22nd Dec 2022
Big data is not completely new. Over the course of centuries, people have been trying to use data analysis and various analytic techniques to support their decision-making process. The ancient Egyptians around 300 BC were already capturing existing data in the library of Alexandria. Moreover, the Roman Empire used to carefully analyse statistics of their military to determine the optimal distribution for their armies.
With online powerhouses such as Google, Meta and Amazon, the volume and speed at which data is generated and collected has grown exponentially. The total amount of data in the world is 94 zettabytes. To put that in perspective, 94 zettabytes is equivalent to 94 trillion gigabytes. Even with the most advanced technologies today, the amount of consumer data available is beyond the measure of human comprehension.
It is easy to feel overwhelmed – not knowing what data you are actually collecting and what it means, not to mention how your collection methods are unknowingly altering the results. Performing research purely based on data does not explain the drivers behind behaviour, sentiment and perception.
Uncovering the local cultural nuance and utilising cultural intelligence is essential to seeing past the numbers and truly understanding what cultural aspects drive the data, particularly in a multimarket and cross-cultural environments.
Cultural intelligence allows you to unlock human elements within large data sets by understanding context and cultural drivers. Tailoring your approach to global data collection and data analysis can help you uncover the real insight within each data set.
For example, an international up-and-coming brand is launching a social listening strategy to understand brand awareness in the Chinese market. They feel a lot of people know of their brand, but they don’t seem to get many mentions across social platforms and are unsure why that is.
With that in mind, it is important to know that in China, people tend to refer to western brands by their logo, their official Chinese name or an acronym of their name, instead of their western name itself, which can lead to the social listening results being skewed.
A clearly recognisable logo helps Chinese markets remember brands, especially those with difficult names (e.g. Abercrombie and Fitch is sometimes referred to as “the deer” or AF) and this should be part of the data analysis planning process.
It’s hard to find comparable demographic data in the complex cross-cultural context. Openness about certain aspects of one’s identity might differ across nations and there are legislative differences. E.g. in the US you can ask someone about their ethnicity (but it needs to be provided voluntarily).
Ethnicity is thus measured in ethnic background. In South Africa, ethnicity is very politically entwined. The employment equity bill empowers the Minister to inter alia identify national economic sectors and to set numerical targets for any identified national economic sector, for the purpose of ensuring the equitable representation for suitably qualified people from designated groups at all occupational levels in the workforce.
Contrastingly, in France the legal framework makes it impossible to ask for someone’s ethnicity or religion. This means ‘ethnicity’ is measured in terms of ‘nationality’ or ‘generation of immigrants’ criteria.
Demographic databases and surveys are often affected by (un)conscious bias. This bias does not just occur during the analysis stage when data analysts make assumptions based on their knowledge of a certain group/country or simply apply what they know is true for their own environment.
It is also often unconsciously applied when survey makers create the actual questions and/or answer options to help slice the data further down the line. Options can sometimes be limited to the viewpoints of survey makers and can therefore make participants feel like they don’t identify with the questions or answers.
For example, the UK census allows people to choose Indian, Pakistani, Bangladeshi, Chinese, or any other Asian background. But despite the diversity of the Arab identity, people can only choose Arab.
The definition of the ethnicity ‘Arab’ is problematic because Arabic-speaking countries span different continents and cultures across the Arab League and identification with the region might differ between generations or regions. Biases in censuses can therefore cause underrepresentation of certain groups.
What this shows us, is that without applying cultural intelligence or a cultural understanding to the data collection process, unconscious bias can sneak into the data parameters. Which in turn can provide inaccurate data and inadvertently marginalise a whole community.
Being more inclusive and understanding of an audience or an individual’s cultural influences prior to collecting their data is essential to determining what drives the decisions behind the data.
Rather than just determining cultural influences and segmenting by place of birth or parent nationality, utilising methodologies like anonymous focus groups and applying a nuanced approach to data collection will yield better results and a deeper understanding of an audience compared to the more traditional segmentation.
Think beyond demographic data.
When setting out your initial data parameters look past the traditional layer of demographics, such as gender and ethnicity. Culture transcends statistics and socially constructed categorisation. Think about your audience, their lived experiences and how that translates to the data you are trying to gather.
Fight fear culture and create dialogue.
When collecting sensitive data, you need to make sure that you create open, safe spaces for people to express their perceptions and provide a place where their voices are heard. Creating a dialogue with your audience counteracts fear culture and will return far better results than traditional methods of research.
Another proven method is allowing participants to use an alias so they can anonymously express their true feelings.
De-bias the research processes and engage with communities.
The way to de-bias your research and make sure it is inclusive is to involve underrepresented communities from conception. Build your questionnaires and methodologies by continuously engaging with these communities to understand what the right way to approach and phrase things is.
This process shouldn’t stop at the research stage, but should continue through product development, service development, and beyond.
Utilising cultural intelligence correctly at the beginning of the data collection process will provide clear cultural data without any bias. The process can be daunting, but the result will be more useful, rather than just more data.
If you’d like to learn more about cultural intelligence and how culture can work for you visit our insights page.
Sign up to our weekly insights and maximise your competitive advantage