Age: its really all in the mind
Lisa Edgar and David Bunker argue that it's time to broaden and change thinking about age, using a new study with the BBC to press their point home.
We often talk about the market for 30 to 45-year-olds or ‘50+ consumers’ or even Generation X or the Silent Generation. When we do so we are, by default, anchoring our definition of age – assuming that a chronological definition of age is king. Marketers and researchers alike love chronological age; it is a linear, sequentially quantitative notion of time (Szmigin & Carrigan, 2001). Yet while this may be a convenient and easy frame of reference, it does not truly reflect a more complex picture – as the Big Window’s hot-off-the-press, large study with the BBC shows.
Even if, as a number of studies show, we age neurologically as well as physically (and the picture here is more complex than we might think), this does not necessarily reflect how we think about ourselves, who we identify with, how we interact and so on.
Perhaps Behavioural Economics should teach us a thing or two about how we see, treat and communicate to older age groups: Whether our frames of reference for older consumers are numerical (50+ years, 65+ years etc) or label-based (‘Empty Nesters’, ‘Winter’, ‘The Silent Generation’ and so on), they are typically positioned at the end of these frames.
In essence, marketers and researchers are Anchoring and Framing (Tversky & Kahnmann, 1974) such that older consumers are always at the end of something, in decline, seen in relation to younger, more significant, others, have had the best of their lives, ‘a good innings’ and so on. In turn, this can lead our strategies to be about what people can’t do rather than what they can.
We are not saying that these frames or strategies might not be right for some older consumers (just as they might be right for some younger ones), but we are saying that that focussing on chronological age alone is potentially misleading.
As a research community have we really asked ourselves why, when people are recruited to groups according to their chronological age, they don’t necessarily relate to one another and why we have found ourselves having to overlay complicated segments and dispositional criteria to make up for this? Have we really asked ourselves why, in quantitative analysis, we might not get the types of age differential that we might expect?
Perhaps the answer is that consumers do not see themselves as being the ages that we see them as. Perhaps they respond to brands, products, services and marketing communications as the age they see themselves as, not the age we see them as. If this is true, not only is this a misunderstood market – but marketing and research might be missing its target too.
Existing knowledge of perceived age
The need to go beyond a chronological definition of age was recognised during the 1980s. US-based academics Barak & Schiffman identified that, while chronological age stood out as the most frequently used variable in research, its socially-imposed characteristic rendered it much less useful than people/researchers/marketers assumed. And, according to Schiffman and Sherman (1991):
“age is revealing itself to be more a state of mind than a physical state (i.e. chronological age)”
Schiffman and Sherman illustrated this using Ford’s positioning of its new Mustang as an inexpensive sporty car for young people. Ford found that, rather than solely attracting young people, the car appealed to a cross-chronological age group, a group defined by a different definition of age: the ‘psychologically young’.
To date, most academic studies have used the four-item definition of ‘personal age’, first suggested by Kastenbaum, Derbin, Sabatini & Artt (1972): ‘feeling age’ (how old someone feels), ‘looking age’ (how old someone feels they look), ‘doing age’ (age-related behaviours) and ‘interest age’ (age-related behaviours). They have explored individuals’ perceived age as defined by an average of these four items. Their findings (following analysis of 324 females aged 55+ years) showed that each ‘perceived age’ measure and the composite reliably showed that chronological age only partially explains the age people identify with. Other studies have since confirmed this.
What was missing
It is clear that perceived age measures are important in understanding and influencing how consumers feel and what they might do as a result. The work in the 1980s and the studies that followed have been critical in developing our understanding of this. That said, our review of the literature has identified a number of opportunities to develop our understanding of non-chronological age perceptions for more effective use in marketing theory and practice:
- Studies have tended to focus on the older consumer – potentially
limiting the ability to explore the relationship between
chronological and perceived age and suppressing the opportunity
for effective segmentation by perceived age.
- Studies have nearly exclusively used the four-factor perceived age
construct first (Kastenbaum et al), potentially inhibiting the
discovery of other potentially important perceived age factors.
- Studies have tended to use the mid-point of a broad ten-year
interval for understanding age perception, averaging the midpoints
to find the composite measure. We felt this suppressed
potential differences between seeing oneself in, say, the late versus
- Related to the above, studies have been limited to relatively small
samples, potentially restricting both the ability to develop more
sophisticated constructs and sub-analysis of the various constructs.
- Finally, studies have rarely had the opportunity of applying the perceived age constructs to the type of data that the BBC holds on a nationally-representative UK sample and its members’ media-related behaviours (i.e. channels, genre, and programme preferences).
As a result, we set ourselves a number of challenges, which became the foundation for our study:
- Devise, using primary research, a solid and reliable measure of
perceived age, not just at the composite levels but identifying any
factors that underpinned it, too.
- Pin-point out the key dispositional (personality-based), situational
(contextual/circumstantial) and behavioural determinants of our
model: explaining why some people feel even younger than others.
- Demonstrate how the approach could explain/predict preferences and behaviours more than chronological age alone.
- Devise market/sector-specific models to help companies more accurately target and effectively communicate to their key audiences.
And so we set about developing the Age Frame.
What we found
It is perhaps unsurprising that we found that, in general, consumers feel younger than they actually are. What is interesting, though, is the extent of the gap between ’reality’ and ‘perception’ (shown in the bar chart below). Using the overall measure of age we found that until they hit 30 consumers typically ‘feel’ older than they actually are, but after 30 they start to feel younger and increasingly so. In fact, by the time they reach their early 70s, consumers actually ‘feel’ in their mid to late 50s.
The BBC team then applied the Big Window’s composite measure of perceived age to the wealth of media-usage data they had on the 3,000 media panel members that completed the first wave of the age perception survey. To really put the approach to the test, they analysed the different radio and TV-related behaviours of viewers/listeners who are particularly ‘young at heart’ (those who saw themselves as even younger than the average perceived age) and the less so, the more ‘mature at heart’ (those who saw themselves as older than the average perceived age).
The results were fascinating and confirmed that we were tapping into aspects of the audience that has real behavioural implications. Knowing someone’s perceived age helped explain their media choices in a much more sophisticated way than chronological age alone. The BBC is confident that the approach has many potential applications going forward. They think that it could help with:
- More intuitive scheduling – enabling them to place programmes
together in the schedule that have similar perceived-age appeals
(whatever the chronological age).
- More effective trailing of new programmes – putting trails in or
around shows that have a similar perceived age appeal to the
message they want to convey.
- Attracting a wider range of potential audiences: helping channels/stations appeal to those likely to be receptive to the content of that channel whatever their chronological age.
In addition to building and testing an overall factor or construct, we regressed each of the Age Frame© items against chronological age plus other determinants and analysed the similarities of the individual models, i.e. groups of items which had similar predictive models.
The models also suggest that we can predict each of the Age Frame items from chronological age with a good degree of accuracy (all with significant correlation coefficients) and the results suggest that the Age Frame construct appears to have three interesting sub-constructs:
- Internal/Cognitive age: perceived age relating to mental energy,
thinking age, interest age;
- Interactive/Social age: perceived age relating to emotions,
identity, social interaction;
- External/Aesthetic age: perceived ‘looking’ age. In other words, perceived age factors appear to position along an ‘Internal-External dimension’.
Lisa Edgar and the Big Window are now developing more effective marketing models for a wide range of sectors, products and brands, including the ability to develop age perception models that are sector or brand-specific. Models that, for example, could be more appropriate to the travel industry (e.g. by developing an Age Frame model which leans more towards the Interactive/social construct of perceived age) versus models that are maybe more appropriate to the cosmetics industry (e.g. by developing an Age Frame model which leans more towards the External/cosmetic construct of perceived age).
Founder, the Big Window Consulting limited
This article was first published in InDepth magazine, March 2012
Copyright © Association for Qualitative Research, 2012