Ludovic Depoortere - A New Decade for Sensory
Welcome to "AigoraCast", conversations with industry experts on how new technologies are transforming sensory and consumer science!
Ludovic is an entrepreneur and keynote speaker on the topics of sensory, marketing, and innovation in research. As chairman and founder of Haystack he supports the world’s leading FMCG, Personal Care, and Healthcare companies with innovations in the area of sensory experience. His passion lays in optimizing consumer experience by using the power of multi-sensory insights to build the connection between emotions and brands.
Ludovic has been active in sensory research for over 20 years. He is a member of the Board at the European Sensory Network, President of Cube (Consumer Understanding Belgium) and Fellow at the KU Leuven University.
Transcript (Semi-automated, forgive typos!)
John: Ludovic, thank you very much for being on the show.
Ludovic: Yeah, you're welcome. My pleasure.
John: Wonderful. Now, Ludovic I think we share many common interests. So there's a lot we could talk about. But I think I'd like to just begin with a general question, which is we just started a new decade. It's now 2020. The world obviously is undergoing tremendous changes in almost every facet of life. And I'd like to see, I'd like to hear what is your view on the role of sensory and consumer science in this new decade now?
Ludovic: Yeah. The one million dollar question. Where are we heading?
John: More than a million dollar.
Ludovic: Yeah. Great question. Thank you. Yeah. I think we live at a very interesting time. I think 2020 will be really a pivotal year in century consumer science. But I think in all industries, a lot of changes are going on. Consumers have other demands than they used to have. Like I think also companies realise that purpose is finally profitable, being sustainable, taking up responsibility. But it also like big brands like the Coca-Cola Company. They are like many of their brands are not so loved anymore as they used to be. So it was lot of change, a lot of uncertainty. And also sensory and consumer science is affected by that. There is a great need for more speeds. It's a rough way of doing research. So I think the sensory and consumer science should really change the way, we do science, we use science so that we can remain relevant because that is a bit what I fear is, if we don't change, if we don't rethink the way we do and how we can impact the world outside, I think the relevance of sensory and consumer science will be reduced. That's absolutely what we want to avoid, because I'm sure that we can play a key role in the disruption that is going on.
John: Yeah. And I would say you and I could be on the same page there when it comes to that. Like for me, I see it as either sensory and consumer science can adopt a host of new techniques and become more relevant than ever, or they can get left behind. But it's really not a middle I don't think just continuing to do what we've been doing is going to turn out well in the long run. Would you agree with that?
Ludovic: Yeah. I totally agree. I see like, I have the pleasure of working like for big corporations for last 20 years. And I see that there is an increased pressure on budgets and also on timings because, you know, the CEO's of these big corporations, they really trust that R&D teams. They can really make great tasting products if you cannot connected to a relevant brand strategy, if you cannot connected to a real strategic angle, like it's just a tactical thing. And we see a reduction of budgets on tactics and an increase in budgets for these corporations for the real strategic issues. And that's where we should connect the dots and not only make sure that sensory and consumer science is there like to create great tasting products, but also like to make sure that we can cope with the big world trends. How can we make great tasting products that are also sustainable? Like how can we make sure that we can make produce products that are as cheap as the now like big fast moving consumer goods, products that are cheap, tasty and also sustainable. So yeah, a lot of challenges on our plate.
John: And what are some of the techniques that you're using in Haystack to help your clients with these challenges?
Ludovic: Well, to start with. We invested a lot in data science because that is one element of overtime. I call it the data bloating. There is so much data available that actually, yeah, we can start using data that is already there. And that is a real difference in the process of how we start doing researches. We used to model would be like, okay, you have a question. We do a survey, we do an experiment, but then we come with results. Today with so much data already available that we can start digging into them. But it's a completely all the way. It's unstructured. The data is everywhere. You can get it on with social media channels. But also, like one thing we do is we start asking our clients, like, what data you're already have and maybe we can already learn a lot from that so that we can come up faster with which solutions. So one rule that we developed is like the data science, the big data, being able to learn more from that. But also like because of these need for agility, speed, everybody wants to be the first in the market to understand the next big thing. We are using much more automation, like the time where we had many consultants here creating charts like copying data from Excel, PowerPoint and then into graphs. Yeah, those days are gone. Like it's automated. So we get faster insights. But also, like we experiment a lot with artificial intelligence. I think the biggest case there is everything that is voice and text processing. So rather than asking people to comment via typing comments in an open end, we ask them, like to record it in voice. It has the advantages that you can have some sentiment analysis linked to that, because the voices of more telling than just the words like we can process from voice to text and then we can do a lot of like natural language processing type of techniques. And yeah, we are there at early stage. But you see that. Yeah. The algorithms become much smarter, faster. And also translation was a big issue. You can also see that the quality of like translating from different languages to English, for instance, but also the quality there is really increasing.
John: I think it's great. I think we're on the same page there, I mean there's so many questions I want to ask you about different directions here. Maybe let's drill down a little bit into the natural language processing, because I think that's really like you're right, we're in the early stages. But if you go back, I mean, have you ever seen a picture of an early car? Like very early cars? You know, they looked like? Two bicycles strapped together with a motor in between, like it was, you know, not anything like the modern car, right? But that's where it gets our early trains, you know, the same sort of thing that the people with vision can see that what's coming right? That it's going to be totally different than it looks like right now. So when it comes to the collection of this text data, so do you have an online survey? And then when you get to the open end, someone could push the button and they record or how does this actually get implemented?
Ludovic: Yeah. So we have an app. We have a platform like we also do a lot in in communities nowadays because that's another thing. It used to be like you do a survey, you ask a question, you get a response. Other big trends where we adopted is that we are much more offering iterative approaches in the digital platform. So rather than participating in a study, you become part of a community for a couple of days or weeks or months. Then one day you can just like send them a question and ask for an open answer. Or you can send products and ask them to give liking score or you can in the same platform. So it's a platform where you jump on a tablet or in the website you can react and indeed you can record a button, they can upload what they want to share or they can do it directly from their mobile phone. And that is the batches of comments are coming in. Also, like we work a lot with video where we do the analysis on the voice recording as part of the video. So all of that is coming in the platform. And then we start analyzing and I like your idea of the bicycle, the bicycle car, because you know, you get a lot of like. But that was also when online research popped up, it was like, yeah, it's not representative, it's the good like it's always better to have an interview asking questions. Now we have the same thing obviously. Of course, it will be algorithms are not so good as the human brain. Of course it's not so good. But if you see what it delivers, it's speed and accuracy and how it can filter. Yeah. I'm really impressed by how good the algorithms are already today. But indeed, we are far from like it's always better to have a conversation Face-To-Face so you can see nonverbal elements. You know, as a researcher, you always want to go for the best. But we will have to accept that, the pace of the world and the pace of the clients, they want to innovate. They want to go fast. They want to learn fast. They are happy with learning something fast, try on something fast. And the quality might be a little bit less compared to the past, but at least it gives them the opportunity to move on. And that is key in today's world.
John: Yeah, I totally agree with that. Chris Findlay on this podcast, who's actually our first podcast guest ever, and he was talking about similar challenges when computerized data collection first came along, that there was this idea that it wouldn't be valid or that, you know, somehow it would not be as good as paper ballots. But of course, you know the world accepted that. I think that this next level where I think what you're really talking about are something like interviews or focus groups at scale that you've got, you know, more like isn't an era. It'll never maybe in 50 years it'll be like a human interviewing somebody. But in the short term, it's the drawbacks are outweighed by the benefits of the speed, like you said, and also the scale that you're able to do this or many more people than you would if you were to send interviewers to people's homes. So let me ask you next about your, so Haystack maintains its own communities. You have your own kind of standing communities that your clients come to you? Or do they have the communities and then you send the platform out into the, how does this typically work for you all?
Ludovic: Well, how this works today is we have worldwide, we have communities available and we are recruiting because, you know, every research, you know, there can be a specific target group. So we have communities to start from. So communities that we own, that we have built and then we have like mixed mode like we can be good people to telephone because not everybody is online and ask them, like, okay, do you have, like, access to a good Internet connection? And then, you know, the platform is so plug and play that it is easy for them to share on the platform what they want to share. So it's very easy. Or we use Facebook, Instagram for recruitment to build a specific community for a certain purpose. The idea is that, you know, you have consumer participation today. So consumers are motivated because they can really feel that they like belong to like an innovation track for a certain company. And like, it's this participation that is also like great. And it has a big advantage also compared to like a classical focus group, for instance, because at client site, everybody can follow the conversations so they can see what consumers are posting, what they are sharing, what they see in the videos. So it's like, oh, that's interesting. I would like to know more about this. So it answers because of its flexibility. It answers the needs of today because one day marketeers wanted to start what type of packaging the consumer would like. And in the next day, the sensory scientist wants to understand what is the flavor direction. I should go to match the concept that we just discussed. So in that way it's much easier to answer all kinds of different questions at a speed, and in that way it increases the relevance. But also it learns marketeers why the taste and the texture and how it looks like or how important it is. And it helps also people in our division to understand like what did need exactly is and what type of language consumers are using. So you can come to a better match of a need and how marketeers and R&D teams are working together to match that need.
John: That's interesting. So what are the kind of data sources then? So you have the data coming in from your app that you collect, which can be in a variety. I guess you've got standard survey data that might be numerical. And then you've got text data. You also video data that you're collecting. And then what are some of the other data sources that you like to leverage in your model?
Ludovic: So the other data sources. So we have access to a database that is actually collecting a lot of what is shared on social media. And it works. Obviously, you have to find words and search terms that you're using to understand the specific topics. For instance, microbiomes, probiotics is like for many, many categories. It is a very relevant research. They want to do an understanding for their category, for their consumers. What does it mean? How can we have probiotics in there? And then we it's kind of social media listening. And then, like, we tap into sources like Twitter, like Facebook, like Instagram, but also blogs, influencers. So it's a huge center. It's a data lake of information. And also there, it's again, the interaction of human capacity where we read the algorithm can actually summarize and find relevant conversations. But it's always like the human brain saying, oh, that's interesting, I want to deep dive in this. So let's look at how we could change the algorithm to really understand what is meant, but also because we collect this data and we can go back two years. We can also understand how. But where did these first probiotics discussion started? And how big is the trend now? How many conversations about this do we have everyday? So you get to and that is the nice thing about it, about working in that way, you get qualitative information at the quantitative skill. So because obviously a good focus group is a great instrument. But here you have such a huge amount of conversations and data that you can go and quantitative and you can go in-depth. So you can have the what and the why at the same moment. So I think that is also like an advantage of these technology that you can you can increase the speed of understanding and commit faster with insights to help clients in their strategy.
John: Right. And is this then your analysts use or do you use support for your clients by providing dashboards through which they can access the information? How does that actually play out in practice for you?
Ludovic: Well, we have analysts at Haystack worldwide, and they will really do an in-depth analysis based on the data they are getting. So we developed a platform that is actually scraping all the data, combining all the data. Also looking into data of the client. And sometimes you build dashboards until the end. The end solution is you have a dashboard and your client can actually look what's going on every day. But more often, the client says, you know, we trust you guys, our researchers, we have questions, we have challenges. So you do a deep dive into that and you come up with what it means for us rather than taking them in-depth to do all of this or be depending on like where you want to go. But like for wide spacing, it is very often used for micro trendspotting to see what would this mean for my industry? And yeah, then you always need, like, the quality of analysis, analyst, and a researcher to make that happen. But yeah, as a matter of speaking, when we looked in signs, we used to have like one or two IT profiles, but now about 20 percent of the people we employ are just developers or specialists in technology. And that is also a big change compared to the past where we just had like SPSS or whatever type of statistics software, and that was the tool. So it's much more complex and much more IT driven than it used to be.
John: Right? Yeah. I mean, it's a trend you see. Almost every big company is becoming a tech company in order to survive. Like Nike's become a tech company that makes shoes, right? Domino's is a tech company that makes pizza. It sounds like, you know, Haystack is a tech company that makes marketing insights that you're moving in that direction. Okay. So then it's actually amazing how time flies. But I have a bunch of questions. So what about helping clients to leverage their internal data? Because most these big companies have a huge amount of data they've already collected. Is it possible then to integrate the data that they have already collected with the data you're collecting to give some sort of comprehensive model?
Ludovic: Yeah. Great idea. It's also what we do like not an easy process because clients, you know, are not so keen on sharing what they have. And obviously also sometimes, you know, they're a bit ashamed of the low quality of data they have. And it's scattered all over the place. You know, that's everywhere so that will be fine.
John: Yeah. I would say if you listen to this and your data is a mess, don't be ashamed, it is the way it is.
Ludovic: So a project like this usually starts with like making sure we can get hold of all the data that we connected. So it's lot of data warehousing in the beginning and then the fun actually starts but the real thing is that I like so much about such really consumer scientists that they are key, they are always have been educated in like working with multiple sources and trying to find the complex things and try to make that easy and understandable. Make that simple. And that's a unique quality we have. And by technology, we have access to that type of data. And I think the key thing there is the word impact. I think that is something that the industry should focus on. It is how can we learn from understanding, seeing a red line in the story. Make sure that we get more impact. I believe we can do better because scientists we love data. We are so proud of, like the charts we create, the insight we have, and look at how fantastic this analysis is. But sometimes we like a bit yeah, marketing that and making sure that it looks nice, simple. And there is a great story build around that. I think if we can develop that skill, that capability better, we will be able to actually show much more value of the capability of being such a good researchers because I see that as a big shift. They don't want to pay us anymore for data collection. I didn't want to pay us like to design an experiment. It's just like, okay, hey, guys, what does it mean? I need meaning. I need you to connect the dots. And I believe if we would go there, because sometimes I'm a bit frustrated that those big five consultants actually, I don't I really don't believe that they are smarter than the average sensory consumer scientist, but they can charge like triple of what we charge. So I'm like, hey, like let show and be brave and and show the value of what we can do. But then we should stop looking at ourselves and we start looking at the data of the client and then actually bring the message in a meaningful storytelling type of way. And then I think we will get more like we were more evaluated for what we do.
John: Right. Now, I totally agree with that. So what would be it when you look at the kind of great successes that Haystack has had? Any themes that like what sorts of things should sensory and consumer scientists be doing to make sure that the work that they're providing is, in fact, returning business impact? Like do you have any kind of a playbook for making sure that your work is valuable or advice for your clients? How do we as sensory and consumer scientist, make sure that we are, in fact, helping the business?
Ludovic: Good question. Yeah, what I always say is that we should stop creating power points. Let's stop sending Excel sheets with lots of data. And I don't know what like if you cannot tell what you found in two or three slides or if you cannot summarize it in a 30 second video, what it means for the CEO of that company, that you better go and do something else. So I think what we can do better is like, hey, put yourself in the shoes of the VP of marketing, the VP of R&D of one of those big corporations and just make sure that we can inspire them. So rather than sending a report, do a workshop, create the better or whatever, do something creative. So I think we miss a bit creativity. Like, if I look at what advertisement agencies do, they do a little bit of research and a lot of creativity. Well, I think we can learn a bit by them, like, okay, find an insight and then make sure that you put a lot of emotion because that is sometimes what we forget to be our researchers. We are very rational people. But yeah, CEO's and VP's. They are just people like you and me, you know, so they like a great story. They like some emotion to it. And then I'm sure that they will understand actually what is the value of what we're doing. This I would say invest more in creativity, storytelling and connect with all the divisions not only research, show that you understand what they have on their plate. And you can answer in a way you can talk to them in their language.
John: Alright. I could agree with that. Yeah, I mean, that's why technology is valuable to us helping to integrate things. I mean, there's a theme I see is you have speed and integration are the two things that technology seems to be bringing the most obvious. There's other changes, of course. So Ludovic, we have somehow made it twenty five minutes and it feels like 30 seconds. So if you could have any, like, parting advice that you would want to give our listeners as a sensory and consumer scientist, for example, what should people be thinking about for the next two years. What do you see as like most fruitful areas of activity for our field? You know this, short term.
Ludovic: I would really recommend to experiment a lot with new technology to embed new technology, use a chatbot, use some algorithms to see what it brings because you know what you don't know you don't love. But we will have to and, you know, embedded anyway. So I would really recommend to spend a lot of your time and invest really in technology. I use that. But keep in mind that actually I believe a lot in the future is not for the machines. The future is not for humans neither. It's for the combination of human machine. So I would really invest in in like how can we increase the value of what we do, everything that can be automated. Automated, so that you keep your mind open for actually what the data means because that is something that I don't see an algorithm do very soon is like connecting those dots and having conclusions, recommendations. That is something that you are way better than any algorithm. So, yeah, so let's try to build and do that together so we can focus on the added value. That's my advice.
John: Yeah. It's great. Very good advice. So Ludovic, this has been great. I really appreciate you being on the show. Where can people find you? People that want to connect with you either personally or maybe hire Haystack, how should they reach out to you?
Ludovic: Well, you can find me on LinkedIn. So if you go to LinkedIn, then you go Haystack Consulting, you will find the company. But you also will find my contact details there. I'm also on LinkedIn. So I'm very very connected to those tools. You can also called me on Facebook, on Instagram. So very easy to find. And I'll be happy if there was anything you would like to chat about. I;m more than happy to respond to your message.
John: Okay. Great. And you'll also be at some conferences, I assume. You'll be at Eurosense, I guess and ASN.
Ludovic: Yes. ASN has two conferences every year. Esomar, IEX. So yeah I love to go to the conferences, chat with people, learn. So yeah. You can find me there.
John: Okay. Perfect. Alright. Well, thank you so much, Ludovic, this has been great. Any parting words? Anything else you want to say?
Ludovic: I want to say that I'm really honored that you had me today on your show, so I'm really really happy with that. I really like the conversation. I wish you a lot of success with everything you do and happy 2020 for everybody.
John: Yeah. Great. Thanks a lot. Okay. That's it. Hope you enjoyed this conversation. If you did, please help us grow our audience by telling a friend about AigoraCast and leaving us a positive review on iTunes. Thanks.
That's it for now. If you'd like to receive email updates from Aigora, including weekly video recaps of our blog activity, click on the button below to join our email list. Thanks for stopping by!