The Big Qual Revolution: How to Get the Richness of Qual at the Scale of Quant

The classic dilemma that accompanies almost every brief: qual or quant? Is it more important to get the exploratory, open-ended, rich qualitative insights that can only come from a free-form conversation? Or will your stakeholders only be convinced by data with a representative sample and significance tests, bar charts and top boxes and conjoint analysis and n=1000?

Each mode of research has its advantages, and each has its place. Qual gives you depth of insight, persuasive snippets of real consumer voices, and unexpected discoveries. Quant has robustness, comparability, volume estimates, and confidence to extrapolate to the real market.Of course, each also has corresponding drawbacks. Qual raises the specters of luck and bias: what if the consumer who said that thing doesn’t represent anyone but themselves? Maybe when they claimed to love the new concept it was because the moderator inadvertently led them down the path and opened the gate for them. Quant has its own risks of tunnel vision: you only get insights into the questions you remembered to include in the questionnaire.

What if you could get the best of both? Research that’s rich and relatable, and representative and robust? Exploratory and empathetic, and quantifiable and quick? Quality data and actionable insights?

 The ‘big qual’ movement promises to provide all this. A host of vendors are developing tools to analyze open-ended conversations at quant scale. These take a number of different forms but all provide consumer insights that blend qual and quant.

 Some tools read in large volumes of text data that you already have, and generate insights from it. These might be your NPS verbatims, or open-end responses to surveys, or customer emails. Natural language processing and AI, often with human-assisted coding of keywords or phrases, are used to find patterns in the data. The outputs can include sentiment measurements, relative frequency of different themes, and similar statistical summaries.

 Some products instead source their data from social media. Typically these read millions of tweets or posts that match specific keywords such as your brands, category names and other topics of interest. These are summarized into trends that can be tracked over time, potentially with geographic or demographic data superimposed (subject to the limitations of social media data quality – Twitter in many cases can only guess at the gender, age or location of its users).

 Variants on these tools might hoover up other types of unstructured data – typically video or audio – and produce different forms of visualization. AI-powered automatic transcripts can be an essential part of the process here.

 A third approach gives you more control over the topics of discussion, and more accurate segmentation of respondents. In these tools (which include Irrational Agency’s System 3 approach), a survey is created and targeted to the specific respondent base you want to hear from. The questionnaire includes carefully designed open-ended questions to provoke the right kind of responses – narrative storytelling and imaginative ideation for example. Then the respondents do their own coding, generating word associations from the stories they themselves have told.

 At the end of this process, a quantitative analysis of the themes, beliefs, desires and attitudes of your customers is generated automatically. This can be visualized in different ways, for example a mindmap-style graph showing the key concepts and mindsets, and the connections between them. Statistical techniques can then be used to identify clusters of related topics, common narratives, and worldviews that are distinctly associated with particular groups of respondents.

 This third approach has a major advantage: you are not limited to the topics people choose to discuss on social media. Nor is your sample biased towards the people who spend most time on social media (or possibly worse – the people who fill in the verbatims in an NPS survey!) You can ask about any topic you are interested in, and target whoever you want to speak to.

 Once you have identified a theme in the data, you can access the verbatim quotes and demographic data of the people associated with that theme. Every topic or narrative you discover can be backed up with examples and stories in the respondents’ own words, to show where the insight comes from. Research shows that communicating via stories is much more memorable than just presenting data. The blend of qualitative stories with the credibility of quantitative information has the best chance of persuading stakeholders and ensuring your message sticks with them.

Whichever method you choose, there will always be a place for small qual and big quant. But more often than not, big qual might just resolve that familiar dilemma. Imagine writing your next brief and instead of ticking the ‘qual’ or ‘quant’ box – you get to pick both!