AI expertise for consumer insights
Welcome to AigoraCast. Conversations with industry experts on how new technologies are impacting sensory and consumer science.
Visit your favorite platform and remember to subscribe:
Apple Podcasts
Spotify
Please leave a positive review if you like what you hear!
Host: Dr. John Ennis, CEO at Aigora
Guest: Roberto Salas, Director of Consumer Insights & Data Intelligence at IFF
In this episode of Aigoracast, host Dr. John Ennis sits down with renowned consumer insights expert Roberto Salas. Roberto shares his fascinating career journey, spanning multiple continents, and discusses how he bridges technical sensory science with deep consumer empathy.
Roberto Salas is a Consumer Intelligence and Growth Strategy executive with over two decades of experience in the food and beverage industry. Operating in the Global Consumer & Sensory space, he specializes in translating complex market, consumer, sensory, and behavioral data into strategies that fuel portfolio growth, innovation, and brand differentiation.
Throughout his career, Roberto has managed global and regional projects for major corporations, including Danone, ConAgra, Cargill, Yoplait, and currently IFF. His expertise lies in bridging technical product design with a deep, empathetic understanding of human needs. Over recent years, he has pioneered the use of AI-powered tools to mine unstructured data for actionable intelligence. A passionate advocate for democratizing research, he frequently conducts workshops to coach cross-functional teams, ensuring that genuine consumer value remains at the forefront of the commercialization process.
Roberto holds a degree in Food Engineering from Tecnológico de Monterrey (TEC) in Mexico and a Master’s Degree in Food Science, specializing in Sensory Analysis, from Kansas State University.
John:
Okay, welcome back everyone to another episode of Aigoracast. Today, I'm very happy to have my friend, Roberto Salas, on the show.
Roberto Salas is a consumer intelligence and growth strategy executive with over two decades of experience in the food and beverage industry. Operating in the global consumer and sensory space, he specializes in translating complex market, consumer, sensory, and behavioral data into strategies that fuel portfolio growth, innovation, and brand differentiation.
Throughout his career, Roberto has managed global and regional projects for major corporations, including Danone, Conagra, Cargill, Yoplay, and IFF. His expertise lies in bridging technical product design with a deep, empathetic understanding of human needs. Over recent years, he has also pioneered the use of AI-powered tools to mine unstructured data for actionable intelligence. A passionate advocate for democratizing research, Roberto frequently conducts workshops to coach cross-functional teams, ensuring that genuine consumer value remains at the forefront of the commercialization process.
So, Roberto, welcome to the show!
Roberto:
Hi! How are you, John? Nice to be here, really happy.
John:
Yeah, it's great. We've known each other for a long time, and it's always good to see you. But I think, you know, for our guests, can you take us through the journey of your career? I know we have a shared interest in AI, but maybe you can talk about your journey to how you got to where you are today.
Roberto:
Yeah, well, it's not going to take— yeah, it's not going to take me 20 years, but just in a— in a nutshell... [laughs] In a nutshell, um, I'm a food sci— I'm a food scientist. I studied food engineering in Mexico, and then I did my master's in Kansas, at Kansas State. And there's where I started learning and started, you know, utilizing sensory widely.
And then after that, I went back briefly to Mexico and worked for Yoplay in the marketing department, but then I moved to different countries, including the UK, South Africa, and then the US again. And yeah, I've been very lucky to collaborate with a lot of people, you know, in this journey of consumer insights and sensory insights, and trying to merge, you know, those two sources of information in a more actionable and commercially efficient way.
Because there's— there's the science on the one side, but that gives you the tools to approach the business problems, right? So, I think just having that sense that what you do is to add value where you are, I think is key for this practice and for the people that are working in consumer and sensory insights.
So, yeah, during the last 10 years, I worked here in the US for IFF in the global sensory and consumer insights, and now it became the consumer intelligence team. And yeah, very, very happy to be talking to you right now.
And yeah, the last few years, it's been crazy with AI and the different ways that we've been trying to work and do the most with it. It's amazing. It can go in different ways.
John:
I totally agree. Well, let's come back to AI in a second. But something I do want to ask you about, because whenever I talk to somebody who's been at multiple companies—and especially in multiple countries, because you've worked in many different countries, so you have experience across the globe—it's always interesting to hear their thoughts on how research differs.
The divide I'm most familiar with is the European-American divide, but I'd be curious about Asian, South African, and just what you've seen regarding the difference in the consumers, and also the difference in the perspective toward research. I think that there's definitely biases in how the research happens in these different areas, and it'd be interesting just to hear your thoughts on what you've seen over the years.
Roberto:
Yeah, no, I mean, of course, I think again, the principle of the approach might be the same in a way across different countries. But you have to really be very aware and empathetic on the consumer from the region. It goes a little bit beyond just— because I hear a lot about how people use scales and all that, and that's all right, you are in that context and then you can work it out. But it's beyond that. It's how is the mood in the country? What are the people going through?
When I was in South Africa, it was very interesting because they had just abolished apartheid, and there was a lot of merging of different cultures into society. So, silly me, a naive me, I came from Mexico. In South Africa, they are very high consumers of maize (of corn), and we are too in Mexico. So, I had this brilliant idea. We have something that we call—most probably a lot of people know—we have tamales, which is basically a corn patty wrapped in corn husk. And you know, people love it, and I was like, "Oh no, I'm going to be a millionaire because I can adapt this to this country!"
So, I did my own focus group, and then I was presenting the samples, and the people were like, totally, totally against it, from many different aspects. It was crazy. Not just the taste—because the corn that we use goes through a process that uses lime—but beyond that, it was the presentation. The wrap was a corn husk.
They were like, "I mean, we became a freely, industrialized, modern society, and you're coming to us presenting us a product that is wrapped in corn husk? What's all that about?" So they were offended. They were actually offended.
John:
Oh, wow.
Roberto:
Yeah! So I apologized, and I was trying to find out more. But it was great learning. It's about the context that the country is in.
In a latest situation I had in Indonesia, there was even more about local taste and localization of products, even though perhaps the product was an American product, an American brand. But they were looking for localized flavors because there's a strong feeling of nationalism right now happening in the world. So, yeah, no, I think you have to be very aware of the human insights and how the society is in each one of those countries, so you can also understand how people are going to be willing or not willing to answer different questions.
For instance, in Mexico, it became really difficult to do consumer research because people were very afraid that you were going to use that information to scam them, right? So, the situation that is right now dangerous over there is making people believe that that can happen.
So, no, it's always very interesting. I think you have to really understand the country, you have to really understand where they come from to do a lot of immersion in the society, and be very empathetic about their situation. And then from there, then you have the tools, right? Then you have the tools, you know how to work them out. But as far as you have that context, I think the rest is very straightforward.
People are people everywhere, and they're very willing to help once you establish some rapport, some understanding, either within the companies or with the consumers as well.
John:
Yeah, interesting. Yeah, I'd be curious, honestly, because we talked a lot about AI before the show, and the promise that AI can help with cross-cultural communication and predictions and that kind of thing. But, you know, it would be interesting to take a situation—so, part of the problem with AI is there's always a temporal lag, right? There's the training data, and then you've got some amount of time after the model has finished training where things can happen.
But I wonder, I'd be curious to see if AI could have predicted that South African consumers would be offended by corn husk or not. That would be interesting to know. My guess, yeah, well... [laughs] Go ahead, I have more to say, but I'm happy to hear what you think.
Roberto:
Yeah, no, I think the amazing thing about AI is that it can tap into multiple sources, right? So, there's a lot of open information out there. There's a lot of data that can be connected, so you can connect the dots even before you organize anything. Even before you start thinking about a certain thing, it organizes your ideas and it can help you, you know, covering those gaps that perhaps you're not aware of.
What I find really amazing is that you can really plan—not even what you want to do, but plan the plan. I don't know if I can explain myself. You know, how to plan the plan. What are the things that you don't have to oversee? Where can you look for information, or how could you ask for information from AI as well? So you can have a really robust plan about what you're about to do.
And that, I think, for me, is one of the things that has helped me a lot: to organize my ideas, to be really like a springboard for planning, and for executing as well. So in the way we all think, I think we have some preference to be more like "jump to action" or be very hands-on, or be very imaginative and visionary. But I think we basically have at our hands multiple minds in AI. We have a whole team that can help us organizing our plans, organizing our ideas, and organizing our execution in a very efficient way. And I find that very, very useful.
And I think if I had AI at that point in South Africa, of course, it was going to let me know, "Just be careful because people are very edgy, very sensitive. You have to be aware of these different points," instead of just running into it and then finding out after paying for focus groups. It could have prevented me.
John:
Right, it might have. That's the thing I'd be curious about: AI is good at warning you of potential problems, but I still think that even if AI says, "I don't see anything wrong here," it may not be right. And so I would be curious. Let's do a little experiment right after the show, let's you and I see what it says, and see if—well, maybe enough time has passed—but it's interesting.
John:
Now, Roberto, you just went through the AI course, and you were part of the first cohort on there. And I think you were one of the star students, in my opinion; you had a lot to contribute to the group. And I was wondering, what are you seeing now? What are the things that you're finding most helpful? What do you see as most useful when it comes to using AI in our field, and what do you see as maybe overrated and, you know, there's a lot of fuss about it, but in practice, it's not actually that helpful? What has been your experience?
Roberto:
Yeah, well, my experience—and this is from a very personal experience on using AI and how I see it applying—is that when you, as a consumer or sensory intelligence person, are out there trying to communicate your findings and trying to democratize the information that you have found, it has been very hard to do two different things: either go one-by-one to the stakeholders, or have like a storyline, like a really nice storytelling narrative of the findings, and how can you apply those findings to solve not only the questions that you have from the business, but perhaps other things that you were not aware of.
So, what I find is that AI can help you putting together and also customizing the communication out there into different layers. Because you're going to have the people that are going to be very interested in the details, right? And then you're going to have people that are going to be very interested in the "now what?" and "what does that mean, and what are we going to be doing with the information?" So you have different levels and different styles of communication that AI can help you to really convey that information and convey those ideas and those insights beyond the numbers, right?
Saying that something is 60/40 and then you reach the target, I think is short-sighted. Understanding perhaps what's behind and the reasons why you reached that point, and having some compelling story using AI to help you either creating nice infographics or nice and very quick videos.
And even beyond that, I was in talks with another company to do an AI persona with the information about different segments. So you could have a persona from a specific segment, and the stakeholders can go back and say, "Oh, you know, hey Max, in your segment, what do you think about this idea, or about this flavor, or about this concept?" and it can actually respond based on the information that you have fed before, right?
So in that way, you can have all the data, and then you can make it very actionable, instead of that stakeholder coming back to you and asking, "Hey, what do you think about this?" and then you have to go back and mine the data, make another presentation, and then give the answer perhaps two weeks after you got the question, right? So the stakeholder now can just ask the question directly to AI based on the nice information, the data that you already have fed into the system.
So, I think communication and really having a really good way of conveying all the knowledge to the stakeholders is a really powerful thing that AI can do. But people have to be very aware that it's not just "press a button," that there is some really diligent feeding on the information, but also a very diligent way of asking, you know? You just don't ask very open-ended questions, you have to also learn how to ask.
I had a teacher a long time ago who used to tell us, "If you don't know, don't ask." So that was very like... counterintuitive. But what he meant eventually is like, okay, if you don't know what to ask, don't ask, right? So you have to know what to ask, you have to know how to ask. There's a lot of richness in learning how to ask to get the right answer.
John:
Right, no, I think you're a very perceptive person, that's a quality I think you have. You're a very sensitive, perceptive person, and I think that's one of the things that makes you particularly good at your job. And I think that AI really needs that, because I think AI is very useful and can automate lots of stuff, but I still feel like there's this human interface, there's this little layer where the AI can kind of get it really close, but the human has to be there to take it across the line.
And it's the same way with the planning, where—and this is where, you know, people say, "What's most overrated with AI?" I think it's the idea that the AI is just going to replace people's jobs. I know that Anthropic is all about that, but I don't really see that happening.
So, how do you see—maybe you can elaborate a little bit more—in the next few years, how do you see AI unfolding? What would a typical day in the life of a consumer researcher be, using AI, kind of building on some of the thoughts you've already shared here?
Roberto:
Yeah, the way I see it, and I agree with you that having the human in the loop is always going to be there, because at the end, you know what the real issue is sometimes, and you also know how that data is going to be used and what's going to be the meaning of that data. And I feel that for AI—and people talk a lot now about the next level of AI, of becoming even more self-sufficient, we're still not there—but how AI helps you is basically to organize that information that you have, to organize the information that perhaps you don't have, into some more compelling way. And then it's your choice and your decision to take that risk or not to take that risk, or to say, "Listen, we don't have enough information, or the information that we have is very generic, it's not unique, and I don't think we can go ahead and use that information. We need to do some other research or we need to feed the system with some other data."
So, I think humans should develop this empathy to the business, and empathy to people's work—you know, to all the developers or the commercial team or the marketing team—because in that way, you can be more aware how to use AI and what type of information you should have backing up that information.
In the way I have used AI, it is almost like to bring people more together. So, for instance, we had a workshop on understanding the future of a certain category, and then I joined people from different parts of the world into one workshop, and then we used AI to synthesize the information that they were giving, and then also to finalize the information that we found out in specific prototypes. So, all that became really good because the people were feeling that their ideas and their input was being taken into consideration. So, it actually was more compelling and also at the same time more personal to them.
So, I think, you know, it helped to create that unity. So that's why I think, you know, without that people, without the input of those people, just doing the workshop with AI, I think I would have fallen short and just restating perhaps the obvious that is already out there in articles and websites and podcasts.
John:
You have to have that human interaction.
Roberto:
So you have to have that human interaction, and that creates all this richness that can then feed into AI.
John:
Yeah, that's right. There's a kind of sensitivity, too. You know, I'm actually reading this book—it's a great book actually, if you've read it—The Creative Act by Rick Rubin. Do you know this book?
Roberto:
No, I haven't seen it.
John:
Yeah, it's quite a good book. Rick Rubin is a music producer, and he's produced all styles of music. But he talks a lot about sensitivity, and how the job of the artist is to really cultivate sensitivity. And I think that's very in line with what still is needed with AI, is that sensitivity. You've talked about empathy, but it's two sides of the same coin. That's what we can bring that is special.
Roberto:
Yeah, and also not very sure at this point—and, you know, like AI is becoming like from 10 miles per hour to a thousand miles per hour—but I'm not sure how creative, in the sense of creativity, AI can be. I also feel—and this is very controversial because that's one topic I always am in some discussion with some people—so AI is fed with things that are there, right? That are available, that some other people did or some other industries. But if you really ask for something totally new, I find that it just creates fusions or mixes.
In one workshop, I asked AI to do a SCAMPER—you know, it's one of the techniques to create some new ideas—and it's very good. I mean, it created really good, amazing ideas. But that's why I feel it is very good: it's very good at giving you some starting points, but I think the creativity still has a long way for AI to be there. I think the real creativity again comes from humans, and AI is the tool that you can use to get there. But if you ask AI to paint as no painter has done ever, I'm not sure what it's going to produce.
John:
Yeah, no, it's really interesting. I just actually finished my first math paper in about 20 years. I was curious about the whole AI math thing that's happening, so I reached out to my former advisor and asked her if she had any math problems that she didn't have time to work on. And so she gave me some problems, and I was able to solve them pretty quickly with AI.
Then she gave me this kind of big problem that her and her colleagues have been working on for years, and through a very extensive process—I mean, using a lot of my experience doing AI coding—I was able to solve the problem. And in the end, the solution came from a big search that I did, where I had all these different agents looking at different sub-branches of math trying to find similar problems that have been solved that we could use the solutions to help us with our problem, okay? So it was a kind of guided search.
And in some sense, this was, you know, you could call it very creative, right? Because, you know, it was using an idea from another area. And human creativity often is like that. But it also was very kind of predictably creative, because it was like, "All right, we're just going to go pattern match all these other subfields, and look, what do you know? Our problem is similar to this problem, this problem's been solved, so we're going to take the same ideas," right?
And so I think you will have creativity like that in the short term, and a lot of problems will get solved. But there is something else that humans have: the ability of like a Van Gogh or a Beethoven to go beyond, to go out of distribution. And that's where—and I think we all do that to a smaller degree, you don't have to be Beethoven to have a new idea. Humans do seem to have some ability to go beyond the distribution that the machines just still don't have, and it's kind of surprising, but I think it's still true.
Roberto:
Yeah, yeah, I think so. I also read some articles about how the cognitive power of generations have changed through time because of technology. Some articles were saying that because of the technology and how people are using technology, the cognitive ability of certain generations is going to decline through time because they won't have this ability of problem-solving without any help of AI.
That is a very interesting view. But I feel like, as you're saying, humans are very plastic. So you solve your problem—you were actually creative in solving the problem, but you were just using the tool to solve the problem. So the creativity still came from you; you still were using the tool to solve the problem. So I think in that same way, I think people are going to develop some other creativity skills, and perhaps it's going to be just powered by AI.
And I feel also, yeah, that sometimes there's the danger of not seeing the outliers, the small segments, the niche, the perhaps the— sometimes the outliers—and you know, because being a mathematician and also in the industry—are the ones that really bother you in your minds, like, "Why is that person there? Why is this group of people there? What does that mean? How can we get them? What is exactly what they— where do they come from? How can we approach them? Why is there a segment that is an outlier?" And with AI, perhaps that might be just washed down by the average of responses or the average of information that it provides.
So, I think it's going to be interesting—interesting thing how people are going to be using AI in 20 years. But right now, I think it's open to a lot of different possible routes. And I think, like you were saying, the best way of doing this is by doing it. If you have some idea, just try it, test it, like you were saying right now. Yeah, I'm also interested in seeing what the South African consumer would say now about the product, but you just have to go there and do it and beta test it and compare it against some other things that you normally do. But if you're not open to these new ways of doing things, I think in the future, someone is going to be open and it's going to be more advantageous to the person. So, rather than just waiting in the background and seeing if it works or doesn't work, just do it. You might be wrong a few times, but still, you've got to learn.
John:
That's right. No, you've got to learn. Well, amazingly, Roberto, we are actually at time, and I want to make sure we have time for you to give final words to the audience. So, yeah, but I love talking to you. And so if somebody wanted to reach out and talk to you, what would be a good way for them to get in touch with you?
Roberto:
Yeah, well, I'm on LinkedIn, so they can reach me over there: Roberto Salas. I believe there's only a few! [laughs] So, yeah, you can reach me on LinkedIn, that's the best way.
John:
Sounds great. And what advice do you have now looking forward through the end of the year? What—you talked a little bit about experimenting with AI, what are the things that are kind of on your mind? What would you recommend that people look at here, you know, with the second half of the year coming up?
Roberto:
Yeah, well, there's one thing I've been seeing more and more, which is experiences. So, AI and other new technology is going to help us understand the real experience of people. You know that if we go and use the normal scales of nine points, and then you get the product out, 90% of the products don't succeed in the market. And I feel like one thing is to like a product, and another thing is to really love the product—like really have a really good experience, a memorable experience, something that you want to go back and repeat, right? But I think the methods that we have had so far are designed to reduce the noise.
But I think in the future, all this technology, AI is going to help us to actually measure the noise, and to really gather that noise and then understand that noise. And that can help us to understand the experience, the memorability of consumption, and how to create—or how to help people to create rituals, habits, and have products that really are more aimed to give them a really good experience. I think one thing is just to give them a product, and another thing is to give them an experience.
And experience is right now, you know—I mean, just look around and people are paying an enormous amount of money to go to concerts, to go to sports events, to have experiences. So I think that's something that we can—if we can understand how to measure the experience of products, that would be one step ahead to have really good use of methods and AI. And again, it's about embracing the noise, trying to do that in the smartest way possible.
John:
Right, right. Basically, you know, we've, for convenience almost, we've sanitized out a lot of the noise so we can have predictable measurements. But the net result is we've lost understanding, and your thesis—which I think is correct—is we have much better ways to collect more nuanced data, we've got better tools for going through it, we should embrace what looks like the messiness, and then use our tools to understand the nuances of the experience. I think that's just right. I agree with that, that's a really good insight.
Okay, Roberto, well it's a real pleasure, thanks a lot for being on the show.
Roberto:
No, it was great to talk to you, John. Thank you.
Thank you for listening to Aigoracast. If you enjoyed this episode, please subscribe and leave us a positive review on iTunes. To learn more about how Aigora is helping businesses leverage artificial intelligence in sensory and consumer science, visit us at www.aigora.ai.
Aigora is a contributor to the Aigora blog, sharing insights on AI-powered sensory science and product development.
