Do you remember in 7th grade math, your teacher would give you a problem to solve and say, “Make sure to show your work”?
So you’d get out your lined paper (yes, paper!) and slowly write down the equation… divide by 6, and then add 2, and then carry the 4, and so on and so forth. And you’d hand it back to the teacher, dog-eared, and covered in white-out and crunchy peanut butter from breakfast.
Why did our teachers do that? At the time, it felt like they just wanted to torture us, slowly, dinging us for every little misstep we made. But in reality, they were making sure we actually understood the mathematical concept we were being taught. Clever, right?
So how does this apply to AI when it comes to qualitative analysis?
Listen, I get it. AI analysis tools seem great. They give us an answer, and pretty damn quickly. But is qualitative analysis only about getting an answer?
I would argue, no. Well, more specifically, not quite. Qualitative analysis should be just as much about getting to the answer. The messy part in the middle.
- The scribbles you left on a dinner napkin (right beside the gob of spaghetti sauce).
- The mind maps in your notebook (that only you can decipher).
- The notes in your phone (that you captured right after a focus group).
- Oh and the Google doc comments from your team (the many, many Google doc comments).
That’s where the true magic lives.
So if you’re like me, sitting there, nervous to use AI to conduct qualitative analysis… good! We should be. In fact, as human beings, it’s our responsibility to be. Because if we rely too much on AI to find themes and insights for us, we may be getting an answer of sorts, but we’d be missing the process… AKA “the work”.
What is “the work”?
When it comes to qualitative analysis, so much more goes into “the work” than just data. The inputs, so to speak. And by inputs I don’t mean prompts. I mean the human inputs we organically (and often subconsciously) feed into our analysis when we’re doing “the work”.
Here’s just a few:
Our intuition – When we’re neck deep in analysis, certain things we read or hear pull on us. Not necessarily on our heart strings (although they could), but at the very least, certain things capture our attention. It could be something unexpected we observed in field, or a feeling we get after a brainstorm call, or one, single word we read in a transcript. And bam! It sticks. Something in us says “Hey! Look more closely, there’s something here…”. As researchers, we gotta keep listening for that. Whatever that is. Because that is something that can’t be explained (let alone programmed). That is the f*cking gold!
Our experiences – We wake up every morning not quite knowing what the day will bring. In fact, we encounter millions of micro-moments, half the time we aren’t even conscious of. Whether it’s the soft smile you get from a passing stranger, the casual conversation you hear between the bus driver and another passenger, the taste of the burrito you’re enjoying for lunch, or the look you get from your partner when you come home without an anniversary card (again). All of these experiences are being lived…processed… and stored, for later use. So, when we sit down to analyze people’s relationships with sunscreen, soda, or food delivery services, everything we read, watch, or listen to gets filtered through our lived experiences. These experiences quietly dictate how we interpret the information in front of us. Some may call that bias. I call it humanity.
Our emotions – A really good qualitative researcher doesn’t just think through a set of data. They feel through it. This requires 3 very important things: 1. The ability to recognize and identify the emotions of the people you’re studying, 2. Being able to recognize and identify the emotions you, yourself are feeling. 3. And being able to connect the two. In short, it’s about building empathy, between you and your audience. Human to human. Having the ability to read someone’s story about the fight they had with their teenage son and how frustrated and infuriated it made them, and knowing exactly what that frustration and infuriation feels like. Make space for that. It’s work, but it’s worth it.
Our vulnerability – The hard reality is, as qualitative researchers, consultants, and strategists, we never really know if we have the right answer or not. No matter how many hours we’ve locked ourselves away in a meeting room, or verbatim we’ve collected, or nods we get from the client in a debrief. There will always be a margin of error. The error could be in what we observed. The error could be in how we observed it. The error could also be in how we interpreted what we observed. The point is, we will never really know if we got it 100% right. And I think that’s ok. That’s the inherent vulnerability of being a qualitative researcher. Surrender to it! But also, don’t be afraid to bring in re-enforements. Other people to collaborate with and help build confidence in your insights.
And dare I say it, our subjectivity – To say I am an objective researcher would be flat out lying. For all of the reasons stated above. My intuition, my experiences, my emotions, my vulnerability (all things I am fairly confident AI, at least for now, doesn’t have)… shape the way I analyze data. But, quite frankly, I wouldn’t change it for the world. It’s what makes me human. It’s what allows me to connect with the people I’m studying on a deeper level and be able to tell their story in an authentic way. Which is the job, right?
So, where do we go from here?
Now, don’t get me wrong. This doesn’t mean we break up with AI altogether, giving it the old excuse of “It’s not you, it’s me”. Nope. It means we need to continue to explore what AI is capable of… while still, doing “the work”.
For example, using AI to:
- Sense check what you think you already know – When the transcripts (and your gut) are both indicating you have an important theme emerging – whether it be a tension, a value, a behavior or an aspiration – and you want to confirm you’re onto something, plug it into AI. If AI surfaces that same theme, you can feel more confident in including it in your analysis. Let’s face it, a little validation can’t hurt.
- Texturize and bring more depth to your thinking – You may not expect AI to identify deep, meaningful insights on it’s own. But if you’ve uncovered a deep and meaningful insight in your analysis, you can plug it into AI and ask it some follow-up questions, like: Why is this happening? What does this look like in people’s lives? What could this mean for the brand? It may not have all of the answers, but you never know what thinking this could spark.
- Fill in the gaps of things you might have missed – As human beings, we’re not wired to keep track of every single detail we hear and observe in field. We actually suck at that. So instead of having to comb back through the transcripts ourselves, we can ask AI to do that for us. I especially find this helpful in completing component lists during analysis. Like a list of triggers and barriers, the steps of a process, or a set of brand associations. It helps to know you’ve covered all of your bases.
- Illuminate the people you’re talking about – Finding the right quote or clip for a report can be crazy making at times. It’s gotta have the just right amount of detail, emotionality, and impact. And so we tend to gravitate to the verbatim we remember in field, often from our favorite respondents (because if we’re being honest, we have our favorites). But if we use AI to source quotes for us, not only could it lead to richer verbatim, but it could also eliminate some of the natural bias we have when making our selections.
None of these are overly sophisticated use cases by any means… but that’s kind of the point. Instead of expecting AI to give us deep and meaningful answers (and be disappointed every time), we need to find ways to make AI “work” for us.
All I know is, AI is not going anywhere. But neither are we. As human beings, we need to recognize and protect what we bring to qualitative analysis that AI quite simply… can’t.
So, let’s make our 7th grade math teacher proud and keep showing our work, shall we?
If you’ve got a brand challenge that needs solving, give us a shout. We’re ready to do the work.
Written By:
Brynn Harris
As a founding member of the Sound, Brynn has played an integral role in shaping what The Sound is today. She works closely with all our teams, uncovering powerful narratives that move people and brands. She is the mastermind of our Sound Track process and ensures that our work is always exceptional.
More by this author
