This article is for the Professionals
Sign up to Inside Retail Professional now for only $5+GST for your first three months.
That's an 85% discount plus you’ll get FREE access to all Masterclasses during Retail Week. 5 retail industry leaders like you’ve never seen them before.Already a professional? Log in
An unfortunate casualty of this focus on big data is other sources of information and insight, those we might think of as more traditional research methods, primary research, or ‘small data’. As retailers race to collect as much data as they can from sources like social media, loyalty databases, or emerging technologies, it can be easy to forget about one of the richest sources: talking directly to customers. Multiple articles have been written about the supposed death of research methods such as surveys, focus groups – even primary research altogether. Yet, like most things that have been pronounced dead (such as physical stores), small data still very much has its place. That place has simply evolved. Let me illustrate with some examples.
The power of customer conversations
Let’s start with the value of simply talking to customers – or traditional qualitative research. Some researchers live by it, highlighting that there is nothing richer than hearing something in a consumers’ own words, and diving deep into their experiences and feelings. Critics point to the subjectivity inherent in qualitative methods, and question whether usually small sample sizes can be generalised across a broader population. Unfortunately, these criticisms lead many businesses to ignore qualitative insights entirely. A recent report on research practices by Greenbook found that around half of all businesses use only quantitative data to make decisions, compared with less than a quarter that prioritise qualitative sources. While an arguably bigger problem is the roughly 20 per cent who use no data, it’s clear there is a bias towards quantitative big data at the expense of qualitative insights.
It is true that qualitative methods are subjective and usually based on a small group of people. Yet that doesn’t decrease their value. In fact, subjectivity is the exact point of qualitative research. When talking to customers directly, we want to hear and explore their personal, subjective views. It’s by diving into those messy, irrational, hard to understand thoughts and feelings that we might discover the key insights that can lead to creative innovations and strategies. In other words, subjectivity helps us understand the ‘why’ behind differing opinions or behaviours.
On this topic, I recently interviewed Pip Stocks, CEO of brand consultancy Brandhook and Founder of Hearsay – a platform that aims to help brands have better customer conversations. During that chat for the Shopology podcast, Pip spoke about how talking to customers can lead to what she calls ‘Wow Moments’: those sparks of inspiration that lead you down an exciting and creative path of ideas. Speaking directly to customers has a uniquely powerful ability to lead to these moments, as we can never quite know where a conversation might lead, or what a customer might say.
As another example of this, my PhD student Beatrice Romano is studying the impact of augmented reality in retail. Her first published paper1 explored the experience customers have when using AR at different parts of a customer journey. By speaking directly to customers as they were using AR, we were able to unearth a range of pros and cons we couldn’t have predicted, such as how it can mitigate the power of brands, and make consumers question their own creative ability. We summarised the results in detail in a previous article for Inside Retail. Suffice to say here that the only way we could reach these insights is with small, subjective, qualitative data.
Qualitative data can go beyond exploration as well. It can even be a means of creation. A team from Griffith University recently published a paper2 detailing what they call the CBE (Co-create, Build, Engage) framework. The team designed it for social marketing; applying marketing theories and techniques to encourage positive behavioural change. The first stage is inviting consumers in to help co-create a marketing program, often through a focus group setting. This process leads to better engagement and more effective marketing. But it can also provide a great insight for retailers into the power of focus groups in facilitating co-creation. For example, say you need to design your loyalty program structure or rewards system. Rather than just relying on data about how consumers earn and spend points, why not bring them in to help co-create and build the program? An immersive focus group setting with co-creation activities could lead to deeper insights than any other means, and help create a program that is more meaningful to consumers themselves.
Can we still trust surveys?
The criticisms of surveys are legitimate. They include reduced completion rates, multi-screening that leads to reduced data quality, and response biases. While improved survey design, attention checks, and better respondent control can mitigate these issues, a key problem still remains: there is a growing amount of evidence that reported intentions in surveys don’t always lead to actual behaviours. In fact, the link can be quite tenuous. Given that, are self-reported surveys still relevant when actual behaviour can be directly observed through big data?
The evidence presents a resounding yes, and suggests that surveys have a crucial role to play in understanding customers. A large-scale academic study recently compared attitudinal surveys with behaviour that can be observed online, and modelled the degree to which each could predict future sales. The study used data from 32 different brands across 14 categories, generating quite robust, generalisable, results. The conclusion was that surveys and behavioural metrics both have a key role to play. In fact, they predict behaviour much better together than either one does alone.
The authors suggest this is because surveys are better at measuring enduring attitudes, while online behaviours display more contextual or immediate interest. While this depends somewhat on product category (surveys may have less value in high-involvement categories), this finding about the complementary nature of surveys and behavioural data leads to the overall solution I’d like to present.
The solution: A data-agnostic approach
When different types or sources of data are compared, it is often done in a competitive fashion. Debates rage about quantitative or qualitative data, big or small data, and so on. Yet, there is an obvious answer these debates miss. Like the young girl in the popular Old El Paso ad, rather than debating which is better, we should ask ‘Why not both?’. Or even better, ‘Why not everything?’. In other words, the answer is an approach that is ‘data agnostic’ – not overly reliant on any one data type, but open to them all.
As data becomes more accessible, so do systems and platforms that make it easier to collect, analyse, and interpret different types of data. With a basic understanding of survey design, it’s now relatively easy to design and collect survey data, and the platforms will often analyse and visualise it for you. Similarly, platforms like Hearsay make it easier to collect meaningful qualitative data from customer conversations. Data visualisation tools and dashboards even make it easier to bring in multiple big data points.
The implication is that now it is easier than ever to combine multiple data types – even big and small. So while big data can have a lot of value, limiting yourself to only the observable behaviours it provides is unnecessarily reducing your understanding of customers. Why not supplement it with small data to fill in the gaps? For example, say you identify an interesting trend or change in purchasing behaviour. You could always try to guess why it happened, and you might be right occasionally. Yet, with some small data, like talking directly to customers, you just might find an answer you couldn’t have predicted, and it might be the secret to success. The evidence also suggests you’ll be able to better predict future behaviour and sales. So next time you see the term big data being thrown around, think about whether it really means all the data available, and if not, why not?
1 ‘Augmented reality and the customer journey: An exploratory study’ in the Australasian Marketing Journal, led by Beatrice Romano
2‘CBE: A Framework to Guide the Application of Marketing to Behavior Change’ in Social Marketing Quarterly, led by Sharyn Rundle-Thiele
3 ‘Enduring Attitudes and Contextual Interest: When and Why Attitude Surveys Still Matter in the Online Consumer Decision Journey’ in Journal of Interactive Marketing by Koen Pauwels and Bernadette van Ewijk