AI and Advanced Methods: Solving the Research Dilemma

Editor’s note: This article is an automated speech-to-text transcription, edited lightly for clarity.  

Thursday, February 1, quantilope presented during Quirk’s Virtual: Innovation series. They discussed AI and its role in the research industry and introduced the audience to Quinn, quantilope’s AI co-pilot. 

Quantilope's Jannik Meyners, director of data science and Katelyn Ferrara, senior director of global clients, were joined by Sara Martens, brand insights manager for Mutual of Omaha for a discussion on the uses, hesitations, benefits and future of AI in the industry.  

Read the transcript below or watch the recording to hear all these three researchers had to say about AI. 

Session transcript 

Joe Rydholm: 

Hi everybody and welcome to our session “Solving the Research Dilemma with AI and Advanced Methods” I'm Quirk’s Editor, Joe Rydholm. 

And before we get started, let's quickly go over the ways you can participate in today's discussion. You can use the chat tab if you'd like to interact with other attendees, and you can use the Q&A tab to submit questions to our presenters and we will address them during the Q&A portion after the presentation.  

Our session is presented by quantilope, Katelyn, take it away. 

Katelyn Ferrara: 

Awesome. Thanks so much, Joe. Hello everyone. Welcome. This is Katelyn Ferrara, and today I am excited to share with you all how you'll be able to significantly increase or double your productivity with AI. We are thrilled to talk about some of the new developments here at quantilope that will increase your research productivity. And then following that discussion, we'll have a fireside chat with Mutual of Omaha on how they perceive AI in relation to consumer research.  

But first, as I said, my name is Katelyn Ferrara. I'm the VP of Client Development here at quantilope and I'm joined today by Jannik Meyners, quantilope’s director of data science, who now leads the automation of advanced methods as well as the application of machine learning and AI within the quantilope platform. 

Today, our goal is to make AI approachable and applicable for you. As said, we're going to demonstrate some of the ways you're already using AI, perhaps without even knowing it. As well as show you some real examples of how AI is going to make your jobs as researchers even easier and significantly more productive.  

At quantilope, we've had AI ingrained into our DNA since day one. The reason why we continue to develop and push the boundaries of AI and machine learning in our platform is simple. Ultimately, at quantilope, we've witnessed firsthand the demand that our clients are receiving for more and more consumer research initiatives as their consumers continue to change at faster and faster rates. However, it's very clear that at the same time we're also seeing capacities shrinking as budgets are being reduced alongside a very challenging economy, not just in the U.S. but globally across all our clients as well.  

When you combine these two things, you get what we call the research dilemma, which really is the notion of having too many requests for projects and insights with too few hands or resources to be able to complete them in a thoughtful way.  

Thankfully, this research dilemma is something that we have been working to solve for nearly a decade, much longer than ChatGPT or any other buzzwords around AI. But I want to quickly explain how we're doing this and how we're thinking about this.  

So, quantilope is working to solve this dilemma through our consumer intelligence platform. And our consumer intelligence platform is more than just a research platform. It is pure empowerment for insights managers enabling over 300 clients just like you. And for those who might not be fully familiar with quantilope, we are an automated end-to-end consumer intelligence platform.  

Our platform connects and automates the entire research process, empowering brands to do research better, faster and more efficiently. By doing so, we not only give you direct access to your data, but also enable you to run more projects and spend more time on the creative and not the repetitive side of research. On top of that, we offer the largest suite of advanced methods, automated advanced methods within our platform that are lean, flexible and most importantly easy to use. All of this is fueled by AI and machine learning. Our platform is combining the power of speed, substance and scale to remove the bulky and inefficient research processes.  

That's ultimately quantilope in a nutshell, but I want to talk about AI.  

So, amid all the hype and skepticism in many cases associated with AI today, it can feel overwhelming or confusing and trying to understand what we actually mean when we say AI. 

For quantilope, that's meant leveraging technology to replace manual and repetitive tasks, which we've been weaving into our platform for years. A few examples of that have existed for a while include data analysis of advanced methods, automated data cleaning, predicting the length of a survey so that clients have a more accurate ROI that they can estimate based off of and analyzing video and text at scale and many more.  

If you work with quantilope or a variety of other technology-based tools, you've most likely been leveraging AI already for years. And we're obviously very excited, as I'm sure many of you are by the latest developments in AI as well. These advances are creating even more opportunities for us to drive efficiency for our clients. 

I wanted to quickly define what we're talking about when we say AI. I know this has been really helpful for me to also learn, but today most people are talking about Generative AI and these are the models that you know from ChatGPT, Bard, Llama and so on.  

In short, generative AI is an artificial intelligence type capable of generating text, images, other media or code. You can use this for things like desk research, concept creation, survey creation or designing your research in general.  

However, recent advances in AI can do even more than just generate information. Perhaps even more importantly, these latest advances provide capabilities for information synthesis. And what synthesizing AI can do is summarize large data sets of information and help you cut through the noise. You can use synthesizing AI for things like insight summaries, data analysis, management and report creation, both Generative and synthesizing AI can be extremely helpful for researchers.  

But the biggest question and the biggest challenge is how do you seamlessly weave this into your workflows where it can provide the most impact? I've spoken to many of our clients, and we've been listening to the industry to understand more about what would be most impactful for their insights organizations. And with that, we understand that we need to really get to the true value of AI by combining three things.  

The first is building AI into an end-to-end platform for AI to reach its true potential in consumer research, it must be integrated end-to-end in a workflow that researchers have direct and constant access to. You should not need to copy and paste across multiple platforms or worry about drafting the correct prompt.  

The second is blending AI alongside advanced research methods. And as said earlier, speed to insights must not sacrifice on quality of insights. And its only advanced research methods like conjoint and turf, implicit association tests and more that can provide robust data that brands and executives can truly trust.  

Therefore, the question for us is how can AI make advanced methods even easier and faster to use? And one of the things I hear most working with clients is, how can AI make research leaders and those newer to research even more comfortable and confident leveraging these types of advanced methods that maybe they typically weren't running themselves?  

So, the way that we view the future of AI in our industry is through the third pillar, which is AI copilots that navigate you through each stage of the research process. And that is why today we are so incredibly proud to introduce you all to our newest partner in research here at quantilope.  

I will introduce you to them now.  

[A video is played to introduce Quinn, an AI co-pilot.]  

Katelyn Ferrara: 

Okay, so now you have met Quinn, our newest partner in research here at quantilope, and I hope that got everybody excited. We are so excited to bring Quinn to all of you. Quinn is the most advanced AI copilot in consumer research.  

And with Quinn, we are introducing an ever-expanding list of market leading AI features to drive efficiency and advanced method consumer research built directly into quantilope’s end-to-end consumer intelligence platform.  

Quinn leverages both Generative and synthesizing AI to collaborate with researchers by providing actionable recommendations and analysis for developing advanced method surveys, analyzing data and reporting. Quinn is already available for quantilope clients to use on our platform.  

With that I am so excited to hand it over to Jannik to really show you Quinn in real time.  

Jannik Meyners: 

Thanks Katelyn. And hi also from my side to everyone. I'm super happy and proud about the opportunity to show you some of Quinn's capabilities in a live demo today. But before jumping to the demo, we'd like to take one step back to what a copilot for consumer research means.  

So, what you see here is the standard quantitative research workflow that we all know. And by the way, this also happens to match the workflow within quantilope’s end-to-end platform that you're going to see in a minute. And here you go from building your survey field, your project, analyzing your data and uncovering insights and building your report within one single platform.  

How does a copilot support such a workflow in general? Every so-called ‘copilot’ is not simply one large AI model solving every problem we have, but it's a smart composition of a lot of small integrated assistance and modules like you know it from modern cast for example.  

And if this integration and orchestration of these small assistants are done smartly, this makes a co-pilot helpful and our task easier, and it's seamless, and we have this pleasant experience.  

This is also how we designed Quinn. By designing a set of integrated assistance all over the workflow, supporting you along all the stages of your research process. 

I would say enough of PowerPoint slides. Let's experience some examples of our copilot Quinn in action. For this we're going to jump into our platform.  

What you see here, and as mentioned before in our platform, all the different steps within a research process are covered from survey, managing your project, setting up your survey, sending data to field analyzing data or creating your report and dashboard for your stakeholders.  

The study we're going to look at today is about the major airline industry in the U.S. and specifically what we want to understand are implicit brand associations with six major airlines.  

So, I'm jumping now directly to our survey section, which by just using drag and drop, I can simply build my survey. And it works also the same way for most, more advanced methods like the ones Katelyn showed earlier.  

So, for the sake of time, I've prepared some of the setup already, but I've used one of the advanced methods we offer, which is the multi-implicit association test. This is an applied implicit test that is specifically designed to answer or to analyze brand associations of multiple brands in parallel.  

So, we are having our six airline brands that we decided here. These are called “static elements” in this method. And now of course I also want to analyze the associations that I want to measure these brands against. And these are called “dynamic elements” in this setup. For the sake of time, I can go into very big detail here, but we are of course happy to share more material about them afterwards. 

What you see here is a set of already predefined dynamic elements, and these are general psychological motives that happen to work across all consumer or product categories. So, I just keep them target on and use them here.  

But of course, I also want to add some more brand specific or category specific associations. And now one of the first of the Quinn modules comes into place.  

We all know the situation. We are setting this up survey, we need to define the survey input, and then the worst happens, we stare at a blank sheet of paper, we don't know where to start. And this is where an AI assistant comes into place, and it's handy.  

Let's say I know, okay, I want some functional benefits, I can say just, ‘Hey, Quinn, suggest dynamic elements.’  

And what happens after a few seconds, I get a first set of associations that perfectly match to the category. And because it takes into account the static elements that I defined before, I can also add another set. And by this I'm having a first draft to start with. Of course I can delete some, so legroom is not that important, or I can add a few more for myself that I want to add manually.  

So, this is not meant to force you into specific setup, but it's really meant as a guide and assistance that helps you to have a first draft to work with. 

Let's see, another set. And let's say now this time, typing in a live demo is always the most difficult part. Okay, let's say we want to have emotional benefits. Again, we just click ‘Hey Quinn, suggest dynamic elements.’ And then we have a first draft to work with.  

This kind of assistance can be added for many more methods. Like say we want to have a need-based segmentation and we want to have a starting point to work with. We want to set up a brand tracker and we want to know which associations or category entry points we want to start with. This is the assistance that helps along the way. 

So, if I'm deciding I'm happy with this setup, now we can just send the data to field and now let's pretend or move forward a few hours and pretend that the data collection already happened.  

We're now in our analyze section, which we call the centerpiece of our platform because here all the data flows in automatically and in live mode while the data is being collected.  

So, now I'm equipped with the task of building a report. And this can take some time because I need to find the right visualizations, comment on the report, annotate it and add action titles. And let's say I want to add airlines flown to my report.  

These are the results, and I just click on, ‘Hey, Quinn, suggested chart, title and a description.’ And what I'm getting after a few seconds is an action title that considers the data, context, question and type of question. It also generates a description below the chart guiding me through the actual data. If I'm happy with this right now, I can just add this to my report, ‘say add chart,’ and that's it.  

Obviously, this is not a very complex chart. So, let's say we split this chart by eight ranges here, it already getting a bit trickier because there's a lot of bars on this chart. I need to see where to look first, I don't know where to start.  

And obviously what I'm most interested in are the key insights and what is the most important finding that I see in this chart and the most important takeaway. Quinn here, also after a few seconds, created a nice action title for us that we can move to the report as is of course later in the report we can edit it if we're not happy and if we want to make some adjustments.  

And also, below this chart, there's a long or a description, a few sentences that basically guides me through how to read the chart and also what key insights I'm finding here.  

Now you can already get a sense that this can save a lot of time, but the benefit gets even bigger when we use more advanced methods like let's say the MIA, multi-implicit association test, that we just set up. This one is empty because it's the one that I just set up.  

Let's take the other one. And here we see the associations of all brands. So, I click the index view or these associations that we tested earlier. This chart looks quite crowded because it controlled for brand size and there's a lot to see in here. And again, I don't know where to start. And then again here, just a click of a button. 

I don't know. As Katelyn said earlier, you don't need to know how to prompt. You don't need to know what to insert. You just click the button and everything is automated in the backend so that the optimal prompt, the optimal data structure, everything is optimized for this specific use case in the backend.  

So again, I'm getting a glance of what are the most important findings and I can just add it to my report as I want. And let's take one last example, this perception map of brand associations. I don't know if you've ever worked with perception maps based on correspondence analysis. This can show how to interpret this and it can be somewhat tricky. And here knowing that there are some more complicated analysis, like a correspondence analysis behind the scenes, also the AI in this case gives us a lot of guidance on how to read it, how to correctly interpret it, and where the most interesting findings are. So, the text here is a bit longer, but it gives many recommendations on what to read, where the key insights are and how to interpret it correctly. 

This kind of method support is implemented for all other advanced methods, be it MaxDiff, turf key drive analysis or conjoint. And this makes the application of advanced methods so much easier and so much more applicable for people who don't have much experience with a specific type of method.  

Moving on to the report section here, you now collected all the different charts that you found earlier with your action titles. I said also in the setup, this is not meant to force anyone into a specific setup or specific framing because I can of course add everything, edit everything as I want, I can edit the text here before I put it or send it out to my stakeholders or clients. But with this support, within a couple of minutes, I can build a very first version that I can send out because everything is correctly interpreted and commented on as I want.  

Now again, moving a bit forward in time. I've used all these findings to build this beautiful dashboard for sending it out to my clients or our colleagues with all the information about which methods that were used and what data we had and the key findings that I put into the report before and the perception map that I had before.  

Now I need to summarize everything again because I need to write an executive summary. 

Again, I just click on, ‘Hey Quinn, write a summary for me.’ And it creates a dashboard summary of the key insights and that I can use, and copy paste for sending out an email to share this dashboard or for just putting it on top of a dashboard as my key insights. And as we know it sometimes can be quite hard to zoom out after we spend so much time with a specific survey setup that this really helps to focus on what's most important. And some nice tweak here is that it adapts the language tone, the style of the color and the language style to the rest of the dashboard so that I don't have to worry that it's completely off to what I showed or what I wrote earlier in the dashboard.  

And of course, I can edit it as I want. This was a very quick run through our platform to see how small, very well integrated assistance can make the process much, much easier without being an AI expert, without knowing how to prompt, but it's just clicking the button and the use case is optimized in the backend. 

Let's go back to our presentation, and we like to end this presentation by addressing two main questions we often hear.  

The first one is about data privacy. And data privacy was and still is and remains an important topic for us as a company regardless of AI. So, everything we do at quantilope is compliant with ISO standards. And if you're not familiar with these standards or these acronyms, this is an ISO standard that covers all topics around confidentiality, integrity and security and availability of the data.  

It might sound cryptic but it is a very important thing because this is what also the big companies like AWS or Microsoft Azure are actually using to keep the data safe.  

Also, we are not new to the enterprise business. In fact, we actually have been serving multinational companies for years. And when it comes to AI, we want to choose, we chose the same standards that we have and these are very high. And we also use Microsoft as a compliant and trustworthy partner for integrating Quinn into our platform. 

Last but not least, and this is very important, this is a question that we always receive. All data and requests that are used with Quinn are not used to improve or train any foundational models so that no one needs to be afraid that the data is being leaked somewhere or found in any other model version in the future.  

So a second question that we often hear is with all these new advancement, what does this actually mean for you as researchers?  

First, Quinn is a new research support system that makes your process quicker and easier, but it doesn't replace any researcher. And in contrast, it just makes you more efficient, it makes your jobs easier and even more fun because you can focus on the creative part.  

Then as shown earlier, one of the big features is that it's blended with a substance of advanced methods. It's not a standalone feature or tool that I can use, you just use outside your regular. But with this integration and blending with advanced methods, there's so much value behind this. And then you don't need to be an AI expert. But the beauty of AI is that it's designed to work automatically and so that it's weaved into your existing workflows. At some point you probably won't remember anymore where AI was and where it wasn't because it's just a natural part of your workflows.  

With that, we hopefully gave, or we are happy to give you a demo of how you can double your research productivity with these nice features and with these blending of AI and Quinn with our advanced methods. And this helps basically to get one step closer to finally solving the research dilemma and increasing your productivity so then we don't have to say no to any project anymore.  

And with that, I'm handing it back to you Kaylin.  

Katelyn Ferrara: 

Awesome. Thank you so much, Jannik, that was fantastic. It was great to see Quinn in real life and get some exposure there.  

I am really looking forward to jumping into our fireside chat. We'll be joined also by Sara Martens, brand insights manager with Mutual of Omaha.  

Sara, are you there with us?  

Sara Martens: 

I am. How are you guys today?

Katelyn Ferrara: 

We're good. Thank you so much. Yeah, I'm so excited to jump into this fireside chat and just for everybody who's also joining us and listening.  

I know I've seen some questions pop up in the Q&A already, but we're going to try to leave a few minutes at the end for some questions. So, if you haven't already asked anything or if anything comes to mind, feel free to put them into the chat. 

As I mentioned, joining Jannik and I today is Sara Martens, who is brand insights manager with Mutual of Omaha and longtime partner of ours. Sarah, I would love for you to just give a little bit of your background, introduce yourself. 

Sara Martens:

Sure, thanks, Katelyn. So yeah, I've been at Mutual now for about seven years. This is my second time around, so such a great company. I had to come back, but in between I had the chance to work in an ad agency as a strategic partner with some of our clients, and then moved from there to the vendor side. So, I worked for a research provider for a number of years as well. 

I started this journey as a journalist. So, I guess my interest in asking questions and having people answer them goes way back. But this has been a great partnership with quantilope, as you said, for years, and many of our teammates in the company Insights teammates use the product regularly. 

Katelyn Ferrara: 

Awesome. Amazing. Yeah, curiosity and it is always good to hear stories of people who head back and love being where they are and with their colleagues.  

Yeah, I mean, Sarah, I'll start with you and I would like to start off with a question. So, we just went through this overview on quantilope, but mostly AI. When you think about AI from your experience, researcher and vendor side across the industry, what excites you most about the potential that AI has within your processes in your organization?  

Sara Martens: 

Yeah, I think a lot about that question, Katelyn, and I'm convinced that my answer today and my answer in gosh, a year or two years are going to be radically different because as we continue to learn and evolve. So, I think any technology is going to grow and be interesting.  

Obviously, the repetitive task was the first place that my head went. So, I had this bad flashback to when I used to hand code verbatims. And the fact that I don't have to do those things anymore, that I could rely on AI to bring me insights from those in a fraction of the time, was exciting. Moving on to other kinds of more manual tasks that Quinn would be great at as opposed to me eating up a ton of my time. Quinn's way faster than I am then in doing these things. So, I found that as my first kind of excitement.  

Katelyn Ferrara: 

Yeah, awesome. Yeah, no, totally agree. I also mean even two years ago we wouldn't have seen AI as a prerequisite on RFPs and things like that. We will only see how that will change over the next couple of years as well.  

Jannik, I'll ask you the same question. What excites you most about the potential that AI has?  

Jannik Meyners: 

Well, first of all, I echo a lot of the things that Sara just said. I mean, seeing technology that is faster than I am in doing some stuff around doing a task that is pretty exciting.  

But in general, to me, the way AI can now be used in the insights process feels a bit like, it's called the third digital revolution in market research. In the 2000s we saw the rise of survey platforms and online surveys in general. Then in 2010s this was shaped by a giant leap in automating more technical processes like data analysis, data calculation, all these things. Now we're moving on to the more cognitive context-based automation. And this is being in the middle of this, I would say third wave of market research right now. I find it super exciting.  

As also said earlier, I think it has a potential to blend with advanced methods. And if we don't forget what makes our research good, which are the methods, this makes it way more accessible to a larger group and relevant because it's fast, but you still have the foundation of good methods behind it. Yeah, I find it, I mean, it's very exciting times, certainly. 

I'm also with Sara, let's see what we say in two years from now and where we stand because this is just so fast, and this is the most exciting part of it.  

Katelyn Ferrara: 

Yeah, no, for sure. And yeah, I think that's what you hear a lot of times people don't know how to approach certain advanced methods or maybe people are newer to market research, they have teams who are learning. And I think that accessibility is a huge component for sure, and something that I hear all the time.  

Jannik, on that note there is also on the inverse, there is some trepidation around AI, some skepticism. What misconceptions do you feel like people have from an AI perspective? I guess mostly as it pertains to market research right now?  

Jannik Meyners: 

I mean, certainly there's a lot of pushback or misconceptions. I actually don't think that the misconceptions in market research are that much different from misconceptions about AI in general because I mean, what we hear most often and what we also address in the presentation is “it takes away my job,” and I am a hundred percent sure it won't, and it won't do our job.  

Even if I want it to, it doesn't do my job, it can help me do my job, it makes it easier. And probably very few people are into repetitive and manual tasks. So, it takes away the stuff that I don't want to do and it makes my job cooler because I can work on the creative deep-thinking part, but it doesn't do my job.  

So, I feel like sometimes it's a bit on the two poles. Either it does everything or I'm not using it at all. And I think it's certainly in the middle and no, even if I want it to, it doesn't do my job, but it certainly makes it better. 

Katelyn Ferrara: 

Right. No, I think that's a good point. And I think the storytelling aspect, I mean, that's what I hear from clients or even prospects when I'm at different conferences is there's an element also about platforms and DIY of, oh, I don't know, that's a lot of time for me to be spending on X. I want to be a storyteller. I want to be an insights leader for the organization. How can we get you spending more time on being strategic, but having more of that, solving that research dilemma and the insights dilemma and things like that too. So awesome. But thank you so much for that.  

And Sara, we just saw the demo of Quinn, and I know that you have had some direct experience with Quinn and have used it over the past two months or so since we rolled it out to clients.  

What were your initial thoughts? How has it been able to either impact what you're doing or be leveraged over the past couple of months so far?  

Sara Martens: 

Yeah, so obviously first things first, I was just super curious how is this going to work and what's it going to do for me?  

Katelyn, earlier you mentioned skepticism about AI. I'll have to admit to being a huge skeptic. The first thing I did with Quinn was I went into a project I'd already analyzed and reported, and I was like, I wonder what Quinn would say. I went back into that same project and used Quinn to generate my summary, generate my headlines, all of those kinds of things.  

I guess first I was relieved to find out that Quinn wasn't a ton smarter than me, but at the same time Quinn was pretty much spot on and was coming to the same conclusions that I was coming to. And that first little pass helped give me confidence and say, ‘okay, this tool is synced in with how I would view and say about this data.’  

Now obviously you're going to tweak it because here's the way your company says these things or here's what's more appropriate for your audience. And you have that flexibility. Obviously you're not locked into what Quinn told you when you have to use it, but just that first pass through of the confidence of this is really kind of in lockstep with what I would do was really reassuring.

Since then, I've done some additional projects where I've taken Quinn more to the beginning of the project as opposed to the analysis backend, which has been super helpful as well.  

Katelyn Ferrara: 

Awesome. Yeah, I mean, I think also what you touched on is the validation. I think what's great is to know that it's along those lines and you can edit it. You're not locked in. It's not like suddenly Quinn's shipping your entire insights dashboard and you're taken out of the equation. And it alleviates, I think a lot of what I hear and how I've even used AI in ways from my even personal side is fear of the blank page. It just gets you started, and there's something to react to, and then you can edit it and craft it from there. And so yeah, that's really fantastic. 

Jannik, you've already talked a little bit about this, but what's your pitch on why researchers should adopt AI?  

Jannik Meyners: 

I would say there's nothing to lose, but only benefits because if there's a technology that makes my job easier, better or faster, and I can use it when I want it, if I don't want to use it, I don't have to use it, but if I'm trying it out and see that it helps me, it's like, why wouldn't I use it, basically?  

So, it's like doing math with pencil and paper instead of using a computer. I would say AI moves in the direction of becoming a fundamental technology that is weaved everywhere. And so, it helps us, and I'm not sure if I'm quoting the correct person now, but I think Steve Jobs at some point said, “well, computers are like a bicycle, but just for your brain,” and this is what we re experiencing right now. If a tool that makes my job easier and I don’t much to lose or not many downsides, it's like a no brainer. But I'm of course also biased.  

Katelyn Ferrara: 

There's a little bit of bias there for sure. But yeah, no, I think that's right. I, yeah, I agree. And I think it is just so interesting.  

Yes, it's taken on this buzzword and AI has taken on this concept over the past year or so, but again, machine learning and AI and these concepts are not new and the ways we're trying to use technology to automate it, there's all of this new talk about it and new technology that we can also overlay.  

But again, like you said, Sara, you have been trying to avoid doing coded open ends for a while now, so why would you go back to that just out of fear of it? So, I think that makes sense. 

Jannik, is there anything that you see in AI that's a step too far right now, too far reaching? I don't know. That's probably not quite the direction at this point 

Jannik Meyners: 

With the pace we're seeing currently, I am certainly hesitant to make predictions.What we experienced in two years because everything we saw maybe today seemed impossible only two years ago.  

But specifically in market research on consumer insights, take synthetic panels or synthetic respondents as an example. Honestly, I think that's a great potential. And I think it's a great thing for the future, but there're still a lot of too many open questions to use it at scale.  

So, how well does it work across industries and in what industries does it work well in? Which use cases, does it work? Does it work in all countries equally? I think it's a great path, but we're not there yet. How fast we get there can be faster than we think. But right now, there's still a lot of open questions we need to figure out. 

There's also a very interesting thing I think, for our industry that we need to figure out when we say, for example, what is the quality of synthetic panels? Is it correlation with human respondents? Or maybe do we get to a point where artificial respondents make predictions for actual business outcomes even better? So, it's super fascinating.

Katelyn Ferrara: 

Yeah. Yeah. And I mean, you even mentioned there are some, there's obviously international components. I know there's conversations around AI and representation in general, and how do we make sure that that is something that is taken into account in all of this from an insights perspective longer term?  

Sara, I guess what excites you about the potential that AI has within your process and where do you see it? You talked a little bit about this already and what you're starting to do now and how you're using it, but really the bigger impact for you and your teams and the industry in general.  

Sara Martens: 

I kind of want to rename the acronym to be assisted intelligence. This idea that it's a tool for me just like a lot of the other tools that I have in my toolbox as a researcher. We've talked about; I don't have a blank slide anymore; I have a place to start. I would love Jannik, you showed the assisted or the AI impact on implicit association. Think about a starting researcher who isn't even really sure what that would be and is that the right method backup from in my report to how do I even choose the right method and how can I tell AI about my use case and it can help me make those decisions.  

And then just the idea that with that time, we talk about time and efficiency, but those two words to me kind of miss the biggest benefit, which is the ability for me to really be the insights professional, not simply to report the data, but to have the ability to churn that in my head and find that I'm always after what's that little golden nugget of stuff within this data that's really going to help me make an impact on my organization.  

And if AI takes away that beginning part that used to eat up a ton of my time and I have more ability to do that, I think that's kind of the holy grail of this thing. 

Katelyn Ferrara:  

Yeah, yeah. No, I totally hear that. And I think that kind of gets to my last question for you, Sara, which is if you could wave a magic wand, I mean, I'm already jotting down on the side a couple things that you mentioned from a potential product roadmap perspective and as I would hope Jannik is, but is there anything we haven't talked about that you wish AI could be kind of supporting you on?  

Sara Martens: 

Oh my gosh. I think I go back to that first comment about when I asked my peers, what would you say about AI? And the first comment is always, ‘oh, it's going to take our jobs.’ It's just going to make our jobs so much more interesting. I think it's going to give us the ability to be the professional experts that we've always wanted to be, when we've had to spend a lot of time on other things. 

Jannik, you have a roadmap, this idea of synthetic respondents and I'm more skeptical than you. So, I still think we need humans, and I think those humans are going to be there and they're going to be as confusing and contradictory as they've always been. But if AI can help us scale over that and find a way to the true insights of that, and I don't know what that looks like for sure in the future, but that would be where I would hope it would help me get.  

Katelyn Ferrara: 

Yeah. Yeah. I don't see humans getting less complex and less confusing for sure. Not at all in the next few years, if anything. Awesome.  

And I think that's a great segue before we wrap up and head to some questions, but Jannik, what I really want to know, what Sara really wants to know, and others is what's next for Quinn? 

Sara Martens:


Jannik Meyners: 

Sure. I mean, could talk for hours now where we see potential integrated and synthetic respondent is not the first job. No, it's one of the most important aspects we currently focus on. 

We're working on facilitating more insights finding and reporting. So, from weaving Quinn and the AI more into the dashboard and report functionalities, updating automatically. Think of tracking studies where you update the report automatically or even one step further, find the key insights, not scroll through a lot of charts and click yourself through your findings and where to start, but get ahigh-level summarizing insights already by the AI.  

And we are experimenting with how much do we build a dashboard from scratch? What's the interaction level between users and Quinn in building a dashboard together?  

As you said, it's assistant intelligence, it's not see it as here's a button and everything is done and it did my job as we said, it doesn't do my job. So, finding this dance between user and Quinn, especially in reporting and dashboarding is our current project.  

And my opinion there is still so much potential, but you still need the expertise, you need the business context, and I'm with you. We need humans with all the contradictions and mines and yeah, that's next for Quinn.  

Katelyn Ferrara:


Sara Martens: 

But if Quinn built those pesky PowerPoint charts for me, I would love it even more. 

Jannik Meyners: 

Not it down. 

Katelyn Ferrara: 


You can at least export it all to PowerPoint now.  

Awesome. Well, thank you both for this amazing discussion. Sarah and Jannik, I think we have a couple minutes. I know we're running late, but I think we can probably go a couple minutes over for some audience questions.  

So I'll pass it back to Joe. But thank you all and if you'll be at Quirk’s Dallas or Quirk’s Chicago later this quarter, please be sure to stop by. Come meet Quinn, come meet myself and our team. So, we will be at those events. But Joe, over to you.