– Nichola Quail
On this week’s On Work and Revolution podcast, I’m joined by a guest that’s been working with AI long before the whole world got wind of this. Nichola Quail is a Global Insights Expert and Founder and Director of Insights Exchange. She has been driving ‘collective human intelligence’ for over two decades. In this episode, we dive into the impactful relationship between AI tools and humans and the potential that exists to free up our time to focus on higher-level thinking, genuine human connection, and contextualizing AI-generated insights.
Specifically, Debbie & Nichola dig into:
✓ The role interpretation bias plays in forecasting, analyzing for future predictions & recommendations
✓ How humans and AI need to work together to contextualize the data we have access to
✓ The need for 100% disclosure when AI is involved
About our guest Nichola Quail
Nichola Quail is a Global Insights Strategist, Research Specialist, and Demography Expert. As Founder and Director of Insights Exchange, a platform that connects businesses with a global network of on-demand research specialists, Nichola has been driving ‘collective human intelligence’ for over two decades.
Insights Exchange partners with Research Technology companies to deliver game-changing data and insights with leading brands such as Xero, Dentsu Aegis, Bravo Media, Bendon, Disney and Panasonic.
Nichola is passionate about a number of key industry topics, including the Professional Gig Economy and how AI will influence market research now and in the future.
“We do believe that ‘slow and steady’ wins the race when it comes to in-depth analysis, but many researchers, including myself, also like to mix some of the traditional ways with the latest AI tools to help us create a full picture for our clients.
We don’t shy away from trying new technology to help save our clients time and money, and our team of researchers have the benefit of using AI to validate their findings and perhaps find some new ‘golden nuggets’ in their analysis.”
Follow Nichola on LinkedIn
Open for Full Episode Transcript
Open for Full Episode Transcript
Debbie Goodman 00:05
Welcome to On Work and Revolution where we talk about what’s shaking up in the world of work. I’m Debbie Goodman, your host, and today we have as our guest Nichola Quail. So Nichola is the founder and CEO of Insights Exchange, which is a disruptive market research platform that connects high growth businesses and brands with insights, analytics and data strategy experts. So their clients include very large multinationals, like Xero, Panasonic VW, as well as some earlier stage companies who are expanding to new markets, new products, and who need data driven insights for their business decisions. And today, we’re going to talk to Nichola about the impact of the massive explosion of generative AI tools on the world of market research and data analytics, what this means for the market research industry as a whole, what this means for all of us as consumers, and how generative AI will further transform the industry, our use of consumer data in the very, very, very near future. So welcome, Nichola.
Nichola Quail 01:14
Great to be here. Thank you.
Debbie Goodman 01:17
So everyone’s talking about AI, or maybe you’re just in my little echo chamber, it feels like everybody’s talking AI. I think on a daily basis, I’m consuming, I can’t actually count the number of articles, webinars, videos, etc, that are really talking about the massive advances even in the last few months. You can ask them for anything, with the right prompts they can pretty much give you some mind blowing information. And this naturally would have the most phenomenal impact one way or the other, on your particular industry. So first question is, how has this started? Or has it started to disrupt the market research industry, which, at its very core, has the collection and gathering and making sense of data at its backbone?
Nichola Quail 02:07
That’s exactly it. It’s funny. I was, you know, at a conference a few weeks ago, and very much the same, it was front and center hot topic, you know, is it going to take over our jobs? Is this industry even going to exist in five years? But it’s with what has happened with ChatGPT. So yes, absolutely has almost overnight changed the game in terms of opening up the power of AI to a wider audience. But we’ve been playing in this space for some time now, particularly in the structured data space. So in the quantitative space, as you can imagine the volume of data that comes out of doing, you know, large scale research. And so companies like Yabble, a New Zealand startup in the AI space, Get Fematic, they’ve been doing AI analysis for many years now. But what it’s been educating the market about that. And so what’s changed the game with ChatGPT is not only the availability of this tool, but now that suddenly everyone is kind of jumping on board and understanding what it can do for them. But also, the ability for unstructured data analysis, which is, as you said, is a game changer. Structured data is where you’re, you know, it’s numbers. So it’s very, you know, it’s your ones, your twos, your 80%, your 30%. And it’s an open ended question, you know, and answer. So it’s very, you know, ask it one thing, and you’re pretty sure what you’re gonna get on the other and you can code that out into different themes. Unstructured data is a conversation and this is where the game changing nature of this next evolution of AI, so it’s, you know, anything that happens on social media, you know, anything that happens in a consumer review, anything that happens in our world qualitative research, so that is where we go into an interview or a focus group, we’ve got some hypotheses, but we’re not totally sure what’s going to come out. And you can imagine the sheer volume of that those conversations. And back in the day, we would have to manually read through, you know, pages and hours and hours of transcripts. So the transcript business model is certainly a key one that’s going to be disrupted. And then we would have to find themes, and we would code that and try and turn that unstructured data into structured data. So it would be like 80% we’ve, you know, 80% are saying this, but it was a very long and manual and expensive task.
Debbie Goodman 04:56
Okay. So as I’m understood it, it’s the ability now to to aggregate synthesize, collate massive amounts of data points incredibly quickly and come up with some kind of sense at the end of 30 seconds.
Nichola Quail 05:09
That’s exactly it. But not only data points. And this is, again, the key difference – human data points. And that’s the difference. Because when we are dealing in, you know, code, numbers, statistics, you know, it’s easier to train a model to understand that because it’s essentially just advanced coding. Where the game changer has happened is, again, it’s called natural language processing. So it’s actually interpreting and aggregating human conversations and human data. And we talk in many different ways, across many different countries, in many different languages, we have our own vernacular, we have our own slang, we have our own meanings for things. And so qualitative researchers had to interpret that and find those meanings, and then translate that into a strategic insight for a brand. But what’s happening now with this next evolution of AI, is that the training models, the huge huge training models that are now been opened up to the world, it’s almost just, you know, it’s speeding up that process. So now that they can interpret at scale, human data.
Debbie Goodman 06:24
Okay, so the interpretation piece is something I want to ask more about, interpretation is always open to bias. And so I’ve heard very conflicting viewpoints on this, the one saying that, if there’s bias in whatever is being interpreted, if there’s bias in the language, then the AI is going to feed off that, and the analysis or the insights, or whatever it comes out of it will have inherent bias. And then I’ve heard other viewpoints that say, actually, you’re excluding human interpretation, you, you’ve just got the AI tool doing that, and therefore, you reduce bias. So what is it?
Nichola Quail 07:03
This is, as you can imagine, a very hot topic, but you know, has been a hot topic well, before AI, and you know, even in just how things used to be coded, you know, dare I say it used to be sort of your young European males who would do traditionally coding. So bias is everywhere, you know, bias is humans have bias. And AI naturally will have some bias in it because of how it’s been trained. But where, again, the evolution of this next stage of you know, with ChatGPT means that the sheer volume and diversity of data sets that it’s now being trained on. So it’s essentially data in data out. And so with AI, it can only learn what has been put into it. But the volume of what’s been put into it now means that, you know, yes, there’ll have to be regulation around certain types of bias. And I think that’s possibly even the wrong. It’s, it’s a misnomer. It’s almost about the diversity of the data that is coming out. And we know for a fact that it is a much more diverse data set, you know, emerging, but then that’s also the role of, you know, in our space, human researchers highly experienced, you know, specialists in this space that deeply understand human behavior, it’s about contextualizing what comes out of the AI. So look, if you just go to a client and present something that the AI spat out, and you don’t disclose that it was generated by AI, then, you know, potentially, you’re opening up a door for somebody to say that’s incorrect, or, you know, I don’t agree, and this is where humans and AI now need to work together hand on hand, because we still need that human lens overlaying and interpreting and ensuring that exactly to your point, what the AI is saying, based on our knowledge of, you know, a culture or a brand context or a consumer buying pattern based on all our years of experience that, you know, it’s it’s on track, and then it can enhance our ability to go deeper with those insights, you know, do higher level thinking around that, because all it’s really doing is just taking away that work and that effort and that time of trawling through all of that data, and it’s just surfacing some of those, you know, all those high level insights, but then it gives us the ability to go go deeper on that.
Nichola Quail 07:28
Aside from the ability to accelerate and trawl through so much data very quickly, my question is, what about the interpretation of that in terms of forecasting and analyzing analytics and insights for future predictions or recommendations that I would imagine is also core to your business, you get you gather the data in order to make some recommendations. Can the AI tools do that well enough yet? Or are they paused at the point of here’s the information, now you decide what to do with it.
Nichola Quail 10:18
That’s pretty much it. But I’m saying that now in April 2023. So, yeah, I’m sure in twelve months’ time that won’t stand true. But look, it’s only learning, you know, it’s not learning even what’s happening today. You know, it’s, it can only learn up to a certain point. But again, it could absolutely predict. And, you know, it’s called predictive analytics, it’s an industry in our in our space, and oh, look, a hand on heart 110% I would say, yes, it can. But again, this is where the human you know, so our researchers, they understand the client, they understand the needs of the client, they understand the client’s business model. Yes, you can feed that all into the AI, and it’ll give you a great recommendation. But then how do you translate that into human speak, you know, and, and even have a person presenting it, and take that client on that journey and say, this is where, you know, we recommend you could go in five years? Let’s workshop that out. Let’s talk that out. Are there any fears? Do you have any concerns? You know, you’re having a conversation. And look, you know, if I was a client, I’ve worked with a lot of fast growth startups, this market research space is very new to them. And they’re nervous about talking to their customers and consumers, because, you know, they’re not always sure what they’re going to say. And they’re like, will they like my product? or will they like my service? And you know, no AI can sugarcoat that. It’s very blunt in how it will, you know, talk about that and summarize that. But it’s then the delivery and how you layer that. A human interface is still required.
Debbie Goodman 12:04
So let’s talk about that the augmentation or supplementation in the workplace, with AI and humans, because I think certainly in my world, we run a group of executive search and leadership coaching companies, it’s all about people and their advancement and their careers and their next jobs. And there’s a lot of fear around particularly those who are witnessing before their very eyes, how these AI tools, particularly ChatGPT, which is the most, you know, it’s got the biggest adoption, how that is taking on certain tasks. I was reading a stat that said that currently even at management level, AI tools can take on 31% of manager level tasks. And so there’s no job, certainly in the professional knowledge worker space that’s not going to be impacted by this. I think a lot of the time when we’re experimenting with these tools, people are looking for, you know, slight opportunities to go, “oh no, but it can’t do this as well as me. And it can’t do this like a human and, oh, we’ll always have our place here for x”. But each day, I’m seeing those beliefs being chiseled away bit by bit. So what do you think, currently, and perhaps even for the short term future, the relationship between the AI tools and humans are where does the one supplement the other? Where’s the other needed? What are the key elements to that?
Nichola Quail 13:29
Look in our space, and I think in many different industries, all it’s doing, you know, if you think about the calculator, if you think about the computer, if you think about Google, you know, we’ve been through the disruption. And yes, this disruption is probably like nothing we’ve ever seen before. But it’s more just the speed than anything else. I mean, Google was pretty disruptive in its time. And really, we will adapt, you know, we’ve adapted for 1000s and 1000s, and 1000s of years, we’ll adapt again to this, and some jobs absolutely are going to be basically non existent in a few years time. But it’s really just how we adapt. And for me, I see it as a huge opportunity. Because, again, like, when we got the car, you know, it’s all it’s doing is taking away some more of those manual tasks. And in our world, it’s, you know, it’s data analysis, it’s churning through high volumes of data, and only because, you know, we now produce so much data. And so then for me, it frees up us to be more human and, you know, in many different industries, having in person conversations, I mean, you think about what we lost during COVID. You know, having those in person, really physical contact, you know, but you know, we are in qualitative research, we’re in the world of, you know, body language and facial cues and, you know, speech recognition, but just again having a human conversation. And I don’t know about you, but I don’t get the same feeling from my computer, as I do having a human to human conversation, and I’m sure that will come. But you know, and you know, I think about, you know, even in recruitment, like, yes, great, it’s going to cut down all that annoying thing of going through CVS and, you know, but having taking someone on that journey, briefing them on their first job interview, you know, that’s, that’s human that’s like, and you’ve got someone in your corner and you trust them, and you feel safe with them. And in our world, you know, we do that with consumers in terms of we want to understand their fears, their concerns, their needs, you know, how they feel about their kids, and what that means for a consumer brand. You’ve really got to take someone on a journey to get them to open up like that. And yeah, it’s sort of feeding that into an AI machine. That’s, that’s past, you’ve still got to go and get the data to feed into the AI. And that’s still a human role.
Debbie Goodman 16:08
In your current world of work. Like right now, April 2023. Are you disclosing to clients when you use generative? And what is their, what is their approach? Do they like it? Are they skeptical? Tell me, how’s that working?
Nichola Quail 16:23
Ah, look, you know, it’s funny, and this came up in a webinar recently that my colleague was on, Catherine Top from Yabble. So we’ve been partners with Yabble for years now. Our businesses have grown up together. And we say it with pride that, you know, our AI partner Yabble is baked into our process, and it’s all about providing efficiencies to the client and value saving them money, you know, we’re like, gosh, we get to, you know, pay our researchers for the high value work, not the, you know, churning through loads of data and spending hours that you’re paying for to do that. We 100% disclose where it’s been generated by AI and, look, I think, you know, pre ChatGPT people didn’t quite understand what that meant, and didn’t really worry, you know, it was just part of the process, some were intrigued by it. Whereas now everyone really understands it. And we certainly haven’t had any pushback, it’s more how much more can you do as a result of this, and, you know, we know what’s coming, sort of, in our space, and many other spaces is, you know, the analysis of video. So, for example, us recording this, you know, the AI will be able to analyze our facial features, you know, our body language, our muscle movement, what we’re really feeling and, you know, images and sound. And we certainly know that’s coming in different products and devices that are going to be game changing, you know, for health, healthcare, and all that. That we’re testing at the moment, but it needs to be 100% disclosed, mainly, because also, I would want it to be because, you know, it’s what we add on top of that, that is the secret sauce, and you know, so then I go, well if it’s all just AI, they’re gonna get the same report from lots of different people. And it’s all just AI generators, you know.
Debbie Goodman 18:15
How are you seeing the impact of the projects that you’re delivering on now using AI? Is it just about speed and efficiency? Or is there something else that’s changing the nature of your work?
Nichola Quail 18:28
At the moment, it’s just about speed and efficiency, and then providing those, you know, handing over those cost savings to our clients, but we know that will change and that’s going to change quickly. So then it will change into the types of analysis we can do. There may be times when No, we don’t actually have humans involved in our types of research deliverables. But where I’m excited about that, because I’ve always been excited about being able to provide as many different businesses as possible with the access to data and insights to help drive growth, you know, it’s something I’ve always believed in, but it’s been quite cost prohibitive for a lot of businesses. And that is because of the human cost of getting fantastic researchers on board and as well as consumers. So this in terms of startups doing, sort of, you know, whether it’s market, you know, you know, analysis, a competitor analysis, you know, even product market fit. So, for them being able to have access, still led by an expert, but having one expert that can help sort of craft that journey, you know, feed in the right questions to the AI and then be able to interpret the results in a, you know, with an expert lens, but the cost of being able to do that will significantly reduce. And then for me, you know, imagine every SME which is the backbone of you know, most countries’ business community, then, you know, for them that’s a huge game changer to have that access.
Debbie Goodman 20:08
So kind of democratizing access to market research as a tool, because now the grunt work of it is being automated in such a simple way. And the real high value is more affordable. So that sounds like that’s a principal very optimistic, look at things. I’m excited and optimistic. Even though I am a little nervous about the dark side of what some of these tools are already showing up as being able to do. I’ve seen some fascinating things on social media, in the last few days, literally. There is the potential for the dark side. But I’m very excited to hear how this can positively impact all businesses with let’s just talk in this one tiny niche of market research, where now all of a sudden, it becomes so much more affordable, that humans can be used for the high level stuff, and provide access to organizations that wouldn’t have been able to even consider using this as a tool for business decisions, because it was just too expensive. That’s like an blue ocean market for market research all of a sudden.
Nichola Quail 21:13
That’s exactly it. So yeah. And that’s what really excites me. So and again, it’s an AI and a human working together. I don’t think you should, my personal opinion is, you know, startup founders shouldn’t go crazy and going into it on their own, you know, unless they really know what they’re looking for, what questions to ask and what they’re going to use with the data. And that’s where having some expert guidance can really help.
Debbie Goodman 21:37
Yeah, well, actually, let’s just hone in on that. Because now all of a sudden, everybody is an expert. You just use the right prompt. And now everybody is a market research expert, because we’ve got access to all this data, but I guess it’s a little lacking in rigor, and, you know, frameworks for proper market research?
Nichola Quail 22:01
Well, and sometimes it makes stuff up, you know, it just, it’s kind of like, oh, I need to find an answer. Because, you know, that’s what I’m supposed to do as an AI tool. So then it just makes stuff up. And, you know, to the untrained eye, they don’t know any difference. So they just take it as gospel. And that’s where I just go, you know, there are people out there with a lot of experience that can just do a bit of a sense check and fact check. And exactly, to your point about, you know, the dark side. I mean, I won’t go into politics, that could get me in trouble. But yes, fact checking is a good thing.
Debbie Goodman 22:37
Nichola, this has been a fascinating conversation, I’m delving into so many different industries that are starting to feel the impact quite suddenly on their doorstep. And some are really fearful and running for the hills. Others are like you, so optimistic about the potential but you know, cautious and also not racing too far ahead in terms of what the AI’s capabilities actually are. This is an evolving story. I feel like, you know, breaking news, but actually an evolving story, and amazing to hear what you’re doing. And thank you so much for joining me.
Nichola Quail 23:12
Oh, great to be here. Thank you so much for the opportunity. You know, it’s an exciting time and the genie’s out of the bottle, we can’t go back. And so now it’s just about how do we adapt, evolve, and use it in the best way possible to improve humanity.
Debbie Goodman 23:27
Okay, I’m gonna end on that note right now, thanks. Thanks for hanging around all the way to the end. It would mean the world if you would rate and review On Work and Revolution on your favorite listening app. It helps people know that the show is worth listening to. And so I will really appreciate that. Thank you so much.
LEAVE A REVIEW
Kind podcast reviews help us make awesome content for you!