Now more than ever before, digital marketers and retailers have a wealth of first-party data on consumers. However, providing an authentic and unique eCommerce experience for audiences continues to prove difficult. On this episode of The Voice of Retail, I’m joined by Frank Faricy, the CEO + Founder of XGen AI. We talk about the chronic pains of contemporary digital retail marketing and the challenges of effectively and efficiently targeting potential customers.
Welcome to the The Voice of Retail , I’m your host Michael LeBlanc, and this podcast is brought to you in conjunction with Retail Council of Canada.
Now more than ever before, digital marketers and retailers have a wealth of first-party data on consumers. Or at least they should - different podcast. However, providing an authentic and unique eCommerce experience for audiences continues to prove difficult.
On this episode of The Voice of Retail, I’m joined by Frank Faricy, the CEO + Founder of XGen AI. We talk about the chronic pains of contemporary digital retail marketing and the challenges of effectively and efficiently targeting potential customers.
Learn about the innovative workings of XGen AI and how Frank and his team had to reinvent the wheel to get the cutting-edge platform that they run today. Their AI drives autonomous and real-time product recommendations with the goal of providing a true 1:1 digital shopping experience that is as dynamic and unique as each individual customer.
Let’s explore the potential that lies at the intersection of retail and artificial intelligence.
Thanks for tuning into today’s episode of The Voice of Retail. Be sure to subscribe to the podcast so you don’t miss out on the latest episodes, industry news, and insights. If you enjoyed this episode please consider leaving a rating and review, as it really helps us grow so that we can continue getting amazing guests on the show.
I’m your host Michael LeBlanc, President of M.E. LeBlanc & Company, and if you’re looking for more content, or want to chat follow me on LinkedIn, or visit my website meleblanc.co!
Until next time, stay safe and have a great week!
Michael LeBlanc is the Founder & President of M.E. LeBlanc & Company Inc and a Senior Advisor to Retail Council of Canada as part of his advisory and consulting practice. He brings 25+ years of brand/retail/marketing & eCommerce leadership experience, and has been on the front lines of retail industry change for his entire career. Michael is the producer and host of a network of leading podcasts including Canada’s top retail industry podcast, The Voice of Retail, plus Global E-Commerce Tech Talks and The Food Professor with Dr. Sylvain Charlebois and the all new Conversations with CommerceNext podcast. You can learn more about Michael here or on LinkedIn.
Michael LeBlanc 00:04
Welcome to the Voice of Retail, I'm your host Michael Leblanc. This podcast is brought to you in conjunction with Retail Council of Canada.
Michael LeBlanc 00:10
Now more than ever before, digital marketers and retailers have a wealth of first-party data on consumers. Or at least they should - different podcast. However, providing an authentic and unique eCommerce experience for audiences continues to prove difficult.
Michael LeBlanc 00:23
On this episode of The Voice of Retail, I’m joined by Frank Faricy, the CEO + Founder of XGen AI. We talk about the chronic pains of contemporary digital retail marketing and the challenges of effectively and efficiently targeting potential customers.
Michael LeBlanc 00:37
Learn about the innovative workings of XGen AI and how Frank and his team had to reinvent the wheel to get the cutting-edge platform that they run today. Their AI drives autonomous and real-time product recommendations with the goal of providing a true 1:1 digital shopping experience that is as dynamic and unique as each individual customer.
Michael LeBlanc 00:56
Let's explore the potential that lies at the intersection of retail and artificial intelligence.
Frank Faricy 01:01
So, for us, we took it a step further by saying hey, you know, every time someone does something, we're gonna send that directly to our, our AI engine, and we're going to get the, the AI engines prediction in real time.
Michael LeBlanc 01:13
Let's listen in now. Frank, welcome to The Voice of Retail podcast. How are you doing this afternoon,
Frank Faricy 01:17
Michael, thanks. I really appreciate you having us on here and I'm doing great and yourself.
Michael LeBlanc 01:22
I'm very good. Thanks, and it's great to, great to meet you, so to speak virtually. We have a mutual friend Neil Weitzman, who introduced us, the great Neil Weitzman. So, thanks. Shout out to Neil and, and looking forward to learning all about your business. Now, where am I talking to you, where am I reaching you today?
Frank Faricy 01:39
Currently, I'm in Tampa. I just got back from Europe 48 hours ago.
Michael LeBlanc 01:43
All right, just getting, just getting accustomed, is that where your company's based, is that where you are based and talk about that a bit?
Frank Faricy 01:49
Yeah, no, absolutely. So, we're based in the United States. Obviously, we're a US company. We just, we have a lot of traction in Europe, specifically luxury. So, it's dragging out there a lot these days.
Michael LeBlanc 01:59
All right. Interesting. Interesting. Well, we'll get into all that. But let's say we've kind of jumped in, but let's, let's take a step back. Tell me about yourself, your professional journey and your role at XGen AI.
Frank Faricy 02:10
Yeah, perfect, great. So, I grew up in England, I have always been, you know, tinkering in computers my whole life, from building computers at a young age with my friends, you know, hacking, hacking together, overclocking machines to play video games, all kinds of fun stuff and that career path really took me into a hardware journey for many years in my life. Then kind of system architecture and design for, actually, live shows. I was in that for many years doing broadcast, like the NFL draft from a tech perspective. You know, some other cool, fun, exciting events,
Michael LeBlanc 02:47
Not without a net, that stuff I actually, you know, I spent a bunch of years at the Shopping Channel, which is like a QVC or HSN and people just don't, you know, generally, even in the in the tech world of broadcast world appreciate, you know, broadcasting for that long for that live. Wow, what's, there's a lot of technology that happens and it's, it's quite, it's quite impressive. So, that's great. That's a great background for it.
Frank Faricy 03:11
Yeah, it's, it's, you're 100% right. It's in high demand for sure. No room for error, obviously.
Michael LeBlanc 03:16
Yeah, yeah. That goes down. There's no take two, right.
Frank Faricy 03:18
Yeah, that's right. You're gone.
Michael LeBlanc 03:20
You're gone. You're off the air. The money's gone. You're gone, bye, bye. Anyways, I'm sorry, I derailed you there a little bit. All right. So, hardware and keep going, tell me how you got into what you're doing, yeah.
Frank Faricy 03:30
So, it's quite interesting. My sister actually, I have her to thank for my current path, she, she runs a beauty salon chain in Australia, and she decided to release her own product line, you know, manufacture these products, really high quality organic, and they took off in the salon line and so she started to, kind of, push on the e-commerce route and she was having all these difficulties selling on Amazon and online and various other platforms and I was going through a change in my life transition to a different location, I was like, hey, let me jump in, see if I can help and, you know, we manage to, we manage to, you know, increase the performance and, and revenue of those products, you know, 1000s of percents in about six months, it really kicked off pretty heavily and that, that really got me heavily involved in e-commerce and you know, retail performance and diving into all things associated to that and, and that was the genesis of XGen.
Michael LeBlanc 04:32
All right, well, though, and tell me all about XGen, what's the origin story, when it started, you give me some background, but tell me about origin story, its scope. You're based in the US. You're doing something fun and interesting with e-commerce. So, take, take us through.
Frank Faricy 04:46
Sure. So, XGen AI was, kind of, born out of this concept of, I had a very unique approach, Michael, to what I considered as, you know, an e-commerce user should you know the experience should be I feel like, so, much of it is focused around, like, crush selling the individuals, you know, you see it in like ad retargeting, right, like you drop them onto something, you see a product and it's like following you around across all your apps, you know what I mean?
Michael LeBlanc 05:11
Yeah. Relevant or not, right? Just based on the fact you may be in close proximity to a product or
Frank Faricy 05:17
Yeah.
Michael LeBlanc 05:17
whatever happened, right?
Frank Faricy 05:18
Exactly. It's kind of the mentality, like, if I'm a brand, it's like, you must see my product, I'm going to show it to you until you crack, you know,
Michael LeBlanc 05:25
Until you say, uncle, I'm going to, until you say, mercy, I'm going to just, you know.
Frank Faricy 05:29
Yeah, right. So, so I've never been a big fan of that kind of methodology and, to me, if you're going to sell a product, you need to serve as the customer, right, like, you really, you need to understand what their needs and wants are, you need to understand that to the core, and you need to be able to drive the right product to that person on the right experience that matches what they're looking for. You know, I'm a big fan of, you know, if you're going to sell something, well, obviously, you find people that are interested in it, but when you do, you really need to approach it from a perspective of like, hey, how can we service you, right.
Frank Faricy 06:04
As like the fundamental question and mixed with the concepts with, kind of, that fundamental hypotheses from day one, and blending it with this concept that I noticed that in high volume ecommerce, meaning lots of traffic, this, the tiniest changes would make specific ones would make huge impact on the performance of the sales. So, you know, taking that kind of philosophy, I ran forward, and I started to look at, you know, the subject of personalization, which, you know, was obviously, attempting to tailor the experience to a group of individuals, you know, Geo based, location based, whatever.
Michael LeBlanc 06:44
At least, at least try and gang tackle them, right. If you can't do one to one, right?
Frank Faricy 06:47
Yeah, yeah. It's like, you know, hey, fast forward, you know, a couple of decades websites were like, you know, one person for, you know, what one experience every single person, right and I have a funny analogy, I have many analogies, but this one's kind of funny is, you know, if, Michael, if you, if you launch a shoe brand, right, and you're sitting in a stadium, and you have like a million people looking at you at the center of the stadium, you get out there, you hold up a couple of shoes and start yelling at everyone buy this product, right, you're going to capture a certain portion of that audience, obviously, but ultimately, the potential you're missing out on is massive and with the advent of personalization allowed, it allowed you to go well, first of all, semi replicate yourself, and then break the stadium into like seating groups, like this is your, you know, age range and we're going to put someone that kind of understands a little bit about these people in front of them and pitch them a different product, right and it really,
Michael LeBlanc 07:48
These are your red shoe fans, these are high heel fans, these are your sports fans, at least, at least some way to, what we would call segment or, or divide them up, right?
Frank Faricy 07:56
Exactly, you've got, you've got basic segmentation rolling, you've got some, kind of, user data coming in, that you can use to trigger responses, you know, hey, they're, they're, they're, these are Japan users, they're, you know, they have purcha-, buying certain products and certain, certain price points, certain colors, hey, we can kind of craft a, semi crafted journey for them, you know what I mean, but the fundamental problem with this is it still bugged me, because if you you look at it, it's like, well hold up, if I'm sitting in one of those audience groups, why am I being treated the same as 10,000 people, right. So, it's kind of the same problem as, you know, the stadium just honestly made it a little bit more complicated and kind of you 10x your, your complication that you've, you know, delivered a little bit more increase, but you're it starts, it's, it's a dark path, it gets really complicated, you start to micro and macro segment. It does not happen; you know what I mean.
Michael LeBlanc 08:50
Yeah, yeah, it can get very confusing and expensive and you, you, as you say, you wind up down a lot of rabbit holes, you get a 10x increase in complexity with a 1x increase in sales, that math doesn't work, right.
Frank Faricy 09:02
Exactly and if you and you know, kind of fast forwarding into like real world application, if you look at e-commerce teams from like big corporations down to like small startups, they don't have time and I mean, they, they are like
Michael LeBlanc 09:14
Yeah.
Frank Faricy 09:14
Hey, you know, sales is everything to them. They're either under immense pressure financially or from their superiors and you can't expect them to spend three months setting up a data strategy and segmentation strategy and building the logic behind who sees what product and experience, it's just, it's not, you know, it's not workable, besides the fact that, that's, you know, unfortunately, you know, technology has advanced far beyond that point of this in this day and age, right.
Michael LeBlanc 09:39
Yeah, yeah.
Frank Faricy 09:40
So, the advent of XGen AI was really, you know, we took this we took that problem and we said, what was the ultimate, what's like the Ne Plus Ultra in this solution, right and we came up with a concept and I'm not going to lie, it was one of the most challenging things I've ever done to build that system, right it, you know, the key points were how do you treat an individual, like an individual, how do you look at what they're doing, understand their needs and wants, look at their journey and then drive not just a unique product to them, but a unique experience, like, if you were to walk into a luxury store, you know, the ultimate luxury experience would be that store, you know, digitally adapts to you and the sales rep is looking at everything you're doing understand has 30 years of sales knowledge to pull upon, and, you know, you walk out of there with a bunch of stuff and you're like, that's actually what I came to look for, I feel happy, I'm good. I'm not being crushed, sold, etc.
Michael LeBlanc 10:37
So, there's a lot going on there. Right, three simple questions, but they're very big questions right they're and then, you know, the only, the only question that you'd add to that as a fourth bonus question is the meaning of life. I mean, if you put those four together, you really got, you know, some pretty powerful questions, quite a bit of a gnarly challenge, right?
Frank Faricy 10:55
Yeah, that's exactly right and, you know, that, that, that category breaks down into several key problems, you know, following on from that, like, well, what data do you need to do that, right and in order to do that, you, you know, people are used to reacting to data, meaning like, if they pick up a red shoe that equals in this segment, a blue belt, and that's the logic that's fixed for the next month, right, but what people don't understand is that a shoppers mind doesn't work that way. It's changing instantaneously, like, it's being, it's being affected by the very thing that you're showing them, the experience. So, you know, the, the data methodology, meaning the value you can pull out of the customer data is that you're not even tapping into like, a decimal percent of its potential. So, again, it was like, the amount of value you can derive from that is instrumental, but the problem is, is how do you do that, right, so the old
Michael LeBlanc 11:57
How do you, How do you, how do you replicate 20 years of sales experience in
Frank Faricy 12:03
Yeah.
Michael LeBlanc 12:03
machine, basically, right?
Frank Faricy 12:05
I mean, and if you look at this, like this hypothetical, super sophisticated sales associate and a luxury store. What are they ultimately doing they're predicting you, right, they're, they're like, you know, they pick up, you pick up a red handbag and, you know, they're, it's a high price point and they're kind of thinking, well, if he, if he puts it down, I'm going to go to blog next, you know what I mean?
Michael LeBlanc 12:28
Right, right.
Frank Faricy 12:29
So it was, it was a question of how to replicate that in a digital environment, which is incredibly cold and unforgiving, right?
Michael LeBlanc 12:38
Very, very efficient, but not much. You know, this is the challenge of e-commerce, right, it's, it's very efficient, but it's not exactly a wonderful shopping experience all the time, right?
Frank Faricy 12:47
Yeah, I mean, I think it's, that it, kind of, follows along the same path as technology, obviously, it's like, you can build one store and invite the world to see it, which is, is seems stupid and but in reality, that is the power of e-commerce, right, but the, you know, the, the unforgiving nature of, you know, if I have 10,000 SKUs up, and the average, you know, customer time on site, for me is like one and a half minutes, the probability of me finding the right product is like, is not even, it's not even close to being accomplishable, right.
Frank Faricy 13:22
So, you know, in product discovery, it's like, well, we break it down to categories, we show, you know, bestsellers, and all this stuff, but it's all, it's all honestly just patchwork to the entire, it's like going into surgery with a baseball bat the way I call it, right. So, so the scalpel in that scenario became, well, how do we take each customer leverage, you know, years of data about this individual, look at exactly what they're doing in a real time environment and adapt the experience only to that customer and exactly what they want to see, but predicting what their next move is going to be and that's where the, the challenge, well, that is the challenge we decided to tackle. It was, it was a painful period of my life. Michael, I'm not gonna lie. You know, you
Michael LeBlanc 14:14
You probably should have started with a meaning of life and work backwards; it might have been faster by the sounds of it.
Frank Faricy 14:18
Yes.
Frank Faricy 14:19
Yeah, I mean, you know, in the beginning as a startup, you know, we literally started in a shed, we were, I remember, we used to sit on like milk crates, and like yoga balls, right, it was ridiculous and we were trying to put this thing together with, like, zero money and you know, that so many times it got to a point where it's just, like, we, the technology. No, the, like the backbone of technology, like the, the services and the servers and the processing power behind what we do is just not ready for this, right, but we found a way we had to reinvent the wheel at every turn and we came up with like a minimum viable product, an MVP that was like, hey, let's test it against, like, existing methodologies and the results were, you know, shockingly profound and that was, that we took it from there.
Michael LeBlanc 15:10
All right, well, tell me about it. Let's, let's pull that apart a little bit. So, you know, lots of AI, folks I talked to say they couldn't do what they do without, kind of, a cloud based horsepower like servers. So, you're probably using that whether it's Google or AWS, are using one of those two services, are you using.
Frank Faricy 15:28
AWS, yes.
Michael LeBlanc 15:29
You're, you're AWS. Okay. So, you're using the power of AWS to, to run this thing. Well, let's talk about this thing, this artificial intelligence, and, you know, it opens up this great conversation about what exactly is artificial intelligence versus clever programming versus learning,
Frank Faricy 15:47
Right.
Michael LeBlanc 15:47
And then I got a bunch of follow up questions but tell me exactly what your product does. So, I'm a merchant and what you've been describing is like, yep, yep, yep, ticking the box. I got those problems and I'd love to do what I do better. How and in what way does your product help me?
Frank Faricy 16:02
Yeah, so in a nutshell, we take the customer data. It's first, first, first party information, right, so we're a first party system, we take, we take the first party data, we leverage that in real time to predict what set of variables are they most likely to engage with, right.
Michael LeBlanc 16:18
So, for, before we go too far, just make sure we take the audience with us, unpack first party data?
Frank Faricy 16:25
Oh, yeah, definitely. So yeah, well, obviously, you know, a customer owns a site, you know, customer, you know, sorry, a client owns a site, shoe brand. Customers come to their site, they get the authorization directly from the customer to, you know, the cookie opt in opt out a banner everyone sees, to use that data to better service them, right, so that's first party versus third party, which is leverage, you know, from other companies, like, hey, I'm gonna pull this data from Facebook users in order to advertised to them, etc, etc.
Michael LeBlanc 16:54
So, this is actual customer behavior on your website. That's what, your, everything, the platform of data, that's what it's all predicated on. Cool, cool.
Frank Faricy 17:01
Yes, exactly and we are huge proponents of customer privacy. You know, so, so, that, that was, I mean it is a big challenge in the beginning to say, hey, how do you leverage just the customer data on site to gain these results and,
Michael LeBlanc 17:14
Yeah, yeah.
Frank Faricy 17:14
It's not, it's not an easy challenge.
Michael LeBlanc 17:16
Well, and the risk, I guess, is, it gives you a slightly, or sometimes more than slightly, incomplete picture from which you develop profiles, right?
Frank Faricy 17:26
Yes. Yeah, exactly and, you know, I mean, AI is moving at such a fast pace, right and it's like, the challenges behind that are immense and, and when you start to get into, like, cutting edge, like R&D, and saying, like, hey, we're going to look at every single relationship of this customer, all the elements about the product, color, the size, all this stuff, and look at all these data points, and drive an accurate prediction is, is already alone, like, quite complicated in a SaaS environment, like, how do you replicate that for every single customer and make it effective.
Michael LeBlanc 18:01
Yeah.
Frank Faricy 18:01
But on top of that, we were like, we took it a step further, we're like, we're gonna do I mean, you could do that on a weekly basis, Michael, right, you could, like, cache the perform, the results to that prediction weekly for the customer. So, their experience would update, you know, it would look like real time, but it's not the intelligence, you know, simply put, the intelligence is not updating in real time, right. So, so for us, we took it a step further by saying, hey, you know, every time someone does something, we're gonna send that directly to our, our AI engine, and we're gonna get the, the AI engines prediction real time directly out of that, and send it back to them all within like, you know, under 100 milliseconds, right.
Frank Faricy 18:41
So that, that was a large challenging component to the system in the beginning, ultimately, we can take any set of variables like we can take, okay, so variables are, hey, you have a bunch of products, right or hey you have a bunch of content, or ou, we have all these different, like, sales banners, or whatever variables you're dealing with, that you're trying to test online, right and we, you know, our system enables our customers to basically just optimize the ideal set of variables to that customer multiple times a second.
Frank Faricy 19:14
And if you really look at e-commerce, you know, everyone's used to AB testing, where it's like, hey, we're going to take a certain portion of the group, make a change, see what impact it made, you know, wait for two, three weeks for it to hit statistical relevance and say, great, this is, this is the change, it's going to go static to all of our customers, but if you really look at what we're doing, we're accomplishing that in real time for every single customer, right.
Frank Faricy 19:34
So, the way we look at it is you are optimizing that experience at absolute maximum to every single customer. So, our concept of the future is that eComm teams and branding teams can focus simply on the products, you know, having a better product looking at product metrics, building better branding, you know, getting involved in merchandising, and, and more content for the site and letting our engine optimize it individually.
Michael LeBlanc 20:04
You know, and I've heard similar things about this and it's not easy to do, that reorientation around metadata that fuels engines like yours, right, predictive engines like yours. So, now, okay, so I'm a, I'm visiting a retailer site and your engine is working in the background, what, what happens, do I get taken to different places, do I get different messages, tell me the next step once, once you your engine is decided that I'm going to be served certain things are taken in certain places, what happens then?
Frank Faricy 20:33
Yeah, definitely. So, you start interacting with a site, the what's happening is, is between our team and the customer, they have enabled locations on the site that are going to adapt to that customer, right. So, that could, that could come in the form of, you know, your atypical recommendation carousel, or it could be, like, literally a dedicated, you know, wardrobe, your wardrobe page, or your products for you page that they can click at any time, right, it's just, like, completely designed for that.
Frank Faricy 21:04
So, as they move throughout the site, this experience changes, you know, they can go back to the same URL, and it's going to look completely different. So, you know, if you look at a customer's journey, and like a sequential series of steps, every, every, kind of, step they take, the engine is crafting those specific locations on site to match exactly their needs, you know, ideally, if you, if we had it our way, we would, we would take a site and you know, make a lot of it dynamic, not just certain locations, right and we are working with some customers that do have, shown for wanting to go that deep, which is very exciting to us because the more you deploy, the more effective the site lift gets.
Michael LeBlanc 21:45
So, your, let's talk about the, let's talk tech for a couple of minutes. So, lots of retailers have sites, you mentioned workers and luxury retailers, for example, in Europe, maybe they're on a Shopify site, maybe they're on Salesforce, whatever. Are you, platform agnostic are you, do you sit on top, talk about that.
Frank Faricy 22:02
Yeah, that's great. So, yeah, we're definitely platform agnostic. We do have, you know, we built plugins for the top five, right, to make it simple, but ultimately, yes, we were on invasive, it's super easy to push in we do, we do treat integrations at the enterprise level, uniquely because everyone has different challenges, right, but the fundamental, the fundamental concept for us is that, from a business perspective, integration needs to be literally as rapid as possible, right. So, if we give you our products, there's a bit of distrust in the industry to tell you the honest truth,
Michael LeBlanc 22:37
The one day one line of code sales pitch.
Frank Faricy 22:39
Yeah, yeah.
Michael LeBlanc 22:40
Been pitched a few times, like, it's one line of code, it'll be in, it'll be done by the end of the day.
Frank Faricy 22:44
And then comes the storm, right, so, but we have, we have spent a significant portion of our lives building systems that can truly integrate lightning fast. In fact, we, we, we stand by our word so strong, that we actually will give the integration to the customer for free on a free trial at the enterprise level. So, hey, use the system for four weeks, we'll, you know, if it takes us a long time to integrate, it's just costing us, it's not costing you, right, so, so we're going to go to the customer and say, look, if we're going to run a test on the system on an AB test, so, you're going to take four weeks, or six weeks, or eight weeks or whatever, and you know, your time commitment through approval of the deployment through integration and setup, we're going to ask for three to five hours maximum of your time throughout this entire time. So that's, that's our goal of what we shoot for with every customer.
Michael LeBlanc 23:37
So, it's a it's a pretty light integration, you sit on top of and work with all the big platforms or major platforms, I'm sure you'd have those details in your site. So, great and you know, one of the challenges I had in the Stone Age of trying to do this, I was in campaign management and CRM, gifting, gifting would always, you know, as we started, you know, we had a great loyalty program and, you know, we had millions of members in it, and we'd get lots of data and two things would throw us off, actually, one, one would be gifting.
Michael LeBlanc 24:07
In other words, I'm not looking for myself. So, the recommendations were always slightly off, sometimes really off and the other thing was understanding, in our particular case, the difference between parents and grandparents, and we figured out a hack for that. The hack was grandparents don't buy diapers for their kids, right.
Frank Faricy 24:27
Right.
Michael LeBlanc 24:27
They go in and they buy other things. So, we figured out a, you know, just get, sitting around a table and figure that out. How do you think about those things, like, you know, the noise in the data, in other words, things that look like one thing, but are actually another, how do you, how does your engine start to think through that horse, you know, use that whole fantastic horsepower and think through those challenges?
Frank Faricy 24:45
Yeah. So, that's a great point and it's funny because you hear about challenges like this in various shapes and forms like all day, every day. You know, back, back several years, you do, you did have to respond to those with triggers, right, like, if under this circumstance that happens, like smart programming, then then, then do this, right, but ultimately, the, the way we approached it is we built models, meaning AI algorithms in multiple forms that are designed, you know, and I'll give you a caveat, there, there has to be a minimum critical mass of, of monthly customer data, right, you can't do this, you know, for the smaller customers, we have to leverage more simplistic models.
Frank Faricy 25:32
And that's one of the benefits of the XGen system is it's going to maximize the potential matter how big your data set is, right, but for the bigger customers, we're running, like hyper sophisticated models that will literally look at every possible patent in that data that is imaginable. So, if you think of a scenario that would apply, like, hey, this product works, when you show it to this type of audience under this circumstance, that's, that, that is a, that is not a true statement because it is going to work for a certain portion of that audience, right. So, so the response to that is to say, first of all, make that decision at the individual level and second of all, you know, if, if, if the model is doing its job, which it does, it will it will visualize that pattern, if that makes sense, like, that, the, the importance of that decision and, and the decision process of that, what you just mentioned, will be taken into account in the models decision process.
Michael LeBlanc 26:33
So, does it, does it pull together all the behaviors from all the clients, of course, not sharing any data anonymously, and kind of help push out those edge cases, so that it just gets better, is that, is that, kind of, that, that gang tackling again or the entourage effect, is that, is that happening?
Frank Faricy 26:49
Yeah, I mean, it's, it's, it's, we have a very unique, confidential approach to doing this in the backend, but ultimately, like, the our key is leveraging a massive amount of customer data points, right, on the sites and then comparing those to everything we can possibly find about the product, right and then when you start to look at stuff like time of year, time of day, and all the other variables that go into this, you can start to see, like, we've, we've seen patterns that just blew customer's mind, it's like, hey, by the way, this is a decision process with the engine, here's a quick summary of it and they're like, wait, we've never even seen this before, yet it is occurring, right.
Michael LeBlanc 27:28
Right.
Frank Faricy 27:28
And the engine, capable of extracting that at each user's level and showing the products it just, it generates instrumental lift. I mean, we have you know, we're, our batting average is anywhere from like 8 to 25%, incremental, AB tested revenue with month over month, right. This is like, empirical, like, undeniable revenue lift and, you know, that's instrumental, right and this is on an optimized experience. This is not some, you know, crappy, you know, Shopify quick build, that's going to require, like, months of, you know, optimization, right.
Michael LeBlanc 28:02
Right, right. Now, you, you've, last question for you, you've been talking to a lot of retailers, sounds like around the world, and what, what do the retailers, you know, what do they get wrong in those meetings, when they talk, think about AI and they think of your, kind of, predictive power and your product, what are the, what are the, kind of, myths or things that you go, hey, that's, that actually used to be the case, but isn't the case or that was never the case or that might be the case in 20 years, what are the, kind of, one or two things that come up most?
Frank Faricy 28:30
Yeah, I know, that, I love this question, right, because, you know, I deal with a lot and I think that, I think first and foremost is, you know, I can understand a non-understanding of what's actually happening. It is, it is complicated, right, like, there's so many things to understand, Michael, you get it, but there's so much that goes into understanding what's actually occurring, right, I also
Michael LeBlanc 28:53
If all the merchants understood they wouldn't need, they wouldn't need great vendors like you, right, I mean, they can't be great at everything. This is where this is the wonder, wonderfulness of, you know, people like yourself who really dig in and, and figure these things out, right?
Frank Faricy 29:04
Yeah, exactly and when you know, when you talk to these guys, I'll biggest challenge the beginning was like actually connecting the dots between the technology and what's going on like, and we had a big issue because we would tell people personalization, but that word has a bad taste to some customers like, oh, we tried X platform, and it didn't work for us, right.
Michael LeBlanc 29:25
Yeah, yeah.
Frank Faricy 29:26
So, ultimately, and, and on top of that, there's a lot of non-transparent, I'll put this delicately, non-transparency in the industry, specifically in SaaS, which I'm talking against my own industry, which I shouldn't do, but there's a lot of like, distrust about performance in this industry and it's because of the methodologies. We just published an article on Forbes about this a couple of weeks ago, but the way that performance is gauged, is not technically trustworthy in multiple circumstances, right.
Frank Faricy 30:01
So, we often have to educate the customer on the different testing methodologies and like, hey, you know, using that way is not really showing you what's actually occurring, right, so we need to, we have to educate people and you know, the first thing we'll tell the customer is, like, hey, we're going to run an empirical AB test to reach statistical events, we'll validate that on a third party system for you and, you know, ultimately, we should not be working together if we can't generate the lift, you're spending money not for feature, but for the results of generating more revenue, right, no matter how anyone cuts it about, you know, feature rich and less than that. It's not the point, the point is eComm teams need performance and that's the bottom line.
Michael LeBlanc 30:41
Yeah. And that, that search for incremental revenue versus, you know, it's this, it's this X, Y axis, how much effort and costs versus incremental lift
Frank Faricy 30:49
Exactly.
Michael LeBlanc 30:49
So, you know, this is kind of what it's all about, well, this has been great. It's a great introduction. Now, we've talked about a lot of stuff. You talked about articles in Forbes, you talked about, imagine you've got different case studies and all kinds of different things. Where can people go to learn more?
Frank Faricy 31:03
xgen.ai, pretty simple.
Michael LeBlanc 31:06
Very simple and you're, you're a LinkedIn guy, or what's the best way to get a hold of you?
Frank Faricy 31:10
I am. I'm a LinkedIn fanatic, Frank Faricy.
Michael LeBlanc 31:14
All right, very good. Well, listen, Frank, thanks so much for joining me on The Voice of Retail, great conversation. I wish you continued success and by the sounds of it, safe travels as well and once again, thanks for joining me on the podcast.
Frank Faricy 31:26
Michael, a real pleasure and it's always awesome to talk to someone knowledgeable in the space. I really enjoyed it. Thank you.
Michael LeBlanc 31:32
Thanks for tuning into today’s episode of The Voice of Retail. Be sure to subscribe to the podcast so you don’t miss out on the latest episodes, industry news, and insights. If you enjoyed this episode, please consider leaving a rating and review, as it really helps us grow so that we can continue getting amazing guests on the show.
Michael LeBlanc 31:50
I’m your host Michael LeBlanc, President of M.E. LeBlanc & Company, and if you’re looking for more content, or want to chat follow me on LinkedIn, or visit my website meleblanc.co!
Michael LeBlanc 32:00
Until next time, stay safe. Have a great week!
SUMMARY KEYWORDS
customer, product, data, ai, site, building, challenge, engine, commerce, people, michael, retail, experience, platform, talk, sales, performance, question, ultimately, understand