#393 - Combating Fake Reviews in the World of E-Commerce with Ming Ooi

#393 - Combating Fake Reviews in the World of E-Commerce with Ming Ooi Speaker 1: Welcome to episode 393 of the AM-PM Podcast. My guest this week is Ming Ooi. Ming is the co-founder of Fakespot.com, a site that analyzes Amazon reviews to determine what's fake and what's not. And there's some pretty interesting stuff that he's gonna reveal, including some things that Amazon's doing when it comes to reviews that might not make you too happy. It's gonna be a great episode full of some fascinating information when it comes to reviews. Enjoy this episode with Ming. Unknown Speaker: Welcome to the AM-PM Podcast. Welcome to the AM-PM Podcast, where we explore opportunities in e-commerce. We dream big and we discover what's working right now. Plus, this is the podcast where money never sleeps. Working around the clock in the AM and the PM. Are you ready for today's episode? I said, are you ready? Let's do this. Here's your host, Kevin King. Speaker 1: Ming Wei, welcome to the AM PM podcast, man. I'm glad that glad to have you here today. Speaker 2: No, thank you very much. Thanks for having me. This is going to be a real treat. I hope I think. Speaker 1: I think so too. I think we might open a few eyes today. It might be some cool, interesting stories that people have always wondered about, but nobody really ever talks about. So I think that's going to be pretty cool. Now, you're Singaporean, Chinese Singaporean? What are you? Speaker 2: I grew up between Singapore and Oklahoma. Speaker 1: Oh, that's a big difference. Unknown Speaker: Also Oklahoma. Speaker 2: Shout out to the town. Speaker 1: In Singapore, that's quite a big difference. I think, like you mentioned earlier, you know Abe. Abe was on the podcast a couple months ago. Abe's one of the top guys in the PPC, runs a PPC agency. If you haven't heard his podcast, go back and listen to Abe Chamalee's podcast here on AM-PM Podcast. You can find it. It's a really good episode. But Abe told me this interesting story when I met you at Rich Goldstein's party, I think, suite party in the Barbra Streisand suite during Prosper. And Abe had told me this story about you, about how he had a little office for his company and you had a little office right next to it. And one day you guys bumped into each other and you're like, hey, what do you do? Hey, what do you do? And it just like, y'all became friends because there was some, you were doing two different, radically different things, but in, in this e-commerce industry. And that was interesting. And so you were the co-founder, was it co-founder or one of the early guys? Co-founder, yeah. Yeah, co-founder for Fakespot. Speaker 2: Yes. Speaker 1: So Fakespot.com, for those of you who don't know, started what, about 2014, 2015, somewhere around in there? Speaker 2: Yes, 2015 it came to us. And then 2016 is the official, I think January 2016, we incorporated, you know, Formed the LLC to take it official by 2015. So what was the concept when you started it? Speaker 1: What started it? Were you and your co-founders searching for stuff on Amazon or somewhere like these reviews just aren't real? Speaker 2: I had a tech partner and I that we incubated a bunch of different startups at the time and so I was really working on two different startups of my own and my tech partner Jeremy Petuto, Jeremy said, Hey, I have somebody like late 2015 or somewhere middle 2015. He said, Hey, I have somebody that is working with me. that brought this to my attention. And it's a review authentication website that he created, but he doesn't know what to do with it. It's interesting. And he knew I came from the retail space. So before I get to Fakespot, in a prior career, I had ran a kids division of a consumer goods product company. And that product company that claimed the fame is we invented the memory foam bath mat, which at one point was ubiquitous. And it could be still ubiquitous in every American household, every household across the country. And I ran the kids division of that company. So, you know, we had worked with Amazon, and this was pre-Amazon market days where you, you know, had to go see the buyers and you do all that stuff. But so I had a fair amount of experience in retail and Amazon. And so when they brought it to me, they were like, okay, what do we do with this? And it was a great idea. I was like, okay, reviews. And the thing is, we had sort of arrived at the same problem from different angles. My wife had bought Beauty products that were supposed to be from Korea and they were starting to be shipped from Ukraine and different parts of the world and we couldn't figure out what was going on and why things are like this and we couldn't figure out the reviews said something different and and Saud, the founder, had the same thing. He had the same experience of he had bought all these supplements that he thought was supposed to be like this because he was into health and he's really into health and fitness and he couldn't figure out why the reviews and Saud To put it mildly is a prodigy in terms of program, to put it mildly. In the circles of people who know, this guy is a bona fide genius in terms of programming and coding. And he managed to create this system of authenticating reviews and he just put it out there. And just with that, when it came to me, he had like 10,000 users maybe with no marketing, no advertising, just... And Jeremy and Sawood came to me and was like, okay, you understand retail, you understand tech. What do we do with this? And we spent a little bit of time tinkering, figuring out and then finally we thought, okay, there might be a plan here. I think there's a plan of what we can do with this. And then so we incorporated a company and the idea was like, we will work together on it. He focused on the tech side, I would focus on the business side and Jeremy sort of also helps, I would focus on the tech side. And that's kind of how we got started. Speaker 1: So you guys were frustrated with reading reviews on Amazon primarily that just were fake, obviously clearly fake. So he developed, was he using early AI back then or he just developed some sort of algorithm where he could actually determine based on several parameters that this is most likely not real or? Speaker 2: When I finally learned how the whole thing was done, I was gobsmacked at how he had managed to pull this off and it's It's to say that it's like a model. Algorithm is a simple thing that there are elements of AI in it. There are elements of machine learning. You know, it goes so deep that it was crazy how somebody could see that and think of it and then execute it. So I'll give you a rudimentary kind of thing of how it works, right? Number one, Amazon has a firewall that people may or may not realize. If you ping the website too much, it triggers a whole bunch of different things on it because they don't want competitors coming to scrape all their information about like their pricing and stuff like that. So Amazon has a near impenetrable firewall. And I say near impenetrable because Saud was able to get it to ping Amazon You know when you're averaging, you know, 10-20 million users at any given time, you know You're pinging Amazon a lot and just to get past that firewall is an art form in of itself so that was number one which surprised me that he could do it and then the second bit of it is that we well I say we fake spot so most most reviewed things will look at the reviews and then they'll do a few criterias like what Fakespot did was take all the reviews that you could see apply and then there were a whole bunch of different criterias behind it which I'll get to later and then they would go to every reviewer And then go grab the reviews from every reviewer that was available. So now all of a sudden you have multiple data sets. You have one data set of all the reviews for this product. And then you have another data set of all the reviewers that left reviews. And then they were able to... Speaker 1: Just to be clear, that's so if I had, it's all the reviews for my, for one ASIN. And then whatever the people wrote, whether it's Tom, Jerry, Sherry, Mary, whatever, what they wrote. And then it also for Mary, it went and looked at all her individual reviews for everything that she wrote on every product she wrote, everything that Jerry wrote, on everything that he wrote. Okay, so I just wanna make sure that people are clear with that. Speaker 2: And then you roll the whole thing up and then you put it through every different criteria and you can catch cadence, you can catch cut and paste, you can catch if they reviewed the same thing multiple times. If they're like, you can catch all that. Like now you can imagine how deep this goes. Speaker 1: That's big. That's a lot. Speaker 2: All that back in less than a minute. Speaker 1: In real time. Speaker 2: Yeah. Speaker 1: So this wasn't scraped data that was stored somewhere. This was like, if I go there and I'm punching in, I want to know if this right now, this product, product B, whatever ASIN it is, it's like calculating it almost in real time. It's not like score that we scraped this last week or two weeks ago or something like that. Speaker 2: So now what happens is like once you scrape all that stuff, you're able to put it somewhere, you're able to store it out and then you can pick, okay, this was X amount point in time and then we'll only pick this X amount point in time new and that's why it gets better and better because you don't have to run the same thing, you can just add it incrementally but there are new products all the time, there's new stuff and so you're always constantly doing the new stuff all the time, like all the time. And it goes that deep every single time. Speaker 1: So when you launched this, was it mostly aimed at consumers, right? So the consumers that were shopping on Amazon, you were like, hey, install this extension. I don't remember when y'all eventually had an extension. I don't know if you had an extension at the first. Speaker 2: No, we did it the first couple of years, no. Speaker 1: So at the beginning, they had to actually go to Fakespot.com, cut and paste a URL or type it in or whatever. Down the road, you actually had a Chrome extension, I think, that actually would just show up. It made it a lot easier and give it a little rating. And y'all did, what was it, like A through F or something like that? Speaker 2: Yes, so that was, I'm going to take credit for that one. That was one of my things that when we first started, people will remember this, we started as a percentage thing. It was like, oh, this is 80% reliable, whatever it was. And my thing to Saoud and given my consumer goods background and a lot of training working with Walmart was you got to simplify it for the masses. It has to be the lowest common denominator. And I told Saoud, I was like, you got to change it to A to F because we Americans understand A to F. You give me percentages and all that, things like that, I don't understand. But if you tell me this is an A review, these products, these reviews are A, I'll know what that means, I know what B means, I know what C means, et cetera, et cetera. And so that has been one of the things that we changed. Almost immediately we changed that, I think. Speaker 1: How did the word go? Was it just viral? Cause y'all weren't really, were you doing any advertising or anything? Or was it just like some social media posts and just one person telling the next person, telling the next person? Speaker 2: So no money, mostly everything is through social media. So Saoud wrote a script and we were good enough where like he wrote a script where every time somebody ran something, it would send out in either, not so much Facebook, but it was sent out on Twitter like, hey, this person did this review and stuff and we're able to tag it that way. So if people were looking online for reviews and stuff, somehow usually they would find us one way or the other. So there's a lot of social media stuff. Speaker 1: Wait, wait, wait. So that's a cool thing. So if I, if Kevin, if I go and I typed in on Fakespot, X, Y, Z, ASIN, it would give me back and say, okay, this is a B. And then somehow he took that and then posted that to your, your Twitter channel and. Speaker 2: Fakespot Twitter channel, just send it out. Speaker 1: Fakespot Twitter channel and put some hashtags or whatever. If he knew who I was or whatever, hashtag it. Speaker 2: On the product or the brand, whatever it needs to be and automatically hashtag it by taking the same tags from the product title. And that's how people found us because when they're doing searches, they were like, oh, this came up, let me double check this. Speaker 1: Oh, that's brilliant. That's really smart. Yeah. Speaker 2: And all we, we started literally when I say we started with no money, it was no money and we built all this with no money for the first three years of change. Speaker 1: So how many users ended up, unique users have gone through it over time? Speaker 2: So when I started with them in 2015, with Salud in 2015, we had 10,000 uniques and when I stepped away in 2018, we were a million uniques. Speaker 1: Okay, and it's still going today. Speaker 2: I think they're probably up to like 20, 25 million unique users minimum, if not more. They don't, I mean, obviously I'm not privy to that information anymore, but yeah, they're still doing very well today. Speaker 1: Now, you said you exited, that's because the company was sold? Speaker 2: No, we, I exited because the agreement with Saud was always for him to be CEO in training. And because I had run my own startups and stuff, and I knew this space, the retail space, He's a great tech guy, still is, but you know, the things that, what I've learned in years over tech is you can over tech something, especially if you're looking at something that's a consumer based and you are not thinking about the consumer so much, you get enamored with the tech so much that you miss out on what you're supposed to deliver. And so what we had talked about was I was sort of like more like the product officer, the strategy officer and the CEO. Where he was a CEO in training and we told him, look, we'll do all this with you. And at some point you will be CEO, but you just have to trust us and we'll work on the product. We'll work on the business plan together and get us and the business model because we had to find a way how to make money. And then when everything gets to a certain point, even if we're early, you get the keys back to the car. Speaker 1: How did you make money? Because it was free service. So was it a data plan? Speaker 2: So when we started, the first six months, I think, we thought we were going to be affiliate marketing because I was like, oh, you know what? There's this thing, affiliate marketing. So we tried affiliate marketing and we thought, oh, so we started it. Fakespot was an affiliate marketing entity to Amazon. And we hooked it up and I think we were making $10,000 the first week, right off the bat. We had like 10 grand in the account. I was like, oh, this will be great. With limited amount of users, we already had $10,000 in affiliate marketing. Speaker 1: So someone would come and type in an ASIN to check it, then you would link it back with an affiliate link, link them back to Amazon. Yeah. You have a hot lead in your hand. Speaker 2: So we had 50,000 the first month and I thought, oh, this is it. We've hit the jackpot and we've got the gold mine. And Amazon promptly cut us off. It was funny because they would keep the dashboard showing how much money we were accruing by our links so we could see the money growing. But they would tell us, by this account, you're not getting any access to it. We've locked this account. But you could see the money growing. Speaker 1: That's crazy. So that didn't pan out, but so then what was the next step to actually monetize it? Speaker 2: So we didn't realize it at the time, but it didn't pan out because our conversion rates were so high. Because by the time you clicked on us to do a research on the review, and then you click through, and you get an A grade, you are almost guaranteed to buy it. So our conversion rates were easily north of 50%. And I think for whatever reason, Amazon caught on to that and then so they shut that account and then we thought, okay, let's try a different way. So we then tried a third party. So we found another company that was an affiliate link of the Amazon and we tried doing that and routing it through them and they caught on and that's when we realized, oh, our conversion rates are too high, that there's nobody that sends 50% of their links through Amazon that way and in terms of affiliate money and nobody's making this level of money that has this little volume. I mean, you gotta think about it, we only had like, let's say even at the end of the first year, we had like 100,000 users, maybe tops. We're like, you know, you're making quarter, you're like all that, it keeps running and how much you're converting. So our last ditch was to try and open up another company that we thought could be a sister company. And at the end of the day, that was just, it wasn't possible. And Amazon claims that two things. The first one, they claimed that we were routing people away from Amazon. And then sending them back, which in their terms of agreement, they claim is illegal. And I was like, well, they're not leaving your site. They go to your site, they come to our site, and then we send them back. And then after that, we had launched the, within the next year, we had launched a sort of an early version of the browser extension. And then they said, well, you're now trying to affect our product via extension. So that disqualifies you from doing anything with us as well. Speaker 1: Okay, so that didn't work. So then how'd you monetize it? Speaker 2: We realized that we had all this data. People around the world and now it's been slowly like you know a year, year and a half in we're like into different countries and that's also another thing is to make this work in the UK, to make this work in Germany, all different sorts of things you have to do to make it work elsewhere because Amazon is different in every region in the world. So we realized we had all this data and So we started knocking on some doors and financial companies started knocking on doors because we had what they deemed as alternative data sets. And this was when data 2017, 2018, data was like the big magic word of like, what do you have? What can you do with data? And so we managed to license out one of the world's biggest quant funds, quant data shops in the world. Contact, somehow we got in contact with them and negotiated a deal where they were willing to trial this for like six months to a year where they were paying us for us to keep sending them data on what people were tracking and what they were looking and they were looking to see if there were some trends based on what people are buying, what people are, you know, leaving reviews about and what people are checking and what basically a way to check on what's hot on Amazon without As sellers, you guys do this because you get like what's hot on Amazon, what you're selling and stuff like that. So it's a different way of tracking that. And so that was the last sort of stage of the monetization. And then the other thing we were developing was sellers tools. How do you help the seller? Manage, not manage as much but like figure out what's going on reviews, clean up the reviews if need to or challenge the reviews if they needed to be challenged. Speaker 1: Also you had an early system, that's a hot thing right now is there's companies out there that are doing it for you or showing you how to use chat GPT to see what the violations were. So you guys were doing that years ago. Speaker 2: Yeah, we were trying to do that early and at that time, the only review management system that companies are out there and there were a few big ones, but they were all only interested in managing reviews as in terms of getting people to write it or getting Or putting bad reviews on your competitor. I'm not saying everybody does that, but I'm just saying that was their review management. There was no real analysis of what's your velocity, what are you looking at, why are things coming the way they're coming, is there a seasonality to it based on how you sell, how long does somebody take before they buy to give you a review. So we were trying to do things like that to the small seller in a sort of platform format. But what we found is like a lot of sellers are busy fulfilling orders and just running the day to day of their inventory and all that. And as Amazon got bigger to manage, they were worried about reviews, but they weren't really worried about reviews to that level at that point yet. So we were a little too early. Speaker 1: And don't forget to sign up for my Billion Dollar Sellers newsletter. It's free every Monday and Thursday. It's like a $25,000 mastermind in a box. BillionDollarSellers.com and in just a few weeks I'll be doing BDSS, Billion Dollar Sellers Summit number 10 in Hawaii. Get information on that if you want to join us last minute at BillionDollarSellersSummit.com. Yeah, they were not worried about them being so much fake as they were just how to get more on my... Speaker 2: You can remember the days where like 5,000 reviews was a big deal or 1,000 reviews was a big deal. Now you're like 1,000 reviews. Is anybody even buying this thing? Right? Speaker 1: Well, in 2015, 2016, there was... How did you guys deal with the, back then you could actually use services like Zonblast or Vyralaunch or a whole bunch of these, there's tons of them, that would charge you 300 bucks, 500 bucks, 700 bucks to run a promotion and you would set up a coupon code on Amazon for free, you can't do this stuff anymore, or 99% off or 90% off or whatever and then they would have a list of people that they curated off of Facebook or wherever they curate them and they would say, hey, Kevin's got a new dog bowl, slow feed dog bowl, If you want it, it's free. All you got to do for getting this for free, use this coupon code and agree to write a review, to go back and post a review. And in that review, you need to make sure you put this terminology as something to the effect of, I received this product in exchange for an honest opinion. or something like that. And then I remember it was October 3rd or 4th or something 2016 Amazon overnight shut that down. Everybody was freaking out like how am I going to rank that because you could literally use one of these services and be number one the next day, the very next day. And then the reviews would start trickling in to help build that moat and Amazon shut that off. So how when you're analyzing these clearly those are Fake, you know, some people are saying their honest opinion, but some of them are just writing whatever and though some of these services had ways to filter it out, like they would first have the right to review on their service and if it wasn't positive, they would pass it through. If it was negative, they would filter it out. There was all kinds of stuff that was going on. How did you guys and then sometimes people will leave. I mean, I used to do this. I don't do this anymore. But people would accidentally, I would guide people to leave it on the seller feedback because on Amazon you have a product feedback where people write a review and give the product to the store and then there's a separate one that's based on the seller. And the seller one is supposed to be for like, oh, he didn't ship it on time or it arrived, you know, damaged, it wasn't packed very well or something very much delivery based, not about is this a good quality product or is it as advertised. And so I would guide people to actually, I would tell them, if you got a positive review, use this link. And if you're not happy, let us know here and we'll go to the seller feedback one, which you could easily get removed. Amazon would cross them out, you can report them and you can knock out all your negative reviews. So it was a way to game the system. None of that works anymore, so don't do that. But how are you guys filtering with so many, there were tens of thousands, hundreds of thousands of these. How did that come into effect? How did that affect the Fakespot analyzation? Speaker 2: So the good news is like, again, the part one of it is like when you have the history, you're able to tell. Because so every, so remember we said like we go to product and every product has their own sort of history. And then you go to every user. So now every user has their own history. Speaker 1: But you can see this is Soccer Mom in Kansas City when you look at the history of Mary and all her reviews are almost the same because she's doing one of these services things. Speaker 2: Yeah, all the reviews are the same and you'll find that most of them have the same cut and paste language. So because we were able to do cadence and language and do that kind of matching, where you can tell these are all the red flags if they cut and paste the same. So I used to give interviews for this. So a lot of these examples are still in my head. There was one person I remember that only reviewed automotive products once and Use the same language every time, the same three, four lines. Oh, this thing. So the first line would always be different. They would call out the name of the product. Unknown Speaker: Oh, this, this, this, this, this. Speaker 2: And then the second paragraph would be the same thing. I really like it, I use it, I would recommend this to all my friends and family, cut and paste. And just on that one thing, we were able to track, oh, this person has done this, this, this. So just on that level, you can track when people use the same language. So even if they don't put the first bit up front anymore, that this is a paid review, which a lot of people stop doing, you can then see when they're lazy and they're cut and pasting. And for us, if you're cut and pasting and saying the same things, It's not a reliable review. Now, there are different levels of reliability, of course, so then instead of A, maybe you're B, and then instead of a B, maybe you get down to a C, and if you see a lot of this, then it's obvious that it's some sort of bot or something that it's a D or F, you know, however that goes. Speaker 1: This was happening back then and it still happens today, but there's a lot of black cat sellers from China and Russia, China especially, that they'll use exchange students. So a Chinese exchange student comes to the U.S. to go to university, they will hire them and they will create an avatar for this person. And they'll say, okay, Ming, you're really into cats. You're a cat person. Anything that's being launched in the cat space, we're going to be forwarding you these links and we want you buying anything that a cat person would buy. And some of that's cat stuff, some of it might be, you know, cat people, I don't know, cat people buy a certain type of pan or whatever and so they would create so that you wouldn't be detected as, oh, all of a sudden you're buying all this random stuff that's all over the freaking place and it's kind of obvious. Speaker 2: Do you remember in 2016, 2017 where there was a brief moment in time where people were getting random packages from Amazon that they didn't order? Speaker 1: Yeah, it's called brushing. Yeah, it's called brushing. Yeah. Speaker 2: That was like, and I remember getting a lot of calls and interviews on that and they're like, why we get packages? Well, because they want to review it. They have to send the product somewhere and they just want to review it. And it's always like crap, like, well, crap. Speaker 1: The way that was working was there was someone had been, someone over in China, I know a specific, I know of, I don't know the person, but I know of a specific group that had somehow gotten 50 million Amazon customer buyers data. I think they got it from the back door from India or from China, but they had their actual addresses. They had my address and so they would just go and they would create virtual credit cards. They would sign up for a Bank of America and create You could do like a, you know, one-off credit cards that are meant for security purposes for your employees or for whatever, but they would go and they'd have multiple banks and there were services that popped up that would let you do a thousand virtual credit cards. So they'd use these virtual credit cards and you could put in whatever your address you wanted. So, so the AVS matching and everything would match. And so they would, they would match those credit cards and, and then use people to just say, Hey, make sure it's an either spoof it, but better than spoofing it was to, One of these students, let's say, go buy this product and just ship it to this random address or they do it from China and ship it to random address. So people are getting, it's called brushing, people will get this stuff just showing up, you know, 35 packages show up at their door. I didn't order any of this stuff, but they didn't care because what they were doing was they would have 10 different seller accounts selling the same product. So one account they would keep completely clean, totally clean, totally above board. And then the other nine would be variations. And those they're doing, they're burner accounts. And so they would have the same product and that's where they do all their malicious stuff, build up the reviews and then transfer those to the main account or use it to get the main account rocking and rolling. And then just abandon the nine and have one account that's stable. There's a whole process there. Speaker 2: And every year we would find these new things just the way that we were looking at reviews and we're like, you see the patterns. Speaker 1: Yeah. Speaker 2: And it's weird that you said that because you're right. They would transfer, like they would sell one thing and then transfer the, the ASIN to something else. And then the reviews would, you know, the reviews would be talking about like, let's say a t-shirt. And then next thing the reviews would be talking about a pair of shorts or whatever, something like that. Speaker 1: Well, that's called zombie accounts. Those are zombies. So what those were, I go in 2016, I saw some guru on YouTube telling me how I can get my own Lamborghini and live in a mansion. I launched a product on Amazon and it was selling on Blender, let's say, and it just didn't work. I just lost my ass because I didn't know what the hell I'm doing. So I just said, screw it. I ran out of inventory. I just let it go. But I did not delete the listing. And so what would happen, but maybe I had some decent reviews. I just didn't have enough money or whatever. So there was tools that would go in and scrape Amazon and find these zombies that are just dead men, dead listings walking and they would take them over with a 1P account and then take them over and then merge those reviews. And they figured if we get enough reviews at the top, nobody scrolls all the way down. Enough of what the product is at the top will be okay. And a lot of people don't even read the reviews, they just look at the stars. So they're not worried if it's about if I'm not selling a blender, if I'm selling a pair of socks, who cares? That was what that was. I remember Slick Deals actually, to get on Slick Deals, which is one of the biggest sites out there for deals. They actually, I don't know if they partner with you guys or they just use you, but you had to have a certain score on Fakespot. Unknown Speaker: Yeah, to get on Slick Deals, yeah. Speaker 2: We actually, I think in 2018, we actually talked about them either buying us or buying a stake in Fakespot or doing some sort of alliance with Slick Deals at some point. So I had some pretty deep conversations with those guys. Speaker 1: So did Amazon always treat you like the brother from another mother they don't want to talk to or do they ever like say, hey, it's kind of cool what you're doing, we're going to knock you off or we're interested in it or are they just like, I wish you guys wouldn't do this? What was their general attitude or did it evolve over time? Speaker 2: I think it evolved over time because in the beginning we were just like a little speck you know like a drop of in the ocean so they didn't really care like you know you're like and then every six months to a every six months or so we would get some sort of contact and then gradually it would get more and more like Hey, this is Amazon you better cease and desist or we're gonna do this or this is this and so but what's funny and This is the inside story of a scoop for you guys Right before the pandemic, Amazon called us to start acquisition talks. And then the pandemic happened and the whole thing went away. But I think that they knew that they had a lot of bad publicity over that point. And I think the company would have probably ended up in Amazon's hands had the pandemic not happened and nobody ended up caring because all they cared about was that things could get delivered. And that you can ship it back one way or the other. And so after a while that went away, but for the longest time, and then they even complained to Apple to try and get Fakespot booted off the app store so you couldn't look at it because our app was essentially an overlay onto Amazon's on mobile. So it was kind of interesting as well. So you could look at it, it would turn us on and it would overlay onto it. I do a review, so they were like complaining left and right. Speaker 1: Did Apple do anything? Speaker 2: They took it off for a while. I think it went back up after a while. But yeah, the more attention we got, the less desirable we became to them or the less Fakespot became to them in their eyes. Speaker 1: So you became, you said you did a lot of interviews, so you became, you were on every reporter from the Wall Street Journal and Business Insider and all these guys, CNBC, you're on their hot list. So anytime there's any kind of story about reviews or brushing or whatever, you were on their speed dial, right? Speaker 2: Yeah, you know, I did the Bloomberg e-commerce podcast with Spencer Sloan, is his last name? Is it Spencer? Like, so a bunch of those Wall Street Journal And then some international stuff. And then you'll be surprised, a lot of local TV stations, like regionally, like Kansas City, Denver, some of the ones, I can't remember, they were always doing these little, every time somebody complained to their TV news about fake reviews and all this stuff, like somebody was like, Channel 5 is on your side. Speaker 1: If something, if you've been ripped off, call us, we'll do a story. Speaker 2: Yeah, I did a lot of remote interviews for things like that, yeah. Speaker 1: So what's some of the craziest stuff you saw people doing like when your system like you said was you could see all these patterns What was something you're like? Oh my god. This is just like I can't believe they're getting away with this or something along those lines. Speaker 2: You know, I think the thing for us was that, you know, most of these companies, most of these sellers were using services. So the patterns were always the same. And we taught everybody, I was like, look, if you don't believe us, look at it that way. Like, and you said just now, like nobody scrolls through, they look at the top. Or usually they don't even look at the top, they don't even filter it. They just look at what are the top reviews and they give it a passing glance. And the most egregious ones we always see were like that they were repeat of the same reviews with the same language. Because if you're going to do hundreds and thousands of reviews, you're not going to get creative. You're just going to throw it up against the wall. And there's going to be so much that it won't matter. There's going to be so much that won't matter. So for us, it wasn't that, it was just the sheer volume of it. At one point, we were doing some analysis for different media companies and we found that like, let's say phone accessories, some categories, like phone accessories were like pretty much 70% of the reviews of phone accessories were all fake. Especially like cases that were being made in China and shipped over here and stuff like that, they're all fake. Not one of them is credible in that way because let's face it, you buy a phone case, are you really going to leave a review if it's a $7 item? But you'll find that these $7 items have tons and tons of reviews and you're like, why is that the case when it's a throwaway item? So there are things like that that we found. Speaker 1: Because reviews are moat. I mean, so that the number is a moat. And I know now even with Vine reviews, there's a lot of people, I don't know how Fakespot's addressing this, but quite a few of the reviews from Vine reviewers are no longer real. They're all AI written. I know Vine reviewers that are not writing their own reviews anymore. They're just typing in, giving a chat GPT or Gemini or something. Here's the link to the listing, write a review. Yeah, and they're posting, they're fake. How do you deal with that? Speaker 2: So the AI aspect, and because we started with the AI aspect of it early, like I remember the first time I really worked with Saud was like when I heard the terms machine learning and how he was training How is training them to learn about how people write and cadence and language and like how language of somebody with a different, different writing style will mean differently than somebody with a from a different background stuff. So the and part of the reason why When Fakespot got bought last year, it was bought by Mozilla's AI arm. And you're like, why is Mozilla's AI arm buying this? Because of that. Because the machine learning and the AI at Fakespot has to outpace All these things that are going on. So if you're using ChatGPT to write reviews, we, well, we... Speaker 1: You can tell, every LLM has a pattern. Right. Every LLM has a pattern. Now you can affect that pattern if you're good at prompting. Speaker 2: Right. Speaker 1: But the average person is not. So you're going to catch 80 to 90% of it. Speaker 2: So you have to get even deeper into your AI as to how to catch these guys. But we, Fakespot has a little bit of a head start because they started this a little bit earlier and I'll tell you another sort of funny anecdotal story. In 2018, we got a call to go present to the NSA. And to present to DARPA and we're like, why does the government want to see us? So we thought we had done something wrong. And as it turns out, they were looking for a tool to analyze research papers and figure out which were real and which were fake and which were plagiarized and which weren't plagiarized, because they were getting so many research papers on these kind of topics from around the world, they couldn't tell which was real legitimate to put their attention on. So the government was look, well, the NSA, I was interested in licensing our engine to do something with this to spot fakes in the research world. Speaker 1: What do you think about, there's been talk and Amazon tested this a little bit last year and they went away from it. But I know it's Casey Goss, back when he had Vyralaunch, young kid, super smart guy. I think he's 20, 21 when he started Vyralaunch. But one of the things he came out with, this is years ago now, he said, Amazon needs to change the way they're doing the reviews. He said that Amazon, the way they're doing the reviews, it's not fair because something that has 10,000 reviews, has a moat over a new product. Maybe I've come in, I've developed a new product and I've actually got a better product that actually solves the problem that everybody's complaining about and actually should be given a chance. And I have no chance against something that 10,000 views. It's almost. 95% of the time, it's going to always beat me. So he said that Amazon should change the way they do reviews and only show like the last year or something like that. So something that has 10,000 only show 820. What are the last year's worth of reviews? And that should be the broader, not this lifetime number. So Amazon kind of checked, tested that a little bit in India and the US last year, and then they went away from it. Speaker 2: He's correct. I think there's got to be a better way of doing reviews. But I also think that we, you know, coming from the retail side, much like in retail, everybody complains that we train the consumer to only focus to buy something when it's a sale time. We train the consumer to only buy something when it's High reviews. How many people do you know look at restaurants or product reviews? As you said just now, they don't even read the most recent reviews. They don't bother to filter it. They're like, oh, this has 1,000 reviews at four and a half stars. The other one has 10,000 reviews at four and a half stars. I'm going with the 10,000 review one. I think part of it has to be how do you train people to read, to understand reviews better? And then the other part of it, yes, you have to change. I do that now, like whenever I look at anything on Amazon or what I'm buying, I only read the last two, three months because I know the product might have changed somehow, you know, because you know, these things, especially if they've been on the market for a couple of years or whatever it is. So I don't bother a review from 2019. I don't bother reading. I'm like, okay, give me what happened the last three months, six months. Speaker 1: And then the velocity, you know, a lot of times what I do as I look at the star rating Okay, this one's 4.5 and if it's they're equal and one is 10,000 one is a thousand I 10,000 reviews with a 4.5 is better than 1,000 with a 4.5 and then I will I will look through the last handful reviews I don't go three months or whatever I'll look through the last handful and then I look for the worst negative ones because and sometimes the worst negative ones are Not necessarily bad. Sometimes it's, not just on Amazon, but if it's on a hotel website, it's someone bitching that, they gave it a one star because their next door neighbor was having an argument all night long. That's not on the hotel. Speaker 2: Even on Amazon, how many times somebody will say, oh, the box came squished a little bit. And I'm like, okay, but did the product survive this? But all you talk about is the box came squished or, oh, they put so much packaging for a little thing. I'm gonna give this one star because I'm an environmentalist. You're like, but did the product survive? Speaker 1: Yeah, I throw those out. Yeah, I throw those out. But there is a, I think it's statistically, maybe correct me if I'm wrong on this. I think it's, is it 21 reviews statistically is the number needed to actually start to trust it? Someone actually put that out. This was a stat that I can't remember where it came from. But some statistician said you need 21 actual real reviews. If it's a fake spot F, 21 doesn't count, but 21 real reviews can actually give you an indication of what this product is. Speaker 2: That's a misnomer because, and the word I've used a couple times is the velocity of it. If you get 21 reviews in a short, short, short period of time, and so that's another thing that Fakespot would compare is like, okay, this is a competitor one. And you get these amount of reviews in X amount of time from a comparable product perspective, should you be getting those? And so we caught a lot of people like that. We had people write into us and said, complain as sellers, they're like, how dare you give me a C grade? You know, I got all my reviews legitimately, blah, blah, blah. And then we were able to say, well, you got all your reviews in a seven day span within launch of, and you had X amount. And we can tell from sales, Because we were able to track all these things. How much you sold versus how many reviews and your conversion rate of reviews and like if you sold 21 products in the first week and all 21 people left you a review, we can pretty much tell that they're not all going to be reliable. And then, you know, there are other things to look at, but the velocity alone of that will tell you. So just saying that, oh, 21 is legitimate. It's true to some point, but not really true to some point because the velocity of accounts, the time period that you get accounts. Speaker 1: Hey, what's up, everybody? Kevin King here. You know, one of the number one questions I get is how How can you connect to me? How can I, Kevin, get some advice or speak with you or learn more from you? The best way is with Helium 10 Elite. If you go to h10.me forward slash elite, you can get all the information and sign up for Helium 10 Elite. Every month I lead advanced training where I do seven ninja hacks. We also have live masterminds and every single week, one of those weeks I jump on for a couple hours and we talk shop, we talk business, Do in-person events. Helium 10 Elite is where you want to be. It's only $99 extra on your Helium 10 membership. It's h10h10.me.me forward slash elite. Go check it out and I hope to see you there. If I'm not doing any hanky-panky on my listing or any kind of, if I'm just selling my product and leaving it at that, I'm not doing a follow-up sequence with a tool. I'm just letting Amazon do their thing. They send out, hey, review this product and people just naturally go and review products. What kind of rate should I get? So, because people always come to me and they say, Kevin, how do I, it's the number one, still the number one thing. How do I get reviews? How do I get reviews? What's the best way to get reviews? And I just tell them, sell more and put out a good product. And so what is the number that you guys figured out, okay, this 2% rate of reviews organically is pretty natural, or is it a 4% or is it a 1% or what? Speaker 2: So we looked at it at, depending on the type of product, and I'll get into this, we always thought that between a 5% and 10% conversion rate made sense. Speaker 1: 5% to 10% review rate? Speaker 2: Yeah, it makes sense. But again, it depends on the price of the product. And it depends on the sort of like the trends of the moment in terms of like, you know, is this a summer product? And so people are more liable to leave a summer review and stuff. And then every so often you will get something that's like a supernova kind of product that's so new. So we were, so I had a couple of different sellers where I would call and like we would talk about What you're seeing in reviews and why and versus, and they're all like top sellers. I always like to pick the brains of sellers. That's how I got to know A because I knew it was like, it was a sell pass. I just wanted to pick his brain of how it worked. And she had, she was the person that had invented the mattress clips that clipped your mattress, your sheet to the mattress. And they were the number one seller. They were made in the USA and stuff. And then all these copycats, It started showing up overnight and it took a hit and stuff. But when I was looking at their velocity in the beginning, there was nothing like it. And people were so ecstatic that at $12.99, well, I guess now $8.99, whatever the price is, a cheap product like this made such a difference in their lives that they were willing to like Share it. So every time you get this sort of like supernova product and then you have to discount for the fact that you have to then take an account that this is a supernova product. And so that's why you're getting that, right? And so, and we, I experienced that ourselves with our memory foam bath mat when we first launched memory foam bath mat. And I was like, I don't know, employee number 11 of the company or something like that. And Our sales kept going through the roof and people kept leaving reviews or word of mouth because they were like, oh, this thing is insane. And every, at that time there were no reviews on sites, but there were blogs and blogs would start talking about this. And people then comment that the blogs, how great this was and stuff. And just, it was just a supernova product, but you can't compare a supernova product with something as basic. Let's say you were launched a new, a new battery product today. You're not going to get that kind of velocity. In fact, you're probably going to be lucky if you convert 1% of your sales into There's got to be a compelling reason why people are leaving a review for you, you know, especially for a lot of throwaway products or a lot of products people think like, yeah, you know, so there are these things that we have to take into account. And so one of the things that Saul would do was like, look at the competitors and what kind of velocity everybody has. So what should you expect in this category for this kind of product? Speaker 1: So what's the future of reviews? Reviews are a very important element when it comes to e-commerce and selling online, not just e-commerce but anything you buy. Where's this going? Are we going to trust anything, especially with AI now? Are we going to have a fake spot, a little icon in their Apple Vision Pro that says, Ming, yes, this is the real Ming. This is not the fake Ming. Where's this going? Where do you see this going? Speaker 2: It's like an arms race on either side, right? Let's say the dark side and the light side and like the guys who wanted to be very useful and very functional and sort of like where it's a community feeling where you can really trust that person who you don't know but you know enough context where you trust it enough where you then you make a decision purchase of it and then you always have the dark side of like, well, I'm just gonna game this as best I can because this isn't irrelevant product and as you spend what's 10-15 dollars if you if you don't like it as a throwaway I'm not sure there's a good answer for either one of these. I think it's always going to be like an arm series that way. I also think that, you know, we as consumers have to be not lazy. And, you know, yeah, we as consumers have to be not lazy. Like, look, it takes you a couple extra clicks To look at something, it's just like when you're in the store, you buy it, you pick it up, you read the, you know, again, this is my training from Walmart, they were like, people come in, they pick up the product, they look at the call outs and the stuff, and then they put it, then they look at what's around it and all the shelves are selling the committed products, what are the committed products saying about it, and then you put it in the cart. If this is the one you want, you put it in the cart. We don't do that now, instead of looking at compared products, because we just want it and you want it, you want to make that purchase in seconds. But an extra minute isn't going to kill you. And I think we just have to get back to be a smarter consumer. Like you said, you read the worst ones and you're like, are these really that bad? Do they really matter? They shipped it a day late. Oh, this is a one-star. Is that really an issue? I mean, it is and it isn't. Speaker 1: Awesome. Well, Ming, this has been really fun. We could keep talking for a while, but we've been going on an hour here. My editor's going to kill me. He wants me to keep these a little shorter, but when it's so good, you got to keep going. So Mel, don't cut this because this has been really good stuff. Really appreciate you taking time. If people want to learn more about you, what you're up to these days or what's happening, is there a good way for them to do that? Speaker 2: Generally, everybody can find me on LinkedIn. I try to use that to sort of like my background. I'm running a different tech company. I run a different tech company now that's in real estate tech that's not really in this space. So I'm not going to push it too much. And then obviously, you know, for me, the better Fakespot does in the long run, the better it is for me selfishly. So I would encourage everybody to like, keep looking at it, keep using it. And, you know, But yeah, look for it on LinkedIn. I think Ming Ooi, there are not that many Ming Ooi's on LinkedIn so you can always find me. Happy to chit-chat anytime. This has been a lot of fun. This is great. Speaker 1: Awesome. Thanks, Ming. And that's spelled M-I-N-G for those of you driving or working out right now. M-I-N-G and then O-O-I. The last name's O-O-I. So if you're trying to look that up or make a note to that. Ming, I really appreciate it, man. This has been awesome. Hopefully, I'll see you again soon. We can have another chat on something. Speaker 2: I'll see you next year at Prosper and we'll have a cigar with Abe. Speaker 1: There we go. It's a deal. It's a date. Thanks, man. Speaker 2: Thank you very much, Kevin. Talk to you soon. Speaker 1: Really fascinating stuff with Ming. He was an open book just talking about the whole history of reviews and sharing some really cool information. I hope you really got a lot from this episode. Don't forget to sign up for my newsletter BillionDollarSellers.com and we'll be back next week with another great episode. We get talking about some PPC stuff that's gonna really help you out. I think you're gonna really like it. But before I let you go today, I've got some words of wisdom. People pay in anticipation of what they're going to get. Instead of gratification for what they did get. People pay in anticipation of what they think they're going to get instead of gratification for what they did get. See you again next week for the 100th episode of me hosting the AM PM podcast.