Trendwatching.com is a global, independent organization that helps forward-thinking business professionals understand “the new consumer” and subsequently uncover compelling, profitable innovation opportunities.
In November 2014, they published a trend-briefing that claims consumers have broken traditional demographic barriers: people are no longer ‘behaving’ along traditional age-group conventions. For instance, if you look at 60 year-olds’ 1,000 favorite artists, there is a 40% overlap with the list for 13 year-olds. Another counter-intuitive example provided was that Twitter’s fastest growing demographic between 2012 and 2013 was the 55-64 year age bracket.
They make 3 key claims that I briefly reframe below. Their claims have significant implications for data science practitioners and the business people who:
In this post, I offer:
Trendwatching.com uses a term called “Post-Demographic Consumerism” to explain the phenomena of consumers no longer behaving along traditional age-group conventions.
The term was originally coined by Richard Rogers, PhD, a web epistemologist at University of Amsterdam and director of The Digital Methods Initiative organization. The term represents a body of research and methods that leverage web-created data in contrast to traditional demographic approaches. It’s a play on words similar to how ‘post-modernism’ represents a departure from modernism.
This claim supports a concluding statement in their presentation that ‘demographics is dead’. After walking through these claims, I provide some reasons for why this is not only false, but possibly leaves us with blind spots in our organizations to longer-term consequences related to sustainability.
Whereas traditional demographic dimensions are categories such as age, education, ethnicity, etc., Trendwatching.com distills Post-Demographic Consumerism down to 4 dimensions:
These observations are possibly VERY valuable to our organizations as it pertains to profiling groups of consumers. They open the door for us to reintroduce practices such as anthropology or ethnography to help us come up with new language to capture how we observe social interactions in the physical and digital world. This in turn would affect how we collect, organize, curate and transform the data into meaningful ontologies or ‘categories’ of observation.
And it would be a mistake for us to think that these dimensions represent ‘reality’. As biologically closed systems, reality exists but we don’t necessarily know what it is. So we make up interpretations that exist as stories inside of our own heads. Those stories may or may not reflect reality. And in more complex situations, we may not know it right away.
For instance, the recent mortgage crisis represents a situation where large numbers of banks and people were operating with various interpretations of what is real. People were over-borrowing, often without realizing it or factoring in for bubbles. Banks were operating with various interpretations of what actions were permissible. And creditors were operating with various interpretations of risk - models created for them with underlying assumptions that were no longer valid. The reality of the situation didn’t play out immediately. But the eventual consequences were devastating.
The claim here is best said in their own words:
“Successful products, services and brands will transcend their initial demographics almost instantaneously. As a result, executives who continue to attempt to navigate using demographic maps, with borders defined by age, gender, location, income will be ill-prepared for the speed, scale and direction of change.” - Trendsetter.com November 14, 2014 Trend Briefing - Post-Demographic Consumerism
While I accept that there is power in accepting post-demographic consumerism, I offer that companies that have a concern for ethics, sustainability and the longer-term continue to incorporate relevant traditional demographic approaches. And here’s why.
Based on work pioneered by Terry Winograd and Fernando Flores (Stanford), I offer the following for consideration:
#1 - Biology: our biology affects us. As we get older, we have less resilience and more breakdowns. And yes, those biological concerns have a plasticity that we need to be wary of as we practice data science.
For instance, there are some fit 60 years olds that can out-run obese 40 year olds. And yet, the 60 year old’s age does mean certain real things in the world - like how much disposable income is consumed in healthcare insurance precisely because of their age.
This is why framing the problem that we’re solving and reflecting carefully on how we ‘bucketize’ or ‘cardinalize’ age-related information is critical. It may be that “age 41-59” is a meaningful categorization. For instance, for an employer assessing costs related to their aging workforce, it may be critical since insurers also use similar tiering processes. It may have certain associations to unemployment when people get laid off. And that in turn may be related to other factors such as cash flow that affect people’s real ability to acquire certain brands.
#2 - Language: language affects what we can observe. Doctors have a specific language for understanding the human body that enables them to do their work more effectively than an engineer operating on a human body (for instance). And likewise, engineers or lawyers have specific language for understanding that enable them to operate in their domains.
So I offer that learning, and its more formal measure called 'education', produces the language with which we observe the world. And it can matter. While traditional demographic approaches may not completely capture this, education (formal and informal) and the level of education may be an important surrogate to understanding behaviors in certain domains of action.
#2 - Culture: we exist in a context of surrounding narratives. This informs and shapes our views of what is real and what success looks like (philosophy).
I am half-Jamaican, half-Egyptian born in Canada to immigrant parents. When it comes to recycling or stopping at crosswalks, my Canadian heritage takes over. I find myself instinctively picking up trash lying around or waiting for the light. And it also says something about the kinds of products I will gravitate towards.
Are demographics dead? I can’t yet see it as dead - can you? Demography, strictly speaking, is the study of structure, interactions and shifting trends among populations. Data is opening up a door for us to study demographics in a different way. And I accept that it will begin with accepting that the boundaries of colonization on our demographic maps appear to have been drawn along some artificial lines.
But I think what Trendwatching.com is driving at is real: our approaches to demographics need more than a ‘make-over’.
The approach to demographics has wide-reaching implications.
As Trendsetter.com points out, it is changing:
But, as we saw with the mortgage crisis, this is not about having a beautiful model. While people are free to make choices as to whether they borrow the money for that nice new car because they can’t really afford it, we are also free to make choices about who we market to. More importantly (and selfishly), we pay in some way for predatory practices. We pay in our taxes, we pay through crime rates, we pay in our bodies when there are problems with food or water sources, and we’ll pay in terms of our dignity when our adult children or grandchildren look back on history and question our judgement.
There are some opportunities for us a data scientists to lead the way of exploring the consequences of the models we build.
This reshaping of demographics requires a different set of distinctions with which to observe. These distinctions are unlikely to be readily accessible by data transformation teams. We need to widen the disciplines of our data science teams: bring in the anthropologists, physically visit the situations we are analyzing, pull in the artists who often see things and capture the moment from a totally different perspective, get help from people who have different modeling disciplines to help frame and understand the problem. Then think about the data analysis problem… if it still exists.
As importantly, we also need to revisit our corporate values. Assessing the longer term implications to our communities requires us to raise the role of ethicists and sustainability leaders.
As data scientists, we need to lead. We have an opportunity to take a more proactive leadership role in understanding, solving or dissolving the problems.
Below I offer a simplified checklist to walk through as an organization before diving into the data.
Five (5) simple questions that drive towards giving us an ‘experience’ of the problem before trying to solve it. They open the door to conversations about our markets, our ethics and our values. And they help us to produce a shared interpretation, bringing the experts that can help us understand more fully the situation at stake.
Trendwatching.com, if you are listening, thank you for opening the dialog. Misbehaving demographics are an opportunity for us to clean up our behaviors.
Copyright © 2014 Susan Bennett · All Rights Reserved ·