The software taking the teeth out of manipulative marketing

Like
+ Add to

As someone who loves whiskey, I know that most purchase decisions come down to marketing and how the bottle looks. Well, today’s guest noticed that too–particularly in the wine industry.

So she created a solution to demystify taste and help consumers make better choices based on their actual preferences.

Katerina Axelsson is the founder of Tastry, a SaaS and insights company using AI and flavor chemistry to predict how consumers will perceive sensory-based products before they hit the market.

The podcast is in all major apps, just search for Mixergy.
You can also use our RSS Feed RSS feed.

Katerina Axelsson

Katerina Axelsson

Tastry

Katerina Axelsson is the founder of Tastry, a SaaS and insights company using AI and flavor chemistry to predict how consumers will perceive sensory-based products before they hit the market.

roll-angle

Full Interview Transcript

Andrew: Hey there, freedom fighters. My name is Andrew Warner. I’m the founder of Mixergy, where I interview entrepreneurs about how they built their businesses. And when I’m not interviewing entrepreneurs, I, uh, I love drinking whiskey. The problem with drinking whiskey is that, um, the only way, you know, what a good whiskey is is if a friend introduces you to it, or maybe you go to the store and the bottle looks nice or it’s priced high.

And it’s got an interesting story. So you go and get it. All right. When you get into whiskey Catarina, I get really excited. Right? So I really believe that most people, for example, would prefer Johnny Walker, red to black. And I think most people who go Johnny Walker green and all these other colors don’t know what they’re talking about, but if they could actually taste it, they would know.

What they really liked the judging based on the bottle and the promotion. Number one. And number two, I think that if I’m sticking with awry, it’s because awry whiskey is all over the place. So I stick with one bride that I know because I’m afraid that some other ride’s going to taste like the stuff that you drink when you don’t have much money, but you want to get really, really drunk.

Anyway, this is a terrible introduction to a product I’m really excited about Karina. Axleson

sorry. Recognize a similar problem with wine that we don’t know what we love until we taste it. We don’t know what, what is the wine for us? And especially now that we’re buying more and more of our products online, it’s all left to marketing and somebody else’s review.

And that both of those things. Are are manipulative. So she said, I think I could create a product software that will actually tell consumers what they, what they would love based on what we know about their tastes. And she’s done it. She started with wine. She’s going to expand to other products. The company is called tasty.

Um, a little too excited about it. Catarina, whenever I’m excited about technology, I should just write the intro because I see that I get carried away with the description, but I think I’ve done a good job explaining it. And now we’re going to talk about how she built uptake street, and we could do a thanks to my sponsor HostGator.

And later on, I’ll tell you why. If you want to host a website, you should go to hostgator.com/mixergy. First Catarina. Thanks.

Katerina: Thanks.

so much for having me, Andrew.

Andrew: discovered this because you were working in the wine industry and you saw something about ratings and consumer sales. What did you do?

Katerina: Well, I’ll tell you real quick. I, I, um, yeah, I definitely got the inside scoop, so I, I paid my way through college by working as a quality control chemist in the wine industry. I went, I went to Cal poly university and we have, you know, what is it? 400 wineries in a 20 mile radius, something like that. And, um, I, Uh, First off, I was given a lot of freedom to run my own mad, signed, to type experiments in the lab.

And I took full advantage of that. Um, and I stayed after work hours and tinkered and, and had come up with inventions before taste story. We don’t want to get into that, but, um, but during this time I, uh, I was became acquainted with the idiosyncrasies, um, uh, when, uh, not only in the wine industry, but really I noticed when it comes to developing a.

And bringing it to market. One of the things I noticed is that, uh, we had a multi-million dollar batch of wine, you know, giant tanks, lots of product sitting there. And we ended up selling half of that product under one label to one client, the other half of the same product to another, like. Another client.

And so the same product went out into the market under different labels, with different marketing, different price points, and then consequently it received different scores from, um, I think it was some of the same critics, if not, not that time than another time. Um, so from a consumer perspective, there’s no way you would know that the same wine is on the shelf somewhere just with different branding.

Um, and that, that was just, I think that was the first inspiration because. There’s an overwhelm of choice. There’s kind of a fog of intuition and a lack of transparency. And, uh, I thought technology could objectify this.

Andrew: Uh, sorry, isn’t the packaging and the whole experience of the bottle and the story behind it. Part of what makes a wine tastes better or worse. So maybe it’s not really about the taste. Maybe most people don’t really care.

Katerina: Yeah, no, that’s a, that’s a really great question. And it’s one that I was asking for, um, years as the technology was getting built. Right. Um, what we found is that, um, what’s inside the bottle absolutely matters after the first time. So you can sell someone almost anything once. You know, I, I, you can convince me that this coffee is going to be great, but if I don’t have a good experience, I’m never going to buy it again.

And the, the Dwight, if we were just to talk about like the business, like benefits and efficacy of this here, 85% of CPGs, um, and, uh, fail after, uh, launching into the market 80 there’s an 85% failure rate. And that’s what the best folk,

Andrew: Packaged

Katerina: uh, yep. So.

Andrew: We’re talking about what types of products beyond wine? What are

Katerina: Coffee grants. Um, you know, uh, what’s a packaged food. Like, like anything you can taste or smell it, you 85% of it fails. And beyond

Andrew: And you were saying they put millions of dollars into marketing these products often, and they still fail. But is it because the consumer, why, why do you think that is.

Katerina: Uh, well, one of the reasons is your product on average only reaches 0.7% of your potential customer base before the fate of the product is determined. Um, and other major driving factor is repeat purchase. It’s very, very difficult to get a repeat purchase. So you’re kind of. Going out, looking for a needle in the haystack.

There are people that are going to like your product. Like we know that but if you don’t find them fast enough, they’re never going to find it. So one of the objectives is.

is we know someone is going to like your product out there. How do you find them? And how do you mitigate the risk of people having a bad experience?

The first time they try something. Um, and I go, like, I could, there’s so much to unpack here, but I don’t want to be a monologue.

Andrew: No. I want to hear more from you. I’ve got to say though, with whiskey, I could totally tell the difference with wine. Maybe a little bit with coffee. I hate to say, because there’s so many coffee snobs out there. I can’t tell the difference unless it’s really extraordinarily different. It doesn’t really matter.

And so the packaging does matter. The price matters. The story behind who created it matters. And so isn’t that true for more, most people? How do you know that it’s not true for most people? And most

Katerina: Yeah, absolutely. By the way, I’m the total opposite. I’m super into coffee. I could totally tell all the nuances and difference and have totally been, I guess I trained myself to do that, but whiskey is still mysterious to me. So maybe we should snuck, swab some, uh, knowledge here. Um, we know that there’s an impact because.

Well, let’s just look as an outsider. These companies are spending millions and millions of dollars focus, grouping the product and trying to find the optimal formulation before it goes to market. The second thing is, is we found that, um, with our methodology, um, when someone buys a product based on a taste tree recommendation, as opposed to choosing a wine, for example, on their own, they rate the product on average 45%.

Um, so people are more likely to like the product when we recommended as opposed to randomly or through a friend recommendation, whatever it may be. So, so off the bat right there, there’s a valuable use case. Um, I couldn’t keep going.

Andrew: All right. And I should say that if I shouldn’t be arguing that nobody cares about taste. Obviously people have flavors that they love and, and they’re unique to them. Um, I think I see where you’re coming from on this. And so you believed that you could somehow understand what people like based on a survey that they fill out.

Katerina: Yeah.

Andrew: was before you got started with pastry, that was your premise.

Katerina: Um, right. We needed to mitigate the cold start problem. Typical of most recommender systems, right? We weren’t Netflix and we weren’t Amazon and we didn’t have 11 million users, uh, where I could just collaborative. We filter that data. And then, um, they give you recommendations on things that are similar to.

To you um, by the way, that method doesn’t really work for sensory based products, the way it does for other products, it’s not optimal. I mean?

um, it like the type of algorithms that are used for providing recommendations that don’t require. Initial user input. They require a ton of data on you, Right.

So like I buy things on Amazon and I buy books and I dunno, bottled water and whatnot.

And they look at that data and compare it to someone who is similar to me and assume that we’re going to like the same or not the same products, um, for sensory based products like wine that’s that doesn’t work. We found before. That person could buy, I don’t know, radio had albums and drive up. I’m just making this up a range Rover or a Honda fit or whatever, but we can go out to dinner and she can love that Chardonnay on the table and I could hate it.

And there’s no latent variable in the data that can help me indicate that. Um, so, so we found that the answer’s in the chemistry. So w the way we provide recommendations is the way it’s different is, is we use information based algorithms and we use. Data sets that we had to generate in house. The first one is on the unique chemistry of those products.

Um, we had gathered that data ourselves because we were gathering the chemistry. The focus in mind that we want to figure out how consumers are perceiving things as opposed to for quality control purposes. The second data is on blind ratings that consumers had on products that we, we, we poured a lot of wine.

A lot of people let’s just say thousands of wines, thousands of people. In order to understand how to improve the odds that you’re going to love, the first product you buy. Um, we have to get an input from you because we don’t have historical data from you. Right. You didn’t tell me these are the 20 wines I already love.

So find me one that’s similar to. So our, the way to short circuit, that was to ask you 10 questions about how you like flavors, you already understand, like, how do you like black coffee, um, or licorice or dark chocolate. Um, so you don’t have to be an experienced wine drinker, which is the other added benefit.

And after about 20 seconds, what it’s doing, and I don’t want to nerd out and I do have a tendency to do that. So feel free to stop me, but what we’re doing with that data, Did those questions are analogs to the underlying chemistry of the product and we’re pinpointing and fingerprinting your palette.

And then we don’t need historical data from you anymore.

Andrew: Okay. All right. So that, that’s what you’re doing today. Revenue wise, where are you?

Katerina: Um, we’re under 2 million in revenue we’ve made over a million in revenue in the two years. We’ve been making revenue.

Andrew: Okay. All right. And so you had this idea and you said I need to get started. What’s the first step you took.

Katerina: Oh, well, so th that’s hard question for me to answer because before it was ever a business, uh, for three years, it was just a. Research project with myself and, uh, PhDs and data science and experts in machine learning and sensory chemists. And it was largely an intellectual exercise. So, um, you know, for the first three years, we just, we wanted to prove that we could teach a computer, how to taste.

I think I incorporated the company in mid 2016, just so that we can get, uh, more equipment. Um, so.

Andrew: Because as a company, you could get equipment. What, why does being a, having a company help you get

Katerina: I knew I wanted a small business loan, so I

Andrew: Okay, got it. So that you could then go and get the equipment to do it because, so for you, it wasn’t even about starting the business and, and building a big business as the end goal necessarily. It seems like from what I know about you, you just love these intellectual exercises.

Katerina: Well, sure. I would say running a business, um, especially in a category. That’s new is certainly its own intellectual exercise. Um, but I, I didn’t ever want to start a business for the sake of starting a business. I think, um, you know, I stumbled upon in a way this, um, invention and there’s a lot of applications and use cases for it and I want to see it through and that was fortunate.

Um, yeah.

Andrew: Okay. The reason that I thought maybe you’re just an experimenter is you were telling our producer that when you were working as a chemist in a crush facility, in your spare time, you were able to run your own experiments, which was one of the things that you loved about the job. And one of the experiments was something you invented that influenced the surf sulfur process in wine.

What was this? And why’d you do

Katerina: Oh, yeah. Um, uh, it was just really interesting to me. Sometimes I intuitively look at things that are like the status quo or the way. Currently done. And I just ask like very naive. We, why is it being done that way? And sulfur was one of those things. So, um, without getting too deep into it, um, you test the chemistry of a batch of wine.

For example, you calculate how much sulfur you need to add to that batch to protect it from, you know, um, damage. Um, and then you just, you put it in a barrel or tank or whatever, and you mix it in. I thought the oxygen, which is why you add sulfury to protect it?

from oxygen. It’s all sitting on the top.

Like there’s a concentration grading it. And a third of the tank or barrel is an aerobic. Right. Um, and most of the time it’s sitting still, even though you mix things. So I thought, why wouldn’t you just have the sulfur concentration, gradient match oxygen? So I, I just, um, what I did was I put sulfur in a teabag, but I added 35% less.

’cause I thought you could use less and suspended it at the top. And the result was is that a better quality wine was made. So it was more protected against oxygen and you use 35% less sulfur. So

Andrew: by putting it in a teabag above the wine, instead of

Katerina: Instead of just mixing it throughout the entire thing where you don’t need sulfur at the bottom. I estimated 35% of the tank because there’s no oxygen there.

So you don’t need to

Andrew: God, it’s just the top that’s exposed potentially to oxygen that you want the sulfur to eliminate oxygen. Is that right? Versus the very bottom, which is just coming into contact with other one. Got it. And this was your invention and what happened to that, to that invention? How much money do you make out of it?

What happened

Katerina: Oh, I made zero money off of it. I first off, I mean, I was young. I don’t know if he could make money off a little contraption. Um, I didn’t patent that invention, but I do think, uh, yeah, no, that, that company took that method and ran with it? Yeah.

Andrew: And it feels to me like that’s one of the influences that may be led you to form tasting into a business where you said, look, I had something, I didn’t know what I had. Cause I was just, I was just tinkering and I had this vision, but I didn’t realize how big it could be this time I’m going to do the right thing.

I’m going to start a company, have the company own the intellectual property, respect it as a thing that could potentially be industry changing. Am I right?

Katerina: Yeah, absolutely. And I mean, I mean, T stree the potential, like the benefits are of that on the industry. It’s so great. Like it’s huge. It’s a huge opportunity. So yeah, not one I would want to pass up.

Andrew: Okay. All right. So you had that idea. You got a business loan of how much money.

Katerina: Oh, for that equipment. Um, oh, must’ve been a hundred thousand dollars. Something like that. Yeah, that was our first one.

Andrew: have to sign a personal guarantee.

Katerina: no.

Andrew: No. Wow.

Katerina: some seed money at that point as well. So I had some very early investors who knew it was just an experiment who, um, uh, invested just to help me hire some, um, help some team members.

Andrew: How’d you get the seed money? How’d you get the investors.

Katerina: went through, um, this program at Cal poly called the CIE, the center for innovation and entrepreneurship. And I, I went into the summer program and just pitched this idea by first time, experimenting with a pitch. And I just said something along the lines of the wine industry is huge and it’s ripe for disruption and blah, blah, blah, or something like that.

I don’t really know what we’re going to do here yet, but. Look at me, I’m a scientist. And, um, yeah, unfortunately, um, that convinced a few people,

Andrew: Sounds perfect. Okay. It sounds like the perfect pitch. All right. So you’re in business. What’s the equipment that you needed in order to run these tests and understand what flavors people would like,

Katerina: um, initially it was just a quadrupole GCMS. Um, now we’ve moved onto more advanced, similar equipment, but basically what it is

Andrew: what is the quadripole CMS?

Katerina: GCMS. Yeah, it’s a, it’s a gas chromatograph. So what it does is is we, um, volatilize a compound, sorry, I’m trying. What we’re doing is, is we’re looking, we’re breaking down the compound down to the ion or molecule count and we’re understanding exactly what compounds that solution consists.

Um, so if we want to quantify how many parts per million of benzaldehyde there are, which is a compound responsible for the cherry flavor. In many cases, the GCMS will tell us that so we can analyze the deep chemical soup of chemistry in a product.

Andrew: Uh, because if you don’t, you don’t want to understand the flavor based on how people had understood it before you had this other vision for how computers need to understand flavors. What was it before and what was your vision for how a computer needs to understand

Katerina: Yeah, absolutely. That’s why I think it took three years to make it a real business. Um, as I was tinkering in the lab, I w needed to understand what the status quo was and why was it not working? Why are 85% of products failing? Um, and that led me down a rabbit hole where I talked to, you know, the head researcher.

IBM. Um, I talked to people at Google brain. I met with the CEO of one of the largest flavor and fragrance companies. And, um, I, I, you know, it’s related research, but what I understood they were doing was testing the chemistry of products and quantifying compounds. So what they’re trying to do is predict, does this compound express this flavor, like for example, does bends out of hide equal cherry?

It doesn’t matter. It could be a coffee, fragrance wine, um, They can’t predict it. And I found my hypothesis was, is they can’t predict the flavor because there’s hundreds of other compounds in that chemical soup that are masking or expressing the original when you’re looking for. The problem is there was no method to look at all the chemistry in one snapshot.

And why is that important? You know, like as humans, we experience all that chemistry at once. So like when I’m drinking this coffee, I’m not experiencing one, um, one compound at a time, the way the equipment is, I’m experiencing them all at once and that’s going to affect how you perceive. So the key innovation was developing a method that looked at everything in one snapshot, the same way the human palette would.

Um, so that was one observation. Um, if that, if that’s helpful

Andrew: It is. And I could understand why it would take me three years to go through this. And then once you were ready, you decided to go after retailers. You said, look, these are the people who are trying to grow sales. They probably, if I understand you, right, your experience in the wine industry helped, they’re probably trying to increase sales of wine.

You went to one and you offered them an API where they able to use the API.

Katerina: Absolutely not.

Andrew: No.

Katerina: Um, well, I mean, We had to build an interface for them because I realized for them to integrate in a tasty would have taken about a year and a half, like retailers do, did not, at least at the time move fast. So, um, I offered the intelligence of taste story and the insight it could provide on their consumer preferences and their.

This is great, but, um, it’s going to take us a year to implement, which blew my mind. So what I meant by that is, is we also had to build a product to enable the product we were trying to sell, which was a consumer facing, um, recommender, um, where customers can interact with it immediately. We had to provide that product.

Yeah.

Andrew: For the retailer or because you wanted to show the retailer what could be done.

Katerina: Uh, well, both, but yeah, if we showed the retailer how a consumer facing recommender should look, it would have taken them forever to implement it. So we’re just like, we’ll just build it ourselves. Yeah.

Andrew: Yeah, I’m used to, um, in the tech space, if somebody has an API that helps a business grow sales, they’ll hire a developer to implement the API. If they don’t have one on hand and they’ll just get it up and running and see results. But I have noticed that retailers are not the most tech savvy or the tech, the most tech eager.

Right.

Katerina: Yes that well, that’s definitely changed, especially, you know, since, um, since COVID

and since the whole foods acquisition, they’re, they’re definitely scrambling to get caught up, but yeah. Um, historically that has been the case,

Andrew: and you knew that that retailers weren’t able to implement your API after you started selling it to them after you had that first retailer sign up with you or before.

Katerina: um, I just, when I started talking to them, everyone was on board. You know, personalization is a huge initiative in retail now and, and has been for a while now. And they’re like, this is great. The wine IO is the biggest pain point for our customers. I think we all know that there’s so many options. There’s an overall whelm of choice.

This could help them. Uh, but, uh, You know, uh, we need help plugging into this API kind of thing. Like it was, the idea is great. How do we, how do we execute? Was the question.

Andrew: All right. And so you built this and now you were in the business of getting customers to try your app. How did that go for

Katerina: Yeah. Um, so, so yeah, app, um, just, just to clarify in this case can mean any, um, just an interface that sits on a website and existing app and app and its own, or, um, an iPad at the end of a wine aisle on a kiosk, um, that has the same.

Andrew: I

Katerina: Yeah, it’s, it’s, it’s a medium agnostic. It’s a powered by taste street, recommender engine, um, uh, model.

And, Uh, uh, we, uh, provided the app and sold that as part of the recommender engine. I hope I answered your question.

Andrew: Yeah. And so now I can’t get it on my I’m trying to get it on the iPad. It does it, it goes into a loop somehow, because when I, when I go to your site to get the ID, to get the app for my iPad, it asks me for my phone number. I put my phone number in it, then texts me right away. A link to the app and then that ends up taking me right back to the page.

I’m guessing that it’s only on an iPhone, is it in the app store?

Katerina: It is, it is in the app store. We’re probably developing on it right now. And. I text my developer right now. It should, it should

Andrew: No, that’s okay.

Katerina: you did bottle bird, right? That’s a, we kind of, yeah. Cool. Uh, let me get back to you on that. What’s.

Andrew: Nope. I see it here. Actually, it looks like for some reason it’s not coming in for me, but I can see it’s a survey followed by some recommendations with an understanding of why those recommendations were made. And I don’t see a lot of use on it. Right. It’s it seems like it’s more like your demo for retailers of what you could build for them.

Katerina: Yeah.

that’s exactly right. Um, I, I am not our C our team. Our company is not, um, really focused on being a consumer facing company. Our core competency is certainly B2B and the data and insight that we derive on customer behavior, and that’s where we focus, but we do have. You can call a technology demonstrator app, um, where we can show all the capabilities of this technology, um, and, and to kind of give a preview of what’s coming next.

So for example, we are very, very good at recommending wine, but beer and spirits are next. They’re in the pipeline. And, um, on top of that, we have so much retail data that now we have this automated food pairing engine to, to pair. Your beverage with the food that is in the store or your dinner, or a group of people that are having a bottle of wine.

We know how to recommend the best one for a group, that kind of thing.

Andrew: All right. I should say this interview is sponsored by HostGator. If you’re building a web. Um, I highly recommend going and hosting it on HostGator. I do it with my site. It’s inexpensive, it’s fast. It just works. And they scale with me as my business grows. They have better packages for me. I’ve definitely saved money by switching over.

My audience did not know until I finally talked about it in the ad. So I recommend it for you to inexpensive, just work scale with you. If you need a host, if you need a website hosted, go to hostgator.com/mixergy. When you throw the slash Mixergy at the end. Yes, I get credit. But also more importantly, you’re going to get their lowest possible price.

hostgator.com/m I N E R G Y. First retailer. You want to show them that they could sell more wine? How much more wine were they able to sell?

Katerina: So, um, it ranges, um, on average, 12% goes anywhere from five to 20%, depending on the traffic and percentage of customers that adopted. But it was, it was, uh, more than compelling. You can say it definitely made a difference in sales, 12% on average and an increase in margin of 18.

Andrew: But 12% of the people who took the food, took the survey who answered your questions about their taste were willing to buy more, right? It’s not overall,

Katerina: All right. So, so, so the way that number, um, that that’s a lift in the sales, the category. So we increased wine sales by, by Overall, by 12%.

Andrew: regardless of whether it’s not just for the people who are using the taste free experience it’s for overall wine

Katerina: no, it was, it was stunning. Um, you know, an average 30 to 40% of customers in that pilot actually got a tasty recommendation. So, so, so of the people that were buying water, In the wine aisle, um, in our initial pilots, 30 to 40% of them were buying wine after getting a taste recommendation. So that translated to an average of 12% in lift in sales.

Andrew: That’s significant. That’s significant. So w w can I talk about one of your customers I’m on their site

Katerina: Sure.

Andrew: Okay. Is it winemaker.com? They’re a customer of

Katerina: That’s right. More of a partner, you would say. Yeah.

Andrew: Okay. So I see that, uh, it looks like it’s a Swiss company. Am I right? Where can I can even do it in English? No, it’s either. It’s either Germany or Switzerland. And as soon as I go to their homepage, I see wines. I, I think the average person would just say, I want this wine. Let me look into, get some more information, tap it for that.

And then add it to a shopping cart. Looks like on this one wine that I. The recommending six, so sure. I’ll add six. At what point in the experience are they saying? Tell us more about, about your taste so we can recommend better wine. I don’t even see us. Somebody would experience

Katerina: There’s a page. If there’s a newsletter, um, they promote it.

heavily through that. So they, like, I got to get on the page, but there’s definitely a page that shows get personalized recommendations. They’re heavily promoting it now. So I think they’re driving traffic directly through, through their promos for that, but I can get you that, um, Uh,

it’s not my site, right.

So I got to find out what they’re doing, but, but it’s really, really personalization is really, really important to winemaker because. They’re featuring wine from all over Europe on their site. Um, and a lot of people in Europe are not used to that. They’re used to drinking the majority of the wine is coming from that country.

So there’s no historical data on consumer behavior drinking wine, cross border. So they’re kind of trying to aggregate all of that in pro wide, provide the wine of Europe for everyone in Europe. Um, so tasty was a great. I don’t know a benefit to them in this case. Yeah.

Andrew: Okay. I see it now they’re calling it wine matcher. The first question is how do you feel about the smell of cigar tobacco? I’m pretty happy with that. Uh, how do you feel about green bell pepper? I like it. I’m going to hit the smiley face on that. How do you feel about black pepper? I’m gonna hit the smiley face on that.

How do you feel about the smell of. Great too. How do you feel about the smell of tomatoes? I made medium on that. I’ll hit the no smiling, no frown. Uh, how do you feel about goat cheese? Eh, either one. Uh, how do you feel about black olives smiley? I’m basically saying yes to all this. Yes. To mushrooms raising.

Sure. How do you feel about the taste of black unsweetened coffee? I’ve wished that there were two smiley faces. I could put on that. All right. And now they’re analyzing my answers. I basically said yes to everything. And I could see this seems like it’s your, your site embedded in their site, so they didn’t have to create any of it.

And once, once you’re done, they automatically show me wines from their selection on my screen. I get it. I see how this works. And based on those simple questions, you can give me a wine that I love.

Katerina: Well, it looks simple on the front in deceptively whimsical. I’m sure. But, um, Yeah.

there’s a lot going on in the backend. Yes we can. We, um, we can understand, we found about 85% of your palate with that set of questions. Of course, there’s a lot of diminishing returns with the amount of questions we ask with that gets us a really good, good starting point for sure.

Andrew: It feels awfully simplistic. You could tell based on questions like that, what type of whiskey I would like what type of coffee I’d like,

Katerina: So, so remember you’re not answering all the questions, but there’s so many permutations of how you answer those questions in combination. And we certainly look at that, like, like I’ll put it this way. We w it’s very important to us when you don’t like something like, we know you don’t like what you don’t like more than you, like, what you like, there’s a weight to that.

And we’re also looking at how you’re answering the questions in combination with each other. So like with this method, we can identify, you know, minimum. What is it over 150,000 unique pallets. So we can get really, really granular as to how you’re different from another person and the chemistry and the robustness of that is supporting that.

Um, so it looks fun. Um, but uh, yeah.

Andrew: but there’s a lot going on. All right. And I like how once I was done, they gave me a few selections and then they said, if I really want all of my information, I, and my scores, I need to enter my email address. So then they have me on the email list, which is a clever, clever thing to do. All right. Let me, let me try to just see if I do all smiley faces versus all frown faces.

Does it end up giving me a different reason? We’re analyzing. Okay, please wait. Data is, is loading. You’re doing this all in real time for them. Oh yeah. Everything is different. Okay.

One of the things you told our producers, because this is brand new stuff, it’s just really hard to explain it. It’s really hard to get new customers. You’re basically inventing a new type of business that embeds itself in somebody else’s business. so what’s, what’s your process for, for doing that for explaining it and for growing sales?

Katerina: Oh, my goodness. Um, well, we, we got really good at efficacy tests, as in proving the technology works in the real world and we, um, we speak very, very strongly to a certain value props. And they’re all about either making more money or saving a lot of money or in many cases being more efficient with both those things.

So, uh, You know, I think, I think we’ve mastered the look, this is new and it, you know, it’s very novel and it’s maybe hard to wrap your head around, but we can make you a better product that will outperform your own blind tasting panels. And we could do it in 48 hours at a 10th to 20th of the cost. Um, uh, as opposed to it taking you the traditional six to eight weeks.

So people pay attention. You know, there’s millions And millions of dollars at stake. Um, so that’s how we got our foot in the door.

Andrew: And are you putting the sales team together? It doesn’t feel like the person who’s the chemist who cares about the, the details of the equipment and how our pallets work is going to also be the person to put the sales process together. Who’s doing that.

Katerina: I would say I I’m actually, I disarm people and I can actually be quite a good sales person, but, um, no, we have, um, we have a very experienced a sales person. He came from, he was the head of global sales at sugar CRM, uh, before a taste stree. And he had a lot of experience, um, talking to high value clients, um, and doing big sales.

So. Joined the company, um, pretty early, um, actually helped us get to product market fit and he, um, helps with hiring, um, our salespeople, to be honest at the moment, we’re not really selling, we’re largely focused on customer success because our customer base is. Um, so advanced, like, like we have more customers than the infrastructure of the company can handle.

And that really largely happened in the last six months. So we’re just, we’re just focused on dealing with the clients we already have today, which is

Andrew: Meaning more retailers are coming on board and then saying, we want updates the way that you’re sending us information, we need more from you so that we can, we can include you in our products and

Katerina: retailers somewhat, but mostly it’s the brain. Um, the brands are all over this. We have.

Andrew: How, how are brands working with

Katerina: Oh, my goodness. Yeah. Um, so, so Yeah.

just, just to kind of like put a pretty bow on this, if possible. Um, we’re a B2B company. We aim to retailers. We sell the ability for them to provide personalized recommendations to their customers, but because we’re getting hyper granular on customer behavior, it’s aggregate, we’re not dealing with private data or anything like that.

They’re able to optimize their assortment in the store and reduce waste on. On the manufacturing side or the winery side, um, they’re using that same nationally distributed customer pallet data or, or customer pallet data in general to, um, identify their market of opportunity. So where are the customers that are going to absolutely love their product and buy it again?

And how do they get into that market? And then the final thing we do. With this technology is computationally blend to that product in many cases to ensure that it doesn’t fail on the market after a multi-year multi-million dollar

Andrew: Uh, so,

Katerina: Yeah.

Andrew: so wineries, or I guess, is it wineries or distributors who are paying you to say where we have this product? Where should, where should we go and sell it? Where are we going to find customers who are going to want it and buy it?

Katerina: Currently, currently it’s wineries, but we’re beta testing. Some of that data. Yeah. With districts.

Andrew: Okay. And then you also have winery saying, we want to create a new type of wine based on people’s tastes, help us understand what they’re looking for, sell us that, um, that expertise.

And then they’re, they’re creating better products based on what you

Katerina: Right. And in some cases it’s more cutthroat and not cutthroat, but like, they’ll say, look, our competitor took half our market share for this label last year. Why did that happen? What did they do? Where’s the overlap? Why was this vintage inferior to last year’s vintage? And we can give them. all those answers.

So, yeah,

Andrew: based on what, based on what.

Katerina: on the consumer data, Um,

that we have based on the fact that we understand. what, is it the majority of the market in terms of chemistry and differentiation, and based on the fact that, um, we, we understand that the answer’s in the chemistry and we know how to measure the chemistry to get the answer.

So we, I say we built the most sophisticated consumer and sensory database slash engine, uh, to answer these questions, that current technology couldn’t answer. ‘

Andrew: cause your, your hardware and software can analyze their wine and compare to the wines that are selling in different markets. Is that it got it. It’s like a super taster.

Katerina: Yes, exactly.

Andrew: then give them feedback. It’s not just what people are answering in surveys. It’s a super taster. And then how, how else do you know what’s selling in different, uh, in different regions?

Are you getting data back from retailers?

Katerina: In many cases. Yeah, we are. We do know what selling, we have channel partners. Um, uh, I don’t know if I should bring them up, but they have, uh, historical purchase data. Uh, what is it? 150,000 locations in the U S market. So that’s pretty helpful. Um, yeah,

Andrew: Did you see taste change during COVID when people drinking at home versus drinking

Katerina: I don’t know. I hasn’t, we haven’t been around long enough to really see or be able to predict that drift. Um, but I think I, you know, I could say everyone’s talking about how people are, um, drinking more and drinking more and more premium wine. And it would say we certainly see.

Andrew: All right. Speaking of COVID, we’ll close out with this. You told our producer that that really changed your timeline. It changed your business. How did COVID change your business?

Katerina: Yeah. Uh, we were very heavy in, uh, into just working with retailers at the time. And we were talking about this big company vision and how we were going to vertically integrate and use our two data sets to kind of power the supply chain with insight. Like that was always the vision, but I didn’t expect to be working with brands.

You know, a year or two down the line when we really got some more traction, but when COVID hit, it became like a necessity. I didn’t want to depend on retail. I don’t want the business to depend on retail. And fortunately that worked out really well.

Andrew: All right. Congratulations on your success. You’ve told me, you know, what your growth is going to be for next year. Can you say that

Katerina: 550%.

Andrew: How, how can you tell that it’s going to find that.

Katerina: Oh, because our clients right now are buying so much product and they already have, we already have a backlog of product we need to deliver to them. The only thing that stopping

Andrew: by product delivered to

Katerina: date away, more software, more data, more insight on consumers, more blends.

Andrew: more insight. Okay. And the only thing that’s stopping you from being able to deliver

Katerina: The front end of the software, it’s still MVP. It needs to be more robust. So we have a backlog of business we’d need to deliver on, um, which means get them front end, caught up with the backend. Um, so good problem. I think, but yeah, that’s our problem.

Andrew: All right. The website is it’s taste street.com. And thanks for coming on here and telling your

Katerina: Thanks so much, Andrew. It was fun.

Andrew: Thanks Katrina. Bye bye everyone.

Who should we feature on Mixergy? Let us know who you think would make a great interviewee.

x