Martha Bajec - Sensory as Input

Welcome to "AigoraCast", conversations with industry experts on how new technologies are transforming sensory and consumer science!

AigoraCast is available on Apple Podcasts, Stitcher, Google Podcasts, Spotify, PodCast Republic, Pandora, and Amazon Music. Remember to subscribe, and please leave a positive review if you like what you hear!

Martha Bajec is the principal and founder of Bajec Senseworks consulting, located in Burlington ON. Martha is a highly-regarded scientist, and an expert at designing, directing, and delivering evidence-based, actionable insights. Martha is known for her clear and concise communication as an individual contributor and as a leader of internal and external multidisciplinary teams.

Martha on LinkedIn

Transcript (Semi-automated, forgive typos!)

John: So, Martha, thanks a lot for being on the show today.

Martha: Hi, John. It's good to be here.

John: Okay, great. So, Martha, just right off the bat, I know this isn't exactly new technologies, but it's very interesting to me. You were recently at Hexa, which is still is a cannabis startup. I guess they're going through different phases of their life cycle. But I would be really interested to know your take on how sensory is fitting into the cannabis space right at the moment.

Martha: Yeah, that's a great question right off the bat, John, way to go. I think cannabis is so new and so poorly defined as a category that producers aren't really sure where to tack on sensory or how to use sensory, so there's a lot of questions around how do we get the user, the experience they want, right? Because there's so much going on other than taste and smell cannabis. You have this sort of multipronged question of what are they sensing? And it's across a lot of modalities going on there. I think companies are trying to get a handle on the business and maybe the sensory end will be a bit slower. I know people are doing studies trying to get research done. It's difficult here because of the regulatory environment, how you can test, where you can test so those things are really difficult. If you're just looking for sensory testing, there's different ways to go about it. But again, the location of the actual facilities where you're doing the testing have to be so well equipped to be compliant with the regulations that there aren't. You know, it's a lot.

John: Do you mean in the event of adverse health events or what is it?

Martha: Yeah. I mean, the concern is in the event of anything, right? Like, if you have some amount of cannabis on premises for testing, where you going to store it? Who has access? How do you control? How do we know, right? Like, what's the tracking cameras? So the government's putting all of that has put all of that together. Now people have to show they're compliant with it. So it's being treated a fair bit different than alcohol in my experience with research between the two. Obviously, alcohol we're very familiar with and we all have the outlines of how to do research with it. But for cannabis, I have a lot of people reaching out and asking, like, what do we do if, how do we do when this happens? What do you do? Or how do you test for this when you can't have a person at the end of the line trying? One from each batch of everything you're producing that day because they'll fall off their rocker there. But, you know, what do you do? So it's been a lot of fun. There's a lot of need for sensory throughout product life cycle, innovation through quality. But it's building that up that I think companies are getting to..

John: Do you think it is still a few years away then of these companies really being ready for sensory departments, or what do you kind of see for the next couple of years here with cannabis?

Martha: I'm going to bet that company, because here in Canada now we've had as of October 17th last year, edibles are legal as well and vapes. And I think they were getting you know, you're into CPG. It's not just an agricultural products like flower that you're selling. You're selling packaged goods. So I think as that ramps up, they'll be more and more like with normal consumer packaged goods, right? What does it need to be to meet the expectations of the consumer, both in terms of flavor and all that goes with that and then terms to the effect as well?

John: Interesting. Okay.

Martha: Yeah. There's a lot.

John: Yeah, I have some questions I want to go back to. But I am wondering about do you think that the industry needs to consolidate first before, right now you have so many of these different start ups. Do you think a certain amount of consolidation has to happen? There have to be some kind of winners and losers. And then at that point, companies will be ready to turn their attention to, you know, really product optimization. How do you see that happening?

Martha: At first I thought, you know, you can do this sort of delight the consumer right off the bat with cannabis-infused, we'll call them if you know it's the easiest terminology. Cannabis infused products. You know that you could really nail that beginning, but then with everything, with the shell of cannabis, you know, being the regulatory environment and not being clear on what kind of testing you need for these things, because it's different than if you're just going forward, if you took the drug regulations and just inserted cannabis. It doesn't work, right? Because it's a recreational product, too. Nobody takes Tylenol for fun like. Well, I shouldn't say that. I don't know what people do. But here you have something that is both utilized in the medical capacity. Yeah, you've got those two sort of avenues, and obviously the branding is different. The products are maybe different or at least somewhat different. You got to figure out where you're going to play and then build out from that, I think, as well. There's going to have to be consolidation, I think. Or just the acceptance that it's sort of like wine where you have, you know, the the craft producers, wine makers that are artisanal in some way or have, you know, terroir. Sorry for butchering the French word, but, yeah, if you have something special or you're a big producer like Gallo and you have a program and you have everything set up. I think that's how it's going to end up falling out. And then I guess whoever decides what services they'll need, whether, you know, on a big scale, like a big CPG producer or artisanal. But I feel like that's the way it'll play out. We'll see, it's exciting.

John: Yeah, it is very interesting for sure. I mean, for one thing, I think it's interesting to think about the opportunity for machine learning to help because you said you can't have somebody evaluate all these products. Well, for example, with Carlsberg and their beer fingerprinting project that they're involved with Microsoft on. They're trying to predict what beers are gonna taste like based on an analytical chemical, things that can be measured. I mean, there may be some also sensory properties, but variables that don't require as much human involvement and certainly don't require the waiting time brewing, right? So this is another situation where if you can start to get at least in the right ballpark as far as what someone's going to taste like without having to evaluate it every time there's a I think a big benefit to that. So do you have opinions on the topic? Would you like to talk a little bit about it like when you were at Hexa, what sort of things were you looking at? Was it in that direction or was it more in the?

Martha: It was every direction. Like, here's I feel like it's a very unique. Both cannabis is a unique opportunity and the timing is awesome because it's an agricultural product and we consume it as such. But there's also components we put into normal packaged good products. So there's so many different ways you can come at that. But just on the on the end of quality, right? You need some kind of electronic device that informs you whether it matches the intended specifications as they were, right? So for thinking about testing there where you have a very strong potentially niebuhrian. How can I reduce the need for human testing as much as possible, especially if the human testing needs to look very clinical for at least for now, right? That's crazy expensive. So thinking about where can you insert an electronic tongue or an electronic nose? Obviously you have to be back calibrated to your human perception of your product specs that are key. But any little bit of equipment can help get you away from there and then you've got reams of data on your control tolerances in your products, too. So, you know, it's a lot better than just checking a box by a person at the end of each lot, I think. If you have the product specs lined up and that's how you're doing your quality and those match back to your sensory profiles, you're kind of you got everything. But it's a lot of information. So I think that's where you come in.

John: Yeah, definitely. So let me ask, I said to follow the other thread from before, I'm guessing not legal to start to ask you questions about the psychoactive experience because my understanding is that cannabis is much more than with alcohol. Like in all the alcohol testing I've ever been involved and there's never been any investigation as to like the quality of, you know, for lack of a better word, the buzz. You know, like I mean, there are many variance, you know, depending on the type of alcohol that you're consuming. So would you say then that the psychoactive experience associate with cannabis is more multi-dimensional than, say, the inebriation effects of alcohol?

Martha: Yeah. That's a really good question, because certainly there's a lot of anecdotal evidence if you just talk to users, but research wise. Yes, there are studies that suggest that you get a different quality of high. I guess, whether you feel social or, you know, the joke or the monikers, Indica is in the couch. That's the easy way to remember it. And, you know, there's no difference between these two plants. You know, per say, except for what they smell like and their cannabinoid content. So the terpenes and, you know, the cannabinoids are the keys and depending on the combination of, and there's so many terpenes, so many possible combinations, so many additional possible combinations of cannabinoids. You have to wonder, you could breed anything. Yeah. But getting around, you know, the quality of that high is definitely challenging because so many of the research papers that have been done are sort of trying to qualify it for thinking about addiction and use behavior sort of more on a clinical end. But once you start thinking about well, you know, all the possible pairings that you could have to like, do I want a sativa or bright up uplifting sort of high paired with dark chocolate notes or you know what do I want the sensory profile to be, which is inextricably linked to the high profile?

John: And that's because the cannabinoids and terpenes are intertwined.

Martha: Yeah. And it's complicated even more because if I get northern lights from one company, it's not necessarily the same northern lights that I'm getting from another company or even that the batch to batch is similar, right? Like, you're putting a lot of faith and consistency, but really, it's a plant. And if you mess with it, well, it's going to change.

John: If you think about the big beer producers, it is an amazing feat of engineering that Miller Lite tastes more or less the same for batch to batch, right? And it probably will be a long time before the, you know, I mean, I guess you'd call it food science comes up to speed on providing the same consistency when it comes to products that contain cannabis, right?Because, I mean, that is a major achievement of food engineering or beverage engineering.

Martha: Yeah. And if you think about, you know, the different qualities, thinking about alcohol, the different qualities, right? Like some people will say, oh, you know, champagne makes you giggly or up here, you know, my husband who's American and Canadian now but we lived up here.

John: Was there's a force conversion? Was he forced to become Canadian?

Martha: No. He did it independent of me already. I'm just the icing on the cake. Yeah. He refuses Rye, rye whiskey. And jokes that, you know, it makes us fight. We're fighting when we drink rye. But yeah, like so then, you know, think about the cannabis world where it's actually the aromatics, you know, terpenes are magic. And there's this whole sort of bubble around them. But really they're aromatic flavor compounds. If you start looking at the alcohols, we drink. Well, they do differ a lot on aromatic profiles. Alcohol is alcohol, right? Like, if you just drink alcohol, you get an effect. But through the different matrices you get, at least we think different outcomes.

John: Right. And it's also true that at least in some beers, there are other psychoactive compounds, right? Like a Belgian beers, you know, some of them can be quite intoxicating. Well beyond the ABV that there is things that aren't being measured, you know, by the traditional alcohol.

Martha: And absolutely, if you look at hops and cannabis, the two are very similar. So there are a lot of terpenes. There are a few really good publications from the last couple of years just comparing the two even. There's tons in there, so you start to wonder, yeah, like, okay, maybe there are this combination of flavor compounds that make you feisty or, you know giggly or whatever. The same as weed or so are cannabis, I should say.

John: Right. Yeah.

Martha: So you kind of ponder that a little bit and think back to Thanksgiving and Christmas this past, what did we serve? Maybe we won't do that again.

John: That's funny. Well, let me ask you a little bit about before we kind of move off this topic altogether. Is it actually legal though, to ask these questions, to ask people about how they're feeling? Or are you restricted to only flavor type questions? Flavor, aroma, you know, standard.

Martha: Yeah. So the last time I looked at the regulations, that's sort of my disclaimer. Everything had to be run through a clinical CRO, like you had to run as a clinical and you had to include I can't remember what it's called, but sort of a drug allowance application to Health Canada to be able to do those kinds of tests. Now where they got a little bit like, oh, wait, what about sensory? They basically said, if you are not asking any questions to do with outcome of the experience, like you're not asking about how do you feel or like your actual emotions or how higher you or, you know, like those kinds of questions. If you were only asking taste and smell and mouthfeel? Or visual attribute questions, that was fine. As soon as you ventured sort of into anything physics, I don't even know what the right word is, because it's all physiological, bio-psychological. As soon as you started asking anything about high, you know, are you high yet or how high are those kinds of things? Are you giggly? Then it's clinical.

John: Interesting.

Martha: So it was a lot of, and then, you know, talking to a few people while liking is a different attributes. It is an emotion. Yeah. So then you get into, okay, well don't you know, let's not stir the pot on philosophical questions, what is liking because you really need that one. I'm hoping it'll evolve, you know, because it's scary when you get rid of prohibition on something so dangerous as cannabis. Full sarcasm there. Yeah, I'm glad there's a lot of regulations and that we have to be very mindful of what we're doing with it, how we're doing it, where we're doing it. But it's a bit too stringent for being able to create a new market for this, I think.

John: Right. That's a tight leash really.

Martha: Yeah. It'll take time, but we're the same, I mean, at least in Ontario, we have government liquor control.

John: Yeah. You can't really do alcohol testing in California to a large extent. It's very hard to do any kind of alcohol test.

Martha: Really?

John: Yeah, I remember that from when I was in IFP, I don't really field studies now the way that I did when I was at Institute for Perception, but yeah, the alcohol controls in California are very strict. So I don't believe you can do any kind of central location. Something now you can do it like home use test. So, yeah, this weird thing where it's like some products you can make them as long as you don't make them too good. Right?

Martha: That's it. And it's true I used to call it like, you know, this industry. Now I don't call it that. That was also a sarcastic moniker for it. But like alcohol, cannabis and tobacco. Right? And now cannabis is certainly not a sin. Thank you. I know, it's not, I went to Catholic school. You know, I used to call them industries I won't anymore.

John: I think Jesus had much to say about cannabis, if I recall.

Martha: I bet he would like it. That's just my own feeling. Catholic school girl here. I think he would have been down with it. There so much that you could learn that you can't do.

John: Yeah, I would love that honestly to get all that data and I could stay, anyway, it doesn't matter. It's not going to happen.

Martha: Yes. Right. Close your eyes and meditate on what could be out of it.

John: Alright. Well, amazingly Martha, we are almost out of time. So let's do a quick kind of look, because we should talk a little bit about new technology, we talked about new industry here a lot and we talked a little bit about machine learning, but as far as the new technologies that you see impacting sensory science, because you've been at Altria, you've been at Coca-Cola, you've been at ACCE, you've been in many walks of life. You know, for our listeners, Martha is a PhD in Biological Sciences with a focus on sensory science, if I recall correctly. So you definitely have seen a lot in your career. What are the technologies that you think are the most interesting or the most promising for sensory science in the next couple of years here?

Martha: That's a really good question. I think all of the data and outputs that could be from these gigantic data sets that must live somewhere in all of these corporations, right? If you could get in there. And so just the different approach to data where if you had an overabundance of actual data, it was like, oh, no, and I remember this. This is a way back when I was young and when so much data, you know, when you're trying to figure out how do we get rid of a lot of it because we just want what's key. Now, you have all that and you can do so much with it, right? Because all of the infrastructures been built up around it. Where 10 years ago that wasn't in there. Now you can, like the work that you're doing is so possible and a few years ago, it wasn't. So it's really neat to see how quickly that's gone. But I feel like probably companies haven't kept up with the possibilities because there's a lot going on. I think it was Danielle Reid from menow and others that put out a paper looking at. Basically taking Amazon ratings and comments and figuring out that, you know, everything's too sweet. No matter how much we tell you, we want it sweeter. It's all too sweet. And some other really interesting outcomes. And that application of like, those are you know, they may not be emotion or intrinsic, but that's really based consumer response. I'm taking the time to go and put a comment on this. And thinking of all the different review sites like here in Canada, we have Lift & Co. and it's a review site. And that's all you're doing is leaving reviews of cannabis companies, cannabis variety strains, whatever what you actually think of. There's so much data there that you could figure out exactly what to make for what sub-segments arguably just from that. Yeah. So don't give them that idea because I haven't talked to them, really. Can I have your data? I think that kind of like getting into where consumers are just being themselves instead of, you know, here's this controlled form that we're going to give you. Of course, you have to have a lot of power behind it to be able to do analysis and understand what's going on. And then I really got into, I'm not a proponent yet, but I really like the idea of using e-tongue's and building out those kinds of databases for, hopeful more and more linkage to human perception. Electronic sensing.

John: Yeah. So few things since I trust that one, I definitely agree that technology supporting this kind of bridging of quantitative and qualitative research that you, like you're talking about like large scale, you mean lots and lots of reviews. Now, I think there might be a question of bias because you may have only certain types of people leaving reviews. That's why our controlled experiment is valuable. But I think you're going to increasingly see controlled experiments that are more like interviews where you're going to have things like chat bots or I just finished a survey on Alexa. You know, we've got a lot of smart, like speaker powered surveys. The text collecting ability of Alexa is still quite limited, but it's going to get better. You know, like it's really good for in the moment evaluations like you want, you know. When someone like in cannabis world where they first, you know, inhale and then you want to get their response right then, they don't have time to put down, you know, whatever the product is. Yeah. They can just cough out over for us. But whatever, you know, like in the moment is good. But you're going to have that bridging of quant and qual. And then to your second point, I think that there was this idea that e-tongue and e-nose was gonna be like all informative and I think that's not true. But it can certainly be partially informative. So as an input to a model that has many inputs, I think that is very valuable. So I think as you see machine learning coming online, you're going to see more and more of these other forms of data coming into the model. And I think that will provide extra value. So, yeah, I agree with you completely. So we've got about a minute and a half left here. So with our remaining time, suppose that I was a sensory science PhD or masters or even undergraduate, just graduated. Now I'm going to go out into the workforce, what advice do you have for me? It could be career advice. It could be scientific advice. What would you tell me if I was that freshly minted sensory science graduate, to help me make, you know, make sure that my career goes as well as possible?

Martha: That's a good question. Probably never say never. That's a big one for me. Across the board. Never say never because you don't know who's going to come to you with some solution. That's like, oh, my gosh, I never would have thought of that if I limited myself and the other one is talked to everyone who will talk to you.

John: Be open minded and network.

Martha: Yeah. Doesn't matter. Ideas and directions come from anywhere.

John: Yeah. That's why I think LinkedIn is actually a technology that's positively impacting sensory consumer science. Connect to various people. You know so many guests on the show, I mean we already knew each other, but like so many guests on AigoraCast come through LinkedIn like it's, you know, it definitely good to reach out and connect with people, talk to people like you're saying. Find out what's going on. The world's changing rapidly, so you can't stand your show and expect to, like, stay up to speed.

Martha: Yeah, well, and even if you are in a company, something I wish I knew at the beginning across departments, talk to people, you know, like your group isn't in isolation. And when you're coming out of school and you have no idea sort of how these things look on the inside. Yeah. You know, talk to the people that are in the factory. Talk to everybody because who knows where you're going to get your next, like, hey, we can save a million dollars with that or there's an innovation on the product. So open, 2020 the year of open.

John: Yes. You have to be open. With all the new technology, everything is open. Open source, you know, open platforms. Everything is being forced to open and you have to just accept it and go with it rather than fighting it because it's just not the way of the future.

Martha: I've now started giving Netflix and Amazon Prime and all of these streaming and review services my true rating. I thought I was being clever before by like getting around them. And then I'm like, duh, you're totally cool with your robot overlords. Let them have it. So now I'm like, yeah, I like that. Give me more of that.

John: That's funny.

Martha: So open.

John: I have a funny story about streaming, but just real quick. My son is 4. He had to wait for a show. There was a show that was all going to be at seven o'clock on Saturday and he could not understand...

Martha: What's wrong with you people?

John: How could he not watch right now. He has grown up in a world where everything is streaming. I couldn't explain to him the idea that you might have to wait for a television show. So anyway, the world is changing for sure.

Martha: Because they don't have TV?

John: No, the television is there. But the idea that it's not available right now, just like couldn't fit in his brain. So funny. Alright, Martha. This has been great. So thank you very much for excellent conversation.

Martha: Thank you.

John: How do people get in touch with you if they want to follow up after the show?

Martha: LinkedIn is the best. I'm on Twitter, absolutely under just my name. Yeah. Or email martha.bajec@gmail.com. Super easy.

John: So we'll have all that information in the show notes.

Martha: Perfect.

John: Thanks a lot, Martha.

Martha: Thank you so much.