Elevating customer experience with AI

Publish date:

Can artificial intelligence be your personal shopping consultant?

Can artificial intelligence transform the way we shop for clothes?

In this episode, Peter Maloof and Samantha Burden from the Applied Innovation Exchange in New York sit down with Michelle Bacharach, Founder and CEO of Findmine to talk about their platform, and finding the perfect fit for AI in fashion retail.

Also available on:

 

 

 

Transcript:

Peter Maloof:       

Hi, this is Peter Maloof from the Applied Innovation Exchange with another installment of our podcast. I’m here with Sam Burden and Michelle Bacharach. Sam, do you want to introduce yourself and then Michelle?

Samantha Burden:

Sure, thanks. Hi Peter, Michelle. I’m Samantha Burden, working with Peter on the Applied Innovation Exchange team here in New York.

Michelle:     

Great. I’m Michelle Bacharach. I’m the co-founder and CEO of a company called FINDMINE, which is a retail technology startup. Glad to be here.

Peter Maloof:       

Thanks Michelle. I think the first thing that we want to talk about is just what you’re doing now, where you got started, and a little bit about your company.

Michelle:     

Yeah, great question. I told you guys before when we were having lunch that I was living in LA before moving to New York, but what I might not have mentioned, or you might not know as I was a professional actress during that time. I had been doing that growing up a little bit here and there, and then I went to Berkeley, which I didn’t study acting there, but I was a part of a private conservatory in San Francisco and I was working as a product manager at a startup doing this acting thing on the side in San Francisco, which isn’t very like, you know, there’s not a lot going on there. And so I was like, “Oh, if I’m going to try it and make a go of this, I’m going to move to LA and see if I want to be an actress, and if I don’t, I’m going to go back to business school and I’m going to get my MBA and focus on entertainment.”

I was living in LA and I looked fairly young for my age, so I was playing like 16 year old girls even though I was 25 at the time. Moving to New York then, when I eventually decided to go to business school, my wardrobe had to completely change and I had to learn how to dress for cold weather, which had never been a thing before and dress for business casual, which I had never had to do before. So I had all these questions like, “Can I wear a suit jacket without the skirt? How do I do that?” I didn’t even know. I had all these questions around the clothing that I was buying and wearing, so essentially I realized that I’m not alone. Most consumers don’t have the immediate answer for how to be successful with the products that they’re buying.

They might have one idea, “Oh yeah, I can wear it with a button down shirt and jeans,” but getting out of their comfort zone or wearing it in a different way that they might not have thought of is hard for most people. And retailers actually sell a lot more products when they can show the shopper, “Hey, here’s three or five different ways to use the product that you’re buying,” But that wasn’t happening very often because it’s manual. The retailer has to manually make that guidance happen for the customer. So our company FINDMINE automates the process of creating that content that helps guide the shopper. So in fashion, we show complete outfits around every product, for home furniture we’ll show you how to style your room around the white couch that you want to buy, and we license the software to the retailer so that they can show that kind of recipe for success on their e-commerce site, in the email campaign, in a messaging platform, in their advertisements, and then in the physical stores as well.

Peter Maloof:       

Wow. That’s an incredible journey. Talking about where you went from acting into business school, and now, part of the way that we got introduced to you is that you use artificial intelligence, which of course is the hottest topic and thing that’s going on in technology now. How did the AI piece kind of come into this?

Michelle:     

I believe firmly that AI is a means to an end and never the end in itself. Like AI for AI’s sake is ridiculous. You always use it to achieve a goal. So it really was just the best means to achieve the end of guide the consumer at scale because merchandisers and retailers and stylists and personal shoppers are manually putting together the lookbook or you go to Michaels craft store and there’s an end cap that has everything you need for an Easter egg hunt right now. But somebody had to think about that and do it by hand. So anytime you have humans doing something, it’s going to be rate limited. Humans are human, you can only move so quickly. But the really unique thing here was we need that human intelligence because how do you train a machine what’s in an Easter egg basket?

And also for our customers who have very strong brands, we work with Adidas, we work with John Varvatos, we work with BCBG, we’ll even work with multi-brand retailers who have a specific point of view from a curation and editorial perspective. We don’t want to get that wrong. And so just letting the machine go kind of do its thing, you need the human input to start. So it’s like the creativity and the inspiration comes from the person, but they don’t … There’s sort of diminishing returns to fun and utility building the hundredth outfit as the person. Building the first five, super valuable, pretty fun, creative, and then you’re sort of like, “I’m over this.” So AI is actually the perfect solution when you have that kind of scenario where you need the human input, that kind of creative or artistic point of view is important, but then there’s repetition after that. So that’s where the AI can be really powerful, and that’s exactly kind of like the solution that we needed for this particular problem set.

Peter Maloof:       

Okay. So essentially the machines are good at the mundane, so when the job turns from kind of creative into the mundane, that’s a good place for the machines to take over.

Michelle:     

Exactly.

Peter Maloof:       

What was your exposure to artificial intelligence before? I mean, you speak so knowledgeably about it right now and yet I’m trying to, other than having gone to school at Berkeley, what was your technical background?

Michelle:     

I didn’t even do any technical anything at Berkeley. I was going to be a Peace and Conflict Studies major and then I switched to … what did I switch to … Psychology, and then I found that they have this thing where you can make your own major and I was like, “Yes, that’s me,” and so I studied innovation. Actually, I made up a major called Managing Innovation Across Cultures and I wrote a thesis on it. None of that had to do anything with tech. And like computer science was one of the hardest things at Berkeley to do and I didn’t want anything to do with that. But I think that my interest is really in the user experience and kind of the consumer journey and the pain points or friction in a user experience, and being able to take those out always requires technology. So I was a product manager even while I was acting. I was acting part time, product manager full time, and then I was acting full time, product manager part time. And then after business school for two years I was a product manager.

I was in around tech enough, like you don’t want me coding, I would be terrible, It’s not a good use of anybody’s time, but I would need to get … I mean, when I worked for Univision, I had teams of engineers in four different countries, three different continents, where they would tell me, “Oh Michelle, we can’t do this,” and I would have to know enough to call bullshit if that was necessary and be like, “No, we can, here’s how and why and maybe you don’t want to, or the way I’ve described it is impossible, but here’s what we’re actually trying to achieve and there is a way to get there.” So I had to be knowledgeable enough to do that, and then just sort of like reading and learning and keeping up with the industry.

I will say though that like, especially if you asked my technical team, they will tell you that my knowledge is very executive summary level, right? So from a practical standpoint sometimes I’m like, “Why don’t you just x?” And then they’re like, “Oh, you have no idea what kind of complexity is involved in that.” So I have to check myself a lot that I’m not actually an expert in any of this because I’m not a practitioner of it.

Samantha Burden:

Well I think when you have that technical team support behind you, it sounds like you kind of, with your experience and your interests and your expertise, kind of understand where the opportunities and problems and what the consumer’s currently dealing with and what solutions you need to solve them, and then you kind of need to go back and say is this feasible? So from your perspective being in retail, what do you see in the industry as the opportunities for maybe AI but maybe other technologies to change the customer experience? What do you see consumers asking for or you see those opportunities?

Michelle:     

That’s a really great way to describe what you said about the technology, like the relationship with the technical team. I’m the hypothesis maker. I’m just like, “I wonder if, maybe this,” and then they have to go figure out is that feasible? What does the data say? Could we do this? And then we kind of iterate from there. But I have those hypotheses all the time about retail because I think retail is really interesting because that’s my industry as a professional but we’re also all consumers so we’re always exposed to this industry whereas that’s pretty atypical for all other industries. And so I think there’s a lot of areas AI can help, but I always go back to what the user experience is.

I would die to have somebody fix the standing in line at a store problem. I went to Macy’s, I bought something on Black Friday and I went to return it. I bought a bunch of stuff actually, but I went to return a couple things afterward and right after Black Friday, a lot of people were returning stuff because that’s a big shopping time and so there’s a ton of people in the store. I was standing in line to return something, so it was like a negative revenue opportunity for Macy’s, but I wanted to be shopping for other stuff but I didn’t want to lose my place in line because I knew five of people were waiting behind me. If they just let me take a number, I would’ve probably bought more stuff than I was returning. It could have solved the revenue issue for them and I would have been happy, they would’ve been happy. But I think people oftentimes, because of tech and all this cool new bells and whistles and whiz-bang shiny objects, they forget oftentimes about the most basic thing.

And for what we do too, I feel like that is always true. Why would any retailer think it’s okay not to show their customer how to use the product they’re buying, that they’re just assuming that they can sell you a thing one at a time, but nobody wears a shirt with no pants and no shoes. So this is like such an obvious gap to me, but I feel like a lot of times that’s what I’m seeing is something that seems like such an obvious gap, why hasn’t someone solved this yet?

Samantha Burden:

Right.

Peter Maloof:       

Yeah. As you were talking about this, one of the things that started coming up in my mind is when we first met you and saw some of the things that you were doing in terms of the complete the look, and I could see where you’re the hypothesis maker because I was trying to do some of it too because you talked a little bit about the black skirt and what would go with it, and then I started thinking about things that were sitting in my closet. If I had a picture of this and I could run it against your algorithm, then I probably would find something to go with it instead of it ending up at Goodwill, right? There’s a lot of abandoned articles that people have. So what are some of the other hypotheses that you’re thinking about with your platform? How do you see it evolving?

Michelle:     

Well, interestingly, the first hypothesis we had when we launched this was it’s going to increase average order value because if a customer can see for the shirt, the skirt and the shoe and the bag and the jacket that all go with it, they’re going to buy something else, so their total basket size is going to go up and that’s how we’re going to increase revenue. So that was true, we found that, but what was interesting was like across all of our customers, everyone has seen between 4% lift in total revenue and channels that we come into and 9% lift in total revenue. But the way they get it is really different depending on the kind of customer. So, average order value makes a big percentage of that 4 to 9% lift in sales for companies that have kind of lower price points on average. It’s an easy thing to add on a bag because the bag is only $30 and your total basket’s only 50 so it’s an impulse buy, easy to say yes to kind of thing.

Conversion rate was never a hypothesis we had that we were going to affect conversion rate. But what ended up happening was when we tracked the data, we found that … We have a customer who is a luxury men’s wear brand and people would buy the … They sell leather jackets, I think one of their big staple products, and for like $2,500. They’re expensive. And what we would find is that the conversion rate on those products would be higher because of our technology, even if they didn’t buy a single additional item, because if you can see three different ways to wear it, you’d be like, “Oh, I get it. I could wear this to work, I could wear it on a date, I could wear it on the weekend.” The mental ROI on this really expensive piece that you’ve been drawn to has now gone up and you’re like, “Hey, it’s justified now. I can buy it.” So that was a really surprising one.

And the other one was repeat purchase. So we’d see customers coming back and repeating their purchase more frequently, which makes sense in that you saw the recipe to how to wear the product, you bought the product, you get the product home, now you have to actually do the wearing of it and you’re like, “Oh yeah, I saw that outfit, let me go back,” and then you end up buying something again.

And the last one that I have a hypothesis around that we’ve never been able to measure is that we can reduce returns because exactly what you said. If there’s stuff that sits in your closet, if you just bought it and you haven’t worn it in two months and they have a three month return window, you’re like, “I really don’t want waste this. I don’t want this to go to Goodwill. I’m just going to return it.” So we’ve never been able to measure that because of how much difficulty there is around getting returns data, which lives in a different place within people’s data infrastructure, that I have a very strong hypothesis that that’s true for us and I would love to measure that.

Peter Maloof:       

Yeah, I bet that’s there.

Samantha Burden:

I think so. I mean, I can think of several things in my closet right now that I wish I knew what to wear it with. What I think is interesting also about kind of the way the tech works and the way that FINDMINE is set up is that you go back to when you’re making these recommendations, they go back to the stylist of that brand, right? So it’s like, you know, when I’m shopping at wherever it is, I’m going there because I want to wear those clothes. I want to kind of embody what that image is like. I want to look at athletic and I want to be stylish, whatever that might be.

But I think it shows the power of brands and with Instagram and social media, all these different things, blogs, I feel like differentiating your brand and showing like this is our look. I think that’s so powerful because I would take that advice over a random, like you mentioned, another company that kind of does this where they’ll just kind of pick things based on you. But the end of the day, I’m not a fashionista, right? I kind of want to go back to the brand and say like, “What do you recommend for me?”

Michelle:     

You want that authority and kind of like a stylistic point of view. That’s such a good point and something I’ve been talking about a lot recently is Amazon because you can buy a lot of the same things on Amazon as you can at the brand’s website or at a multi-brand retailer, and Amazon’s great at variety and convenience, but it can be overwhelming and they don’t have an editorial point of view. Whereas if you can buy that same thing on a multi-brand retailer’s website, who’s got some kind of fashion editorial [inaudible 00:15:46], and hold your hand, “You’re going to look great, we got your back,” then there’s an added benefit to going there even if you don’t get two day free shipping. You can almost make up for all the things that Amazon has in convenience because you’re providing this other kind of service.

When that multi-brand retailer doesn’t communicate that point of view stylistically at scale consistently across every single product and every single customer interaction, and they only show you what they think you want to buy, they’ve eliminated their competitive advantage. You might as well just go buy from Amazon.

Peter Maloof:       

Yeah, I agree. And I think because brands spend so much time focusing on their products, but the lifestyle that they’re projecting, right, I mean the way that you described it was an athletic lifestyle rather than kind of an athletic cut, right? And so I feel the same way where it would sure make it a lot easier if using FINDMINE to complete things that I just don’t want to continue to hunt and peck for, especially across these really large department stores that have a bunch of different brands, just being able to have something do that for me that still projects that lifestyle. It’s great that it’s true to the brand and it’s true to the things that they’re doing. So I can see where that would be a value.

The other thing that I started seeing is that if you do kind of embody that, something that I bought a few seasons ago or a few years ago, somebody who is … I came from a technical background where I wanted to know my customer’s install base, right, because if I knew what they had installed a couple of years ago, then I can kind of build on that. From a human perspective, it’s like, “Well if I had something from a couple of years ago, but it’s not completely out of style, is there something that is coming out with this new season that could update the look?” If our clients, right, our collective clients start thinking about that, they can kind of continue to to build on that. So I’m wondering if that’s going to be a future hypothesis for you guys or something you’re working on?

Michelle:     

Yeah, I mean that’s honestly like my little bit of the Berkeley, very liberal part of me coming back into the business because one of my hesitations when I started the company was, “Am I contributing to frivolous consumerism?” And the way I’ve justified it, whether others agree with it or not, is that if we can help consumers be more thoughtful about what they’re buying, you buy less of that one off, “Oh, I saw it, it looked cool, but I never really thought through how I was going to use it and it just sat in my house and then I ditched it and I got rid of it and ended up in a landfill somewhere.” If we can come back to you and say, “We know you have this in your closet, you bought it in the spring, now it’s fall, here’s your fall style guide. We can incorporate it into all these new things,” you might still buy new products but you’d be more intentional about each product.

And I believe that brands will win by having a bigger share of the customer’s wallet, not by getting customers to buy one or two things from them here and there, and that just contributes to that frivolous consumerism. But if they can be a go-to brand for you because you start realizing over time, “Hey, every time I go into my closet I pull out brand X and I feel like they got my back. They send me this email every quarter with my style guide. I know that they understand my needs and they meet my needs.” And then you’re like, “Well, when I think of ‘I need to buy a new jacket,'” where are you going to think of first? Brand X. So we’re actually helping retailers sell more, but everyone’s selling and buying more in a more thoughtful way that I’m hoping eliminates that whole, you got to go through your closet and donate, hopefully donate, but a lot of people don’t. It does end up in the trash, two thirds of the stuff that they have, every couple of years.

Samantha Burden:   I think that’s such a good point because when you talk about … it’s like another way to build loyalty and trust other than other programs that you can do, or all those different things. But if you’re using FINDMINE or having their … you said you go back to your closet and you always pick their product.

Yeah. It’s not just share of wallet share, it’s mind share. Like how often can you be a part of their thought process and if you’re always there in their Instagram feed or inbox in a helpful way, not just a, “Buy from me, buy from me, add, add, add,” kind of a way, then that changes the relationship. It feels more like a trusted friend and not a schlocky salesman.

Peter Maloof:       

That’s right. You’re creating more of that emotional connection, right, and as we’ve seen and as we talk with our clients, emotionally connected customers are far more valuable than even highly satisfied customers, right? So it fits into what you were saying is that they’re less price sensitive, they actually take advice from you. I mean, there’s all these other things that we’ve seen about emotionally connected customers and so you’re doing great to kind of help build that connectivity or that emotional [inaudible 00:20:41]. I know we’re coming up on it, on time, one thing I did want to ask you about since we just heard it, you said that you wrote your honors thesis on innovation across cultures? So how did that go? What was the premise behind that?

Michelle:     

I was actually interested in business back at my freshman year of Cal, but I took the intro to business class and I thought everyone in there was just like this smarmy capitalist that I didn’t want anything to do with.

Peter Maloof:       

Even at Berkeley?

Michelle:     

Well the business school is called Haas and there’s a joke that the people who go to Haas are called Haas-holes and they’re actually.

Samantha Burden:

They set themselves up for that one.

Michelle:     

Yeah, they totally did. I mean, there’s tons of people who go there who are my good friends and I love them, they’re wonderful, and it’s still Berkeley so it’s still very liberal. But I just kind of had this notion of business is not something for me. But I was interested in management still so I found this make your own major thing and I had to pick stuff from two or three different disciplines. I picked psychology, linguistics, and business and I kind of combined those. And what I ended up going down a rabbit hole of was like, it wasn’t first party research, it was all third party research, but was how different cultures sort of react differently to creativity tactics or innovation tactics.

So for example, Eastern and Western cultures have different values. And so in Eastern culture is a lot of times about harmony and so you’re not trying to be super different, whereas Western culture is like you’re trying to be different, actively. But one of the biggest things that influences creativity and innovation is divergent thinking, so anything that breaks you out of the norm. So if you’re working with a group of people from an Eastern culture, just being divergent is divergent enough for them to kind of like have that brain aha moment that sort of opens up pathways to creativity. Whereas in Western cultures, that’s kind of expected. Everyone’s expected to be divergent, so you have to find other angles to really be divergent and disagree and that’s why diversity is so important. And mixing up the kind of points of view, whether it’s from a ethnicity standpoint, racial, socioeconomic, gender, where you live in the country, suburban, urban, rural, that all unlocks creativity. So that was kind of the point of the whole research.

Peter Maloof:       

That’s cool. Yeah. So you get a little bit of the cognitive diversity manifests itself out of whatever the social, economic, or regional geographic diversity.

Michelle:     

Absolutely.

Peter Maloof:       

It’s fascinating. I mean, I think that’s something that we look at and here at the Applied Innovation Exchange, we’re always looking at what’s driving innovation, how people get inspired by things. Love the story that you have here. It’s great been working with you and some of the stuff that you’re doing for clients and looking forward to continuing our relationship and really appreciate you coming in to help us out on the podcast.

Michelle:     

Yeah, this was fun.

Related Resources

Technomics Episode 4: Cybersecurity

As we move into a world of near-zero infrastructure model with the rapid adoption of Cloud,...

World InsurTech Report 2019 Infographic

Get an overview of World InsurTech Report 2019 findings with our quick-read infographic.

World InsurTech Report 2019 Video

View video on World InsurTech Report 2019 key findings and see how established insurers and...