April 22nd, 2020
How much does a person’s age, education and income affect their decisions and behaviour? That used to be a very straight-forward question, and the answer a simple one: a lot.
However, in today’s data-rich environment, the answer is becoming increasingly nuanced. Recently, a person’s psychological profile, behavioural traits and belief systems have proven to be much greater factors that impact decision-making. Psychographics are now very much actionable and can even be weaponized, as proven by events such as Facebook and Cambridge Analytica.
So, it brings us to the question of whether we still need demographics and why.
What role do demographics play in this world that’s drowning in data?
In this increasingly noisy world, it often feels like we are ploughing through a tsunami of data with no way to navigate. We're shooting in the dark blindfolded. Demographics provide us an easy, straight-forward framework to take a first stab at slicing this data. After all, we’re all humans. Socialisation typically separates us into groups by sex, skin color, ethnicity, national identity, physical capability, cognitive abilities and class.
The short answer is, if you don’t collect demographic information, you’ll never know if there are any major differences in preferences and behaviours across various demographic groups. It’s also usually easier to draw hypotheses on specific demographics or find commonalities across demographics - because our mind and worldview are already shaped along demographic divides. For example, if you test a new product and find out that females are 80% likely to purchase it and males are only 40% likely to purchase it, you can take actions to market towards the female segment or make changes to increase the appeal to males.
It’s a short leap to see how this could lead us down strange roads - finding correlations where they don’t exist, pigeonholing data, etc. As with all techniques, as researchers we have to approach them with great caution and restraints - never conflating correlation with causation.
Whether we are doing marketing, improving a product or launching a new business line, we are constantly seeking to understand what motivates our customer. Or rather, to put it crudely, what would motivate the customer to consume what we have. Attitudinal or psychographic segmentation gives us clear directions on what is driving a customer’s decision-making process. For example, French food company Danone delivered a 40% lift in ad recall for its Volvic water brand by segmenting and targeting its marketing by passion point.
However, it's dangerous to form assumptions around your customer based only on mindsets and attitudes. Mainly, because mindsets and attitudes tend to shift and evolve over time. And similarly , if you look at just at demographics, you will also end up with an incomplete view of your customer. Just segmenting based on demographics will give you something closer to a stereotype than actionable personas and segments.
When looking for insights, a blended approach, where attitudinal segments are overlaid with demographics, will give you highly contextual segmentations. When used in conjunction with qualitative research (e.g. deep shadowing, ethnographic interviews, etc.) we can build nuanced and sophisticated scenarios and personas.
The last point is specific to running Gen-Pop (general population) surveys. It’s applicable to other forms of research, but we’ll focus on surveys here.
The facelessness and quantitative nature of surveys make it so that we sometimes conveniently assume recruitment is truly random and includes every walk of life. The reality is often far from it: surveys have a huge tendency towards sampling biases, and the smaller the N-number, the less representative the data can be. Without demographics as guardrails, we are unable to see the discrepancies between the distribution of our recruited sample vs the true general population.
For example, your survey of n = 100 living in City X said that 80% of participants find the city affordable. Your participants had an average cost of living of $10,000 / month.
However, according to census data, City X actually has an average cost of living of $6,500 / month. How much does the difference in your data affect how affordable participants think the city is?
This is, of course, a rather extreme example to illustrate the point, but a difference in stats show the importance of understanding demographics in your survey. For this example we should immediately want to know: what is the breakdown of income levels, age, professions, etc. of the sampled data as compared to actual census data?
Once you understand the differences in your data, you can understand how those differences influence the rest of your survey. You cannot find them without including demographic questions within the survey. Questions like:
Including demographic questions seems easy enough: just copy and paste standard demographic questions you find online. Done! Actually, you’ve still got some work to do.
Demographic questions need to be adapted for your survey so that the questions and responses are contextually, culturally and socially relevant. That’s not easy; you have to devote hours to market research, fact-checking and due-diligence, which can be very time-consuming and tedious.
This happens to us all the time. We often run surveys in countries and cities with vastly different cultures and languages. Every time we need to launch a survey for a different market, segment, or product, we end up spending a good amount of time researching to build accurate demographic questions. For example, income levels in the US compare differently to Singapore, so we need to adjust our ranges. And the way we ask a question in English might not be the right way to phrase it in Chinese or Vietnamese.
We spend a lot of time testing and learning with different questions in different contexts. When we finally get it right, we reuse the relevant demographic questions from one survey to another. This led us to put together a list of commonly used demographic questions contextualised for each of the countries we’ve worked in - and why should we be the only ones that have access to it?
Sign up now for early access!
Liked this? Drop your email here and we'll send you something interesting whenever we can.
Your missing survey design toolkit.
You’re not alone anymore.
© 2020 Bellini Slushie
This is a database of basic survey questions. Free for anyone to access and use. New questions added everyday!
Tzeying has been doing UX Strategy & Research across 7 countries in the Asia-Pacific for 10+ years.
Alexandria has been doing concept validation and development in the US and Asia for 3+ years.