Meet The Founder Who Uses AI To Help Retail Brands 'Elevate' What They Do

  • 5 months ago
Founder and CEO of FindMine Michelle Bacharach joins "Forbes Talks" to discuss her company, which uses AI in the retail space, and her thoughts on AI over the past decade.

Subscribe to FORBES: https://www.youtube.com/user/Forbes?sub_confirmation=1

Fuel your success with Forbes. Gain unlimited access to premium journalism, including breaking news, groundbreaking in-depth reported stories, daily digests and more. Plus, members get a front-row seat at members-only events with leading thinkers and doers, access to premium video that can help you get ahead, an ad-light experience, early access to select products including NFT drops and more:

https://account.forbes.com/membership/?utm_source=youtube&utm_medium=display&utm_campaign=growth_non-sub_paid_subscribe_ytdescript

Stay Connected
Forbes newsletters: https://newsletters.editorial.forbes.com
Forbes on Facebook: http://fb.com/forbes
Forbes Video on Twitter: http://www.twitter.com/forbes
Forbes Video on Instagram: http://instagram.com/forbes
More From Forbes: http://forbes.com

Forbes covers the intersection of entrepreneurship, wealth, technology, business and lifestyle with a focus on people and success.
Transcript
00:00 (upbeat music)
00:02 - Hi everybody, I'm Brittany Lewis with Forbes.
00:05 Joining me now is Michelle Bakarach,
00:07 founder and CEO of FindMine.
00:09 Michelle, thanks so much for joining me.
00:10 - Thanks for having me, Brittany.
00:11 - Of course, and before we dive into the conversation,
00:14 can you explain what FindMine is?
00:17 - Yeah, so I think like most founders,
00:19 I started a company to solve my own problem,
00:21 which is that I'm not the most stylish person,
00:24 I don't have an eye for interior design,
00:26 I don't know how to do a smoky eye,
00:28 and I found that when I would walk by a window display
00:31 in a retail store, I would be like,
00:33 whoa, yes, this, all of it,
00:34 that's my aspirational life, right?
00:37 And I feel like a lot of consumers have that struggle,
00:38 so I started asking brands and retailers,
00:40 like, why don't you just do that for every product you sell?
00:43 And what I found is that, you know,
00:44 there's just not enough resources in the world
00:46 to do a window display for every single thing
00:48 that most retailers and brands sell,
00:49 so they get about 5% or 10% of the product catalog
00:52 into those inspirational moments,
00:54 and then consumers like you and me are sort of left
00:56 without that for everything else.
00:58 So what FindMine does is we apply artificial intelligence
01:00 to deeply understand sort of the brand's creative vision,
01:03 so we can do that inspiration for every single product
01:06 in the catalog, and the consumer gets more of that
01:08 more of the time.
01:10 - I wanna read some numbers.
01:11 WWD reported that FindMine raised nearly $9 million
01:15 in Series A funding, bringing the total raise to 17.6 million
01:20 so first, a big congratulations,
01:22 but second, the company was founded in 2014,
01:25 this was a few years before the generative AI boom
01:28 of late 2022, so how did you get investors on board so early?
01:33 - Great question, and I think that like,
01:35 this fundraising round that we just did
01:37 was like the easiest I've ever experienced,
01:40 partly because I've been beating the drum for AI
01:42 for 10 years, and a lot of times people would be like,
01:46 cool, but like, what's the market for this?
01:48 Or cool, but I don't really understand it.
01:50 And so now the fact that we are here in this moment
01:53 and the market's fully developed for it,
01:54 people are like leaning in to AI,
01:57 and the fact that we kind of grew up on this old school,
02:00 if you will, style of discriminative AI,
02:04 which is really about like, taking the data
02:06 and to understand what it means,
02:07 and then blended that with the generative side,
02:09 which is really about producing something
02:11 that is typically done by a human,
02:13 is really how we crack the nut on getting that right blend
02:16 of AI to really deliver high quality results for what we do.
02:20 - That's really interesting that you saw this coming
02:23 a decade ago, but did you expect to see AI
02:26 where we are now then?
02:29 - I think that, I mean, it's obviously changing
02:32 every single day, and I have this reputation
02:35 in my own company of being like AI skeptic,
02:38 because I keep having the worst possible user experience
02:41 every time I ask ChatGPT to answer something for me.
02:44 So this is a running joke,
02:45 I'll screenshot it and send it to everyone.
02:46 So I'm constantly being surprised by what AI can
02:49 and can't do in this moment,
02:51 and I think that if we would circle back to 10 years ago,
02:54 and I forecasted what the future might look like,
02:57 I think there was maybe one or two things that I saw
02:59 that actually came to fruition,
03:00 and then everything else has just surprised me
03:02 and will continue to do so,
03:03 just given the pace of change in AI.
03:06 - AI has touched really every industry from education,
03:09 obviously now fashion, how has it impacted retail overall?
03:14 - Great question, so for our customers
03:16 who are predominantly like Fortune 500 brands,
03:21 they might be internet retailer top 500 type customers,
03:24 so they're big companies, they have lots of scale.
03:26 What we're doing for them is we're getting $89 million
03:29 in incremental revenue, because we're inspiring
03:31 their customer, the consumer more frequently,
03:34 which is causing brand loyalty to go up.
03:36 So we're actually seeing that customer come back
03:37 and buy again more frequently.
03:39 They're coming back for one of our top 500
03:42 internet retailer companies 8% more frequently.
03:44 So that repeat purchase effect
03:46 is building a more loyal customer.
03:48 And the impact of that really can't be understated
03:51 in an industry where people are sort of like,
03:54 you buy once from that brand and then never again.
03:56 So the financial impact is very real,
03:58 but then also the sort of, just what you can do with AI
04:03 when you don't have limitations of hours in the day,
04:05 or people getting tired, or all that stuff
04:08 just unlocks this whole world of possibility.
04:10 And in retail in particular, there's so much data.
04:13 Think about how many products, how many customers,
04:16 how many data points, products go in and out of stock,
04:18 everything changes.
04:19 That's a blessing and a curse when you talk about AI.
04:22 - FindMine is fashion focused.
04:25 And when you're talking about,
04:27 hey, maybe I wasn't the most fashionable person,
04:29 when you said that at the beginning,
04:31 that really resonated with me.
04:32 I don't know how to do a smokey eye either.
04:34 I don't have an eye to put everything together.
04:37 So the product resonates with a lot of women.
04:40 AI, as we know, is a very male dominated field.
04:44 So how did you get investors to come onto this?
04:47 And was it harder, especially as a woman,
04:50 going into a male dominated field?
04:52 - Great question.
04:53 I definitely feel like I had the push and pull
04:56 because fashion is female dominated.
04:59 And we do beauty and home,
05:01 and those tend to be spaces where there's just more women,
05:03 maybe not at the leadership levels
05:05 as much as they should be.
05:07 But then we have also the AI side,
05:08 and our AI technology is very, very deep
05:10 and very, very robust.
05:12 So I kind of had to play both sides of it.
05:14 And I think one thing that really helped me get investors
05:17 excited in this round in particular
05:19 is the length of history we have.
05:21 Because there's so many companies coming in
05:23 that are, you know, FindMine's AI first
05:25 and use case specific,
05:26 and I think that's a big differentiator.
05:28 There's so many companies coming in
05:29 trying to solve something in retail,
05:30 trying to solve styling,
05:31 and then bolting on AI on top of it.
05:34 It's just not differentiated.
05:35 And so if you're an investor putting money in,
05:37 it's like, well, anyone has access to these technologies,
05:39 how do I know you're going to be the one to succeed?
05:41 The fact that we have proprietary
05:42 and very deep data sets on what we do,
05:45 I think just put us kind of to the top of the list
05:47 and helped overcome some of that, you know,
05:49 there's just less people looking like me in this field.
05:53 >> Let's talk about that a little more,
05:55 because as we know,
05:56 AI has disrupted industry after industry
05:59 over the last year and a half.
06:00 So how does FindMine differentiate itself in this field
06:04 aside from the longevity?
06:06 >> Great question.
06:07 So the first thing is that
06:08 there's kind of these two camps within AI,
06:10 discriminative and generative that I mentioned before.
06:12 So the discriminative stuff is sort of more old school,
06:15 and then the generative is like kind of the new shiny object.
06:18 And we believe that you need both
06:20 in order to succeed in this particular pursuit,
06:22 which is styling.
06:23 The reason for that is,
06:24 if you think about generative AI
06:26 and the big players in the space,
06:27 Google, Amazon, Meta, et cetera,
06:30 you can ask it, hey, make me a style guide,
06:31 make me a lookbook, right?
06:33 And maybe you're Abercrombie.
06:35 And if you ask it that,
06:36 it might give you something that looks like American Eagle.
06:39 That's a really scary place for brands to be
06:41 'cause they want to differentiate from each other.
06:43 They don't want to look the same.
06:44 And so this generative AI
06:46 that is trained on sort of the wisdom of the crowds
06:48 collapses the differentiation between those brands.
06:51 And that's exactly what FindMine solves.
06:53 To solve that,
06:54 we have to use that discriminative side of things
06:57 to deeply understand brand.
06:58 We actually map brand in multidimensional space,
07:02 which is a crazy thing to kind of picture.
07:04 But being able to articulate what is brand to an AI
07:08 using computer speak, using vectorization,
07:11 is the way we're actually able to then tell generative AI,
07:13 make me a lookbook that looks like Brand X
07:15 and have it actually look precisely like Brand X.
07:18 - Can you dive a little deeper in there?
07:20 How are you distinguishing brand to brand?
07:23 Because when you think of Abercrombie and American Eagle,
07:26 they have similar looks.
07:27 So how are you breaking those down?
07:29 - Yeah, so as a consumer from the outside,
07:31 and we don't work with either of those,
07:32 I'm just using them as examples here.
07:34 You can see that they're similar in some ways,
07:37 and you maybe get a sense of how they might be different,
07:39 but maybe not.
07:39 And that's kind of where you stop thinking about it, right?
07:42 'Cause you're a consumer,
07:42 you just kind of move on with your day.
07:44 If you work for either one of those brands,
07:45 if you're the creative director of either of those brands,
07:48 they can rattle off a whole bunch of ways
07:50 that they're different.
07:50 But they also don't have the vocabulary
07:54 to describe in human language
07:56 exactly what is specific and unique about each of those.
07:59 So if they were to crack open OpenAI
08:02 or Stability or any of these kind of LLM technologies
08:06 and try to teach it what their brand is about,
08:08 what ends up happening is all the promise of time savings
08:11 that generative AI is here to tout,
08:14 you end up losing,
08:14 'cause you have to torture yourself into a pretzel
08:16 to try to explain to it what is it about your brand.
08:19 And that's really the thing
08:20 that FineMind has become uniquely good at,
08:22 is to very quickly get a sense of what it is to mean,
08:25 what does this brand mean, basically,
08:27 and then feed it into all these different models
08:29 for the different outputs that look high enough quality.
08:31 - Meta, as I'm sure you know,
08:33 is coming out with smart Ray-Bans
08:36 that provide outfit inspiration.
08:38 So how does FineMind differentiate itself
08:42 from those types of products?
08:43 - Yeah, so first of all, we're not competing with Meta.
08:46 That's sort of the wisdom of the crowd's styling response,
08:49 which has its place in time.
08:50 Wearing and getting dressed,
08:51 I just wanna know, quick fit check, is this okay?
08:54 Am I gonna get laughed at in the mom's group
08:58 that I go to if I wear this, right?
08:59 There's a very specific purpose for that.
09:01 And then there's also the leadership and expertise
09:04 of an individual brand, which builds that loyalty
09:06 and helps you be a better version of yourself.
09:08 And so I see, potentially,
09:10 when brands start flooding into that space,
09:13 trying to say, okay, Meta,
09:14 suggest my brand, suggest my brand.
09:16 There's a bidding environment or a retail media
09:19 or advertising ecosystem built around that recommendation
09:22 that you get in that lens.
09:24 That's where FineMind can come in,
09:25 because we can say, all right,
09:26 well, don't just push this specific product
09:30 for any old reason.
09:31 Push what's important to the brand
09:33 and style it in a way that is authentic
09:35 to the creative direction of that brand.
09:36 So you're getting the brand's point of view,
09:38 not just the wisdom of the crowd's point of view.
09:40 - Speaking of paradigm shifts,
09:41 we have seen a shift in the way people shop
09:44 with the rise of social media, particularly TikTok.
09:47 I know some of your clients include Lululemon
09:49 and Vineyard Vines.
09:51 They both have a large TikTok presence.
09:53 Are they nervous, apprehensive about a potential TikTok ban?
09:57 - I think a TikTok ban is a blessing and a curse
09:59 when you think about what brands are gonna feel about it.
10:02 On the one hand, obviously, there's tons of traffic
10:04 and sources of revenue that come from that channel.
10:08 But on the other hand, there's also a leak
10:10 from the brand's sort of core properties
10:12 to a place like TikTok,
10:13 because we talked about it earlier,
10:15 the scale issue of trying to inspire your customers
10:18 for every single product.
10:20 If I go to buy this blazer
10:22 and I don't see it styled in a lookbook
10:24 or I don't know all the different outfits to wear with it,
10:26 guess where I'm gonna go?
10:27 I'm gonna go to TikTok or Instagram
10:28 or one of these places and just put in
10:30 thick blazer, thick white blazer,
10:33 and just see what kind of inspiration I get.
10:35 And that's really an opportunity for the brand
10:37 to reclaim that expertise in that moment
10:40 and then sell the customer the jeans and the shoes
10:42 and everything that goes with it.
10:44 So that's kind of the blessing side of it.
10:46 And then I think all the brands that we work with,
10:48 they're all kind of thinking about user-generated content.
10:50 They're all thinking about influencers
10:52 and how they can capitalize on the kinds of consumers
10:55 who don't necessarily want the brand
10:57 to tell them exactly what to do.
10:58 'Cause I think all of us as consumers,
11:01 we want expertise from the brand sometimes,
11:03 we want it from our friends sometimes,
11:04 we want it from influencers sometimes,
11:06 we want it from non-celebrity people
11:09 who are just like us sometimes.
11:11 And so you kind of have to cover
11:12 all of those different bases
11:14 to give the customer everything that they need.
11:16 So I think it's gonna be both a blessing and a curse
11:18 if TikTok does get a ban.
11:20 - I know you have been embracing AI for a decade,
11:24 but do you think if a company is not embracing AI overall,
11:28 they're being left behind in this space right now?
11:31 - I think it's still early days.
11:32 I think there's plenty of time to get your feet wet.
11:35 I liken this to sort of the beginning of the internet
11:38 or the beginning of using computers for stuff, right?
11:42 Now, if you say, "We use computers to do banking."
11:45 It's like, well, duh, you and everybody else, right?
11:48 That's how it will be to say,
11:50 "We use AI to do X in the future."
11:52 Maybe three years, maybe five years, maybe 10 years,
11:54 we don't know how long, but it's coming.
11:56 It's just such early days now
11:57 that if you don't, like, I hate to see companies be like,
12:00 "What's our AI strategy?
12:01 "We need to figure out how to bring AI in."
12:04 That's the wrong way to approach it.
12:05 It's what are our problems?
12:06 What is the strategy we're trying to solve for our business,
12:09 for our industry, for our customer?
12:11 And if AI can support you in getting to that goal,
12:14 that's a great application for AI.
12:15 But AI for AI's sake tends to always be a fool's errand
12:18 of a lot of money and a lot of waste of time
12:20 and a lot of frustration.
12:22 - Do you think AI in your case promotes brand loyalty?
12:27 - Yeah, absolutely.
12:28 I mean, one of the things
12:29 that I started the company based upon
12:32 is when I had a really great interaction
12:33 with a store associate or a beauty counter expert
12:36 or someone who really oozed the brand expertise,
12:40 I just found myself going back to that more often.
12:43 I can feel that part of my brain
12:44 almost like clicking into gear,
12:45 like this is one I trust,
12:48 versus when you buy a product
12:51 and you don't wear it right away,
12:52 or it doesn't quite ever fit into your living room,
12:55 and you're just like,
12:56 every time you look at that chair,
12:57 you're like, "I just don't know."
12:59 You question yourself, you doubt yourself,
13:01 but you certainly doubt the brand.
13:02 So if you can make the customer feel confident
13:04 with their purchases
13:05 and feel like they're living their best life,
13:08 it 100% does drive brand loyalty,
13:09 and we see that in the data.
13:11 - And what do you say to people who are more bearish on AI
13:14 who think, the first thought that comes to mind
13:17 when they think of AI is,
13:17 "Hey, is this going to take over my job?
13:20 "Is this going to take over the world?"
13:22 What do you say to them?
13:23 - I think that's a fair concern,
13:25 and I absolutely believe that the guardrails
13:27 need to be in place for things like rights issues.
13:30 You know, the SAG Afterstrike was really all about that.
13:34 Training on people's creative pursuit
13:35 to then disintermediate them from the process.
13:37 I don't believe that that's,
13:39 I believe that people are right to be concerned about that.
13:41 In FineMind's particular case,
13:43 we have to have the seed from the people.
13:46 The people are necessary but not sufficient.
13:48 They have to create that beautiful window display
13:51 and those, like, what is this season all about?
13:53 But they only do it, again, remember,
13:54 for five or 10% of the product catalog.
13:56 No one has the time and resources
13:58 when you're a multi-million dollar, a billion dollar brand
14:02 to create beautiful editorial moments
14:04 for every single product you sell
14:06 and then go back and update them
14:07 when this product went out of stock
14:08 and this one's come in and now there's a new trend.
14:11 There's just, the people to do that don't exist today.
14:15 So there's no jobs to disintermediate.
14:17 And then what we do is we cast the people
14:18 who are doing that work for that five to 10%
14:21 in this very elevated light.
14:23 Like, they're the creative directors,
14:25 they're the Anna Wintour behind the whole operation
14:28 and then we're just executing it.
14:30 We're just taking a baton and running with that.
14:32 - So you're saying, essentially,
14:34 we're not stealing anyone's jobs,
14:35 we're making your life easier.
14:38 - Making their life easier
14:39 and almost elevating what they do.
14:41 One of our investors is this woman who is,
14:44 she's just been in retail and C-suite
14:46 and kind of senior VP roles her whole career.
14:49 And the way she described it was,
14:50 you give me time back to do the part of my job that I love,
14:54 which is the deeply creative work.
14:55 It's really fun to create one outfit,
14:57 it's really fun to design a furniture set,
14:59 but it's really boring and tedious
15:02 to do the 15,000th one, right?
15:04 You max out super fast.
15:06 So like, let's let them do what they do,
15:07 which is creative and fun.
15:09 And then we just replicate it at scale and everybody wins.
15:13 - Can you talk about who are your investors?
15:16 What do they look like?
15:17 What are their backgrounds?
15:18 - Yeah, great question.
15:19 So we have a whole kind of running the gamut of investors.
15:22 And so who led our series A
15:24 was a company called Greyhawk Capital.
15:27 The partner is this guy, Brian Smith,
15:29 who is just a very pedigreed kind of Silicon Valley investor,
15:33 SaaS software, not necessarily retail, not necessarily AI.
15:37 He's a generalist SaaS investor
15:39 and just very disciplined in what he does.
15:41 And he saw two things.
15:42 He saw kind of the market opportunity for AI
15:44 and then just the roster of customers we have
15:47 paying us real money to solve a real problem.
15:50 And so that's kind of one camp of investors
15:53 is like the Silicon Valley VC type.
15:56 But then we have angels like the one I mentioned before,
15:58 her name is Sarah Wallace.
15:59 She's the COO at Minted,
16:01 which is a home design and furnishings marketplace.
16:04 And she came from Gap
16:05 and then she came from Walmart prior to that.
16:07 So the retail folks, I think really get it.
16:09 And then the AI folks as well.
16:10 We have some investors who are just,
16:13 they've been doing this for more than a decade.
16:15 They've been doing this for 20, 30 years
16:17 and really have seen the opportunity that AI has opened up
16:20 for a really big financial return
16:23 in the application that we've chosen.
16:25 - So you, along with some of your investors
16:27 have years and years of experience in AI.
16:30 What do you think's missing
16:31 from the national conversation then when it comes to AI?
16:33 Because there is this debate of,
16:35 A, this is really helpful.
16:36 This will move our society at breakneck speed.
16:39 This is a paradigm shift.
16:40 There's the other debate of people more scared.
16:43 Is this going to steal my job like we talked about earlier?
16:46 - It's such a good conversation to have
16:48 for any paradigm shift.
16:50 I mean, we had to have it around computers.
16:52 We had to have it around the internet.
16:53 We had to have it around social media.
16:54 There's so many good things that social media have done
16:57 for humanity and there's so many dirty, bad, awful things
17:01 that it has done and you can't paint it all
17:03 with the same brush.
17:04 So I think it takes a very disciplined
17:07 and time-intensive approach to have these conversations
17:10 that a lot of times I think that people who are involved
17:13 in the conversation aren't willing to take the time
17:17 to have that, right?
17:17 I'm talking about in Congress.
17:19 I'm talking about kind of like think tank groups
17:22 and stuff that talk about AI.
17:24 We're missing a lot of what we should be thinking about
17:27 and we're spending too much time talking about the things
17:30 we shouldn't spend so much time talking about.
17:32 It's tricky.
17:33 - Michelle, I appreciate the conversation.
17:35 Thank you so much and congratulations.
17:37 - Thanks again.
17:38 (upbeat music)
17:40 (upbeat music)
17:43 [BLANK_AUDIO]

Recommended