘intergenerational design principles’.

For many years, Lee ran a design consultancy, and argues that design thinking should be inclusive for young and old. She tells the story of the Ford Focus, a car designed originally for older drivers – easy to drive and to get in and out of. In development, testers wore a customised ‘third-age suit’ to mimic lack of mobility. The result was a car that was successful with all ages, because (not surprisingly) younger people also liked a car that was easy to drive, and to get in and out of.

Lee argues that our obsession with age-based demographic targeting, from Gen X to millennials, “defies logic”. TAONR research shows that, across all ages with no statistical variation, 86% of people wanted to see the design principles in action, but only 16% said they saw products that adhered to them.

“If you design well you’re not compartmentalising your customers into age-based demographics; you’re designing from an ageless perspective,” she says.

This is not just a challenge to designers, says Marisa Tuffnell, partner and customer insight manager at John Lewis & Partners (JLP). Advertising and marketing practitioners can also be drawn from a narrow demographic. “Sometimes, we’re in our little bubbles and it can be hard to look beyond them. It’s important not to stereotype specific groups and make assumptions about them, especially if you’re not that close to them… but the average age of a member of the IPA [Institute of Practitioners in Advertising] is 33, and less than 5% are over 50, so they are often in different categories from the people they are trying to target.”

Recently, a consultancy came in to JLP to discuss the impact of demographic trends in retail, and how to address growing customer segments. One of the slides showed the increase in the number of consumers aged over 65. “The icon used was a stooped person with a walking stick,” Tuffnell recalls. “I don’t think that’s helpful.”

Group identities

Stereotyping is complicated because our group identities may feel positive or negative – often depending on whether they have been pushed on us or actively sought. At some level, we are all compartmentalised. Sales, product design, innovation and marketing are not done for audiences of one, so – at some level – it becomes important to guess that a person is comfortable being identified with a peer group in some way. If the guess is good, it is tolerated or even liked by customers– the brand has tried to be empathetic, even if that is not always successful.

And while we may dislike being stereotyped, we all want to belong to defined groups at some level, even if that means altering our behaviour to conform to its values. Researchers, not least those looking to influence behavioural change, have discovered the power of group norms in influencing our preferences and choices, as Crawford Hollingsworth and Liz Barker have written about for Research Live.

Individuals may have many group identities, not all of which are salient, drive purchasing behaviour, or are easy to see in the available data. This means that any segmentation fits some customers better than others.

In the past, lack of data meant that demographic distinctions were a least-worst option for segmentation. No-one has ever seriously believed that all people between 18 and 25 share identical preferences while no-one outside that demographic shares them – but they were a better-than-nothing proxy for certain preferences, which increased profitability if groups were targeted in this way. As consumers, we have long tolerated inappropriate, demographically inspired marketing and advertising – perhaps because we had low expectations; perhaps because advertising and marketing have been a relatively small intrusion in our lives. As marketing messages become increasingly intrusive, however, so they also need to become more appropriate.

Combining customer-tracking surveys with operational and sales data offers the potential to identify what motivates us, rather than just what demographic category we fall into. O2 is one brand combining those data sources to understand its customers in a more nuanced way. Sandra Fazackerley, director of customer experience at Telefónica, says: “We’re pushing ourselves really hard to get away from generic segmentation – family would be my example. Quite often in the marketing world, we have talked about family as a segment, and maybe in our minds that concept was mum, dad and a couple of children. We’ve learned from our data that it is so much broader. Grandparents can have children; there are blended families; divorced and single-parent families; same-sex families. The old way of advertising to families can be very negative in some people’s eyes. It’s just too stereotypical.”

O2 now prefers to consider a customer’s affinities, not just attributes. Discovering the things that customer likes offers the chance to segment based on preferences, but this is much more data-intensive, and requires knowing the ‘why’ as well as the ‘what’. The reward is that customers are more likely to respond to messages positively, because they feel they have been heard. Fazackerley explains that affinities must be “the things that people have a strong bond with. So I might like football, but my affinity is with Manchester United. That step is hard to get to.”

Not all of this can be intuited by choices, but some can; for O2, some affinity may be as simple as discovering that a customer consistently prefers Apple or Android, which immediately narrows down the offers they are likely to appreciate.

The availability of bigger and better data may mean an overdue end to demographic segmentation. But we should also worry about the extent to which stereotypical thinking can be passively enabled or encouraged by digital services.

In the past, landlords placed signs in their windows: ‘No blacks or Irish’. In 2016, ProPublica, an investigative journalism site, posed as a landlord to buy advertising on Facebook in the US. It found that it could exclude African-American, Latin and Asian Facebook users from viewing its apartments-to-rent ads. This wasn’t a bug; it was a feature of the Facebook platform – or, more precisely, its ‘ethnic affinity’ marketing product.
As Facebook later pointed out (after it had hurriedly disabled ethnic affinity for housing ads), ‘ethnic affinity’ has positive applications for marketers who want to actively target particular groups, rather than just exclude them. It’s just that this particular unintended consequence was not only racist, it violated the Federal Fair Housing Act of 1968.

Machine learning creates more difficult-to-fix stereotypes in decision-making. The problem is that expert systems need to be trained, and much of the training data confirms stereotypes, especially when using images. Training using datasets – such as stock photos that embrace stereotypes (female scientists wearing glasses, business people that are male and wearing suits) – risks embedding those stereotypes in the next generation of products and services. Tom Evans, director of data science at Kantar, gives the example of artifical intelligence (AI) that was trained to recognise and categorise people in images in the home. “In the training data, the only pictures of a kitchen had a woman in them, so the algorithm classified every person in a kitchen as a woman.”

“I’m looking to build intelligent algorithms to analyse film scripts,” says independent consultant Shaf Choudry. “One area of research is taking movie subtitles as a corpus of text to train how to communicate – but if these scripts are problematic in the first place we’re training our AI systems with biased data.”

Choudry knows more about this than most, because he is one half of the team behind the Riz Test, a simple way to identify Islamophobia in films and on TV (see ‘Who fails the Riz Test?’ below). It was created by Choudry and Dr Sadia Habib, a specialist in identity, because of their – and their friends’ – dissatisfaction with relentlessly negative media stereotyping. “It was a conversation many people were already having – we’re just providing a framework to capture and codify it. The conversation exists, but the data does not,” Choudry says.

The test was named after Riz Ahmed, an actor and activist who spoke passionately to parliament about the problems of media stereotyping in March 2017 (he has tweeted his approval of the Riz Test). The impact of stereotypes was that many young Muslims would “switch off and retreat to fringe narratives”, he argued at the time, unless they could see balanced portrayals of their community
on screen.

Sense of community

In 2018, a wildly successful crowdfunding campaign raised £102,000 for the London LGBTQ+ Community Centre, a community-driven initiative to provide a shared space in Hackney. But the centre realised it had slightly different ideas of what that space would be. “It’s a very important time for this initiative because so many community spaces are being closed down, and this has contributed to increasing fragmentation among what is already a huge cross-section of people,” says Chris Illsley, a strategist at Flamingo. It wanted to answer the question: “How could it built something that spoke to the needs of the LGBTQ+ community?”

First, it needed to attempt to define that community. Flamingo used a two-stage process: in the first stage, qualitative research among the many groups within the community found a longlist of values that the centre should represent. In the second stage, qualitative and semiotic research helped examine the participants’ unconscious associations of the attributes of a community centre (‘boringness’ being one), and to come up with the list of services the centre could offer.

“We knew we needed to speak to the people on the fringes who needed this the most,” Illsley says. “People who traditionally don’t have this space – such as older people, people of colour, disabled people. They bring a different lens and perspective to what it means to be part of this community.”

Michelle Fan, Flamingo senior strategist for semiotics, helped this process. “When we have a lot of subconscious forces at play, I may have some stereotypical assumptions that I need to question. Semiotics helps us whittle them down into specific opportunities.”

Illsley adds: “As soon as you create something, it’s no longer owned by you. You create a framework the community will shape, so it is important not to hold on too tightly to your ideas of what it should be.”

This, they explain, inevitably introduces a tension, because when we think beyond stereotypes, there is no simple definition of an LGBTQ+ community, any more than there are homogenous communities of seniors, women or Muslims. “We had no idea how fragmented the culture is, and we should not homogenise it. But that means it becomes difficult to cater for the entire community in a meaningful way. There is way too much richness and culture. One of the insights we took back to them is inclusivity as an amazing value, but it’s not a strategy. We do not live in a society where we can do off-the-shelf thinking any more.”

“Using multiple research methodologies, this can be a powerful process,” Fan adds. “You can be concerned with the deeper questions of ‘why’, and really see where the culture is going. But even if you don’t have concrete answers to everything, you’re asking the right questions. That will push you in the right direction.”

“When people feel empowered to express their whole selves – and feel that it’s the expression, not suppression, of their point of difference that is being called upon – then we’re getting somewhere,” says Ben Skelton. “When that happens on projects, we can see a tangible difference in the quality of insight.”

Who fails the Riz Test?

The Riz test asks five questions. In a film or TV programme with at least one identifiably Muslim character, is the character:

1. Talking about the victim, or the perpetrator of terrorism?
2. Presented as irrationally angry?
3. Presented as superstitious, culturally backwards or anti-modern?
4. Presented as a threat to a western way of life?
5. If male, is he presented as misogynistic, or if female, is she presented as oppressed by her male counterparts?

If the answer to any of these questions is “yes”, the film fails the test. It is no surprise to find that a film such as Taken ( 2008 ), in which Liam Neeson battles evil Muslims who kidnap European girls and sell them as slaves, is a 5/5 fail. More troubling is that fantasy and superhero cinema often uses Islamophobic tropes as lazy plot devices: Iron Man ( 2008 ) is a 5/5 fail, Black Panther ( 2018 ) fails in two categories, and Batman v Superman ( 2016 ) fails in three by inventing an imaginary country beset by Muslim terrorists.

But the worst recent offender? Bodyguard, the most popular TV show of the past decade. It is a 5/5 fail within the first 12 minutes of the first episode. “Jed Mercurio is an incredibly talented writer, but he seems to have let himself down here,” Shaf Choudry, who created the test, says diplomatically. He now uses it as a case study in his Riz Test workshops.

How different are young and old?

Surveys of young and old show a remarkable commonality in response to questions on topics – use of the internet, feeling overwhelmed by technology, attitude to politicians – which are commonly considered to have a strong age-based skew.

Source: Age Does Not Define Us, download from www.ageofnoretirement.org
Base: Online survey of 2,000 adults, aged 18 to 99

Coca-Cola through the ages

Two years ago, Coca-Cola called a company meeting to discuss long-term trends that would affect it. One such issue was how it engaged with seniors. The racially and sexually diverse group that stood on a hilltop in its famous 1971 TV spot, singing ‘I’d like to buy the world a Coke’, would all be past retirement age now.

“We have been over-focused on recruiting a new generation; we never focused on seniors in our plans, in our merchandising. I think it’s a miss,” says Véronique Chéné, new beverages and innovation senior manager for Western Europe at Coca-Cola.

The company realised it understood little about what motivates seniors ( 65+) when selecting beverages – how they made choices, what design formats they preferred, and whether they passed on their brand preference for Coca-Cola to younger generations. So Chéné asked Engage Research to recruit online communities of seniors to discuss their opinions on the topic.

“Seniors are a huge demographic – one that is growing – and it has money. You can’t say that about a lot of target audiences at the moment,” says Deb Sleep, director of Engage Research. “But there is a disconnect between the stereotype of age we have in the back in our minds and what we see in front of us.”

It was not hard to find online respondents – another stereotype of seniors. “The online community we ran had some of the best engagement levels we have ever seen,” Sleep said. “If you can’t find those people, you haven’t looked hard enough.”

In follow-up, internal Coca-Cola workshops, Engage encouraged participants to bring photos of seniors they knew, which “totally blew stereotypes out of the water. One person danced the tango and had just come back from three weeks in Cambodia; another baked bread for the community. But, in the back of many people’s minds, is a stereotype of a 70-year-old with a walking stick, glasses and grey hair,” Sleep says.

Coca-Cola’s work is driven by the insight function rather than strategy or marketing, but similar projects are influencing its innovation, advertising and marketing. Chéné adds. “We want to share the information broadly. Across the globe, we are making sure it will be in marketers’ minds.”

Using customer closeness to break through stereotypes

By Alison Camps, deputy chairman and partner at Quadrangle

Researchers need to create an environment in which individuals feel empowered to share their world view, invite insight into their culture and reveal why they feel the way they do. This is the only way to engage communities that are often sidelined or misunderstood because of their religious beliefs, race, sexual orientation, age, family type and gender identity.

Having deep links to an extensive professional network of social commentators, cultural lobbyists, semiotics experts, journalists and, of course, highly skilled socially aware researchers to tap into allows us to lay the right foundations for the work we do. This is how we uncover the ‘unknown unknowns’, particularly around structural prejudge, which often plays into the formation of stereotypes.

Some of our most successful results come through customer closeness workshops. Running these goes to the heart of Quadrangle’s most important customer consultancy principle: know your customers and understand them as people. Workshops help a brand reconnect with what their customers think. Undeniably, they can bring some ‘and-breathe’ points and can be pretty humbling. Ultimately though, they have a series of light-bulb moments that push clients to a new point of discovery.

We equip our facilitators with a nuanced understanding of how certain stereotypes and unconscious biases have developed and the effect they create, both on the community depicted and on the dominant culture. This enables the right set-up and creates a trusted foundation.

The sessions are typically run across a formula that starts with ‘listening’, evolves to ‘asking’, builds in ‘observing’ and ends with ‘connecting’:

• Listening – stakeholders watch focus groups to understand customers’ functional and emotional needs and ‘walk a mile in their shoes’.
• Asking – one-on-ones and speed-dating exercises to allow stakeholders to gain greater depth in areas key to them such as what ‘being inclusive’ really means.
• Observing – understanding what matters to customers and building empathetic insight into the issues faced by participants through customer presentations.
• Connecting – presenting proposition ideas to customers and getting the happy or unhappy emoji treatment live in the room… ouch!
  

Ironically, brands often treat diversity and inclusion as a stand-alone issue. This is counterintuitive. Not only does proper representation go to the heart of every aspect of a brand, it also plays into trust, personalisation and voice of customer – all fundamentals that brands need to get right to win in the post-digital customer landscape. Integrating greater cultural awareness into all projects through customer closeness workshops pays dividends.

This report was first published in Issue 26 of Impact.