Technology has the power to subvert long held beliefs. AI’s impact on chess is a good example, with computers introducing radical and effective moves that have reshaped how humans now play the game. What’s true for chess holds for insights, too. 

Twenty-five years ago, collecting quantitative research responses would take weeks, with data tables landing on desks much later. Then someone realised the internet could speed up the process. Today, we can conduct a quick poll in minutes or a robust one in hours. 

Insights haven’t just got faster; their accuracy has also improved. Ten years ago, analysing the detail of a national election campaign meant scrutinising small changes in demographics in a nationally representative poll. Now, with advances in statistical modelling and machine learning, as long as we use good quality data, we can make highly accurate constituency-level predictions. For example, in recent elections, the best MRP (multilevel regression and post-stratification) models projected results correctly in over 90% of constituencies, all while voter volatility reached new heights. 

But while technology has helped lift quant, qualitative research has progressed at a slower pace. Online focus groups have improved efficiency – with things like text-based focus groups adding anonymity and personal video responses capturing a personal picture – but the gap between quant and qual has widened. However, AI is changing that dynamic. It’s increasing the speed, sophistication, and quality of reporting in qual, bringing the two disciplines closer together in a tighter race.  

In the near future, it’s likely that quant and qual will no longer be seen as distinct pursuits; instead, they will merge and blend, complementing each other by providing simultaneous instantaneous insights into both the “what” and the “why” and giving a clearer view of reality. 

The relatively higher cost and lower speed of qual services created the perception of it being the less favoured sibling in the research family. But AI is elevating qual, making it faster and more effective. AI models can analyse large volumes of text and quickly identify themes, allowing us to process responses from thousands of people in a day and extract key insights within hours — a feat unimaginable even a few years ago. 

This AI-driven theming also enhances quant research – in reducing researcher bias, for example. Now, we can listen to respondents’ experiences in their own words and draft questionnaires that accurately reflect their language and concerns. For example, a 30-year-old researcher in a large city might struggle to fully appreciate the concerns of a senior citizen in a rural community discussing fuel prices. But with AI, we can hear those concerns directly and ensure our questionnaires capture them authentically. 

The practical applications of AI in qual are vast. If a company is unsure about an ad campaign, they can show it to a group of consumers and get immediate feedback. Qual can be the canary in the coalmine, surfacing negative public reactions before an ad goes live. This real-time feedback is becoming more important in sectors like tech, where UX and UI use similar techniques to traditional qual methods like ethnographies and in-depth interviews. 

These tech advances also allow us to properly analyse people’s thought processes in the right way. Daniel Kahneman’s Thinking, Fast and Slow describes two systems of thinking: system 1, the fast, instinctive mode, and system 2, the slower, deliberative mode. For years, we’ve only had the tools to look at system 2 behaviour at scale, but with AI, we can capture their system 1 responses more easily. 

For example, we can now collect thousands of responses from people via their phones, with participants recording voice notes or videos explaining their decisions. This approach allows us to capture immediate, authentic reactions — why someone chose one supermarket over another, for example. AI not only enables us to do this quickly but also allows us to scale the process across markets and languages, with responses transcribed, translated, and analysed in days rather than months. We can also extract key terms and ensure companies reflect their customers’ language. We can track their behaviour online, identifying the exact point where they chose one service over another.  

AI is putting qual on a fast track, and the gap with quant is closing. In a few years, it’s likely we won’t distinguish between them. We’ll ask questions in a way that mirrors how people naturally communicate — through voice notes, short messages, and videos. Their answers will merge the what, why, when, and how, and AI will help us disentangle and quantify these insights, offering scale and depth.  

As AI continues to evolve, the line between qual and quant will blur, and the insights we gather will become faster, more nuanced, and easier to access and understand than ever before.

Jake Gammon is global head of qualitative research at YouGov