Qualitative research is a human process – but AI can help

Anyone involved in research will tell you a lot has changed in the past couple of years. New AI platforms are everywhere. Clients are being asked to do more with less. Business models are under strain. 

But we are now at the point where we know what AI can and cannot do. 

This month, I hosted an AQR webinar examining what makes great qualitative research with Kath Rhodes from Qual Street and Anne-Claire Pierron from Unilever.

Now all three of us use AI tools and platforms and get a lot of value from them. Getting an instant transcript from an interview is great. Being able to “pose questions” to your dataset dynamically is fantastic. None of us wants to go back to a time before these tools existed.  

But it’s important to note that AI tools are a supplement, not a substitute, to great human-centred research.

Human communication goes beyond words

In the session, I discussed how humans are multimodal beings: when we spend time with another person, we pick up on their body language, emotions and unspoken cues. We read the context. We intuit what is left unsaid. All of this gives us psychological proximity to that person. None of this stuff is in a transcript.

I’ll give you an example. I interviewed a small business owner recently. Her business was closing, and she was deeply unhappy. Her office was full of packed cardboard boxes, bills pinned to the noticeboard. She was wringing her hands and not making eye contact. There were many subjects she just didn’t want to discuss. 

The transcript of the interview was an incomplete record of our meeting. An AI summary of this transcript didn’t get to what mattered. The “data” which allowed me to convey her experiences in technicolour to our client, then to advocate for change on her behalf, was ephemeral and invisible to anything other than an in-person human connection. 

Getting to the meaning

Part of the issue with qualitative methods is that the analysis process that gets to meaning and insight, is a bit fuzzy and circular. To provide THE answer, not just AN answer, you have to think, draft, edit, revisit the data and reformulate multiple times.

Kath Rhodes shared her framework to help shine a light on this process:

  • Perception: Ongoing observation, listening, and sensing, rooted in human cognition and context.
  • Connection: Engaging with people and context, challenging assumptions, and building rapport, best achieved in person.
  • Reflection: Analytical thinking, synthesising observations, and allowing time for creative, unfocused thought to generate insights.
  • Direction: Translating insights into actionable recommendations, focusing on business impact and future strategy.

This model helps surface the value that qual offers. Analytical thinking is iterative – depth and accuracy require continuous revisiting and correction. We need to talk about this process with clients to get credit for it. 

Advocating for change at Unilever

Anne-Claire talked about her experiences as a researcher on both the agency side and as Global CMI Manager at Dove Masterbrand for Unilever. 

Taking AI outputs at face value can result in bias and oversimplification. Qualitative thinking is needed if you are using AI tools to avoid making wrong calls. You have to go back to interview recordings to verify and check emerging themes and align with what the data really means. 

She also reminded us that clients are flooded with data. But abstract data does not produce the kind of emotional response that compels action. 

This is where qualitative research has power. 

In her view, conviction is the superpower of the qualitative researcher. Deep listening means we “feel” the data we collect. This means we can convey it powerfully. It allows us to inculcate change in organisations like Unilever. 

In summary 

So in summary? For great qualitative research, AI is a supplement, not a substitute. What is easy for humans is hard for machines, and vice versa. AI tools can help with speed and scale, but humans will thrive at the uncertain boundaries where human connection is essential.

SPEAK TO OUR EXPERTS