TMBA 587: The Real Problem With Fake Reviews
Podcast 49:31 | Download | Spotify| Stitcher | iTunes | Comment
When we received an email from today's guest, we were fascinated by his story and knew that we wanted to speak to him.
Curtis Boyd is the founder of two reputation management businesses but with a very c21st spin.
The first is a bespoke consultancy for high-level service companies and the second is a software as a service business called Objection.co.
Both of these were created with a singular goal in mind: to identify and dispute fake reviews on the internet.
Curtis' entrepreneurial journey began by disputing a single review for his first client and he soon found himself competing in an arms race with fake reviewers in what has become a multi-million dollar industry of deception.
Curtis joins us this week to discuss how a chance encounter while he was still a student nurse led to his consulting career, why he decided to go back to school to learn about artificial intelligence, tips for identifying fake reviews, and a whole lot more.
We've also opened up a limited number of one-on-one podcast strategy meetings as a way to test out our new services platform at Dynamite Jobs.
Listen to this week's show and learn:
Why businesses are so serious about removing malicious reviews. (3:29)
How Curtis was able to pay off his student loans by getting one fake review scrubbed from the internet. (10:27)
Why he chose to go back to school to learn how to develop artificial intelligence. (20:44)
The cat and mouse game of detecting fake reviews. (33:21)
What Curtis believes is the next step in his battle against fake reviews. (39:27)
Mentioned in the episode:
Before the Exit - Our New Book
Partner With Us
The Dynamite Circle
Dynamite Jobs
Dynamite Deals
Tropical MBA on YouTube
Post a Remote Job
Dynamite Jobs - Remote Recruiting Sales Page
Let's Talk High-Level Podcast Strategy for 1 Hour
Curtis Boyd
The Transparency Company
Objection.co
Yelp Questionable Content Review
Yext
Guaranteed Removals
Enjoyed this podcast? Check out these:
TMBA571: There is Only One Kevin Kelly
TMBA582: The Upwork Hustle
TMBA584: From Services Failure to Software Success
This week's sponsor:
Today’s podcast is sponsored by Dynamite Job’s Done-for-You Recruitment Service.
If you run a growing seven or eight-figure remote company, your next productive team member could be just one simple phone call away.
Our recruiting services are designed to save weeks or even months of your own effort in the hiring process. While we’re finding your next dream team member, you can avoid the ‘job board grind’ and ‘hiring hustle’ to instead focus on building a better business.
The process starts with a free, no-obligation call with one of our senior recruiters. We’ll then go out and execute the entire job search on your behalf. The job will be marketed to thousands of candidates, the applications will be filtered down, the talent will be interviewed, and you’ll receive the top picks hand-delivered.
Hiring can be a total pain but the team at Dynamite Jobs is doing it successfully and efficiently every day. Schedule a call with their team and learn more here.
Listening options:
Subscribe via iTunes.
Listen on Stitcher Radio.
Android users can grab our RSS feed here.
Right-click and "save file as" for direct download.
Or click here to get sent to the top of the page where the on-site player is.
Have comments about the show?
Do you have ideas for things you'd like Dan and Ian to discuss on future episodes?
Our producer Jane would love to hear from you at jane@tropicalmba.com or leave us a voice message using the record button below.
Thanks for listening to our show! We'll be back next Thursday morning 8AM EST.
Cheers,
Dan & Ian
Full Transcript
Dan: Happy Thursday morning. Welcome back to the TMBA pod. A little behind the scenes talk - every Monday night we have a production meeting where we go over basically every pitch that comes our way, and we get a lot. And it's cool, y'all email us things that you think would be interesting, whether it's your stories or people that you know, and we debate, we review, we see if it's something we want to talk about. And so I just wanted to put that out there because it matters. It matters what you email us to Jane at tropicalmba.com that's our producer, and to myself, Dan at tropicalmba.com. Of course, we get tonnes of these things. So most of them don't make it to the show. But today's guest, man, his pitch really appealed to us. He pitched us the story of creating two businesses: one a bespoke consultancy that led to a software as a service business, a progression we commonly see, called Objection.co. And what the software identifies malicious and fake reviews. And man he has got some interesting insights to share.
[bg_collapse view="button-blue" color="#ffffff" icon="arrow" expand_text="Show More" collapse_text="Show Less" ]
Curtis: Just like dolphins and whales, fake reviewers travel in pods. They want the highest margin of profits. So they'll use the same accounts to review the same businesses. And you'll see these accounts used repeatedly across multiple businesses. But you'll see you know, the same 20 reviewers reviewing the same 20 businesses. And as a consumer, you don't see that because you're just looking at that business's reviews.
His name is Curtis Boyd, and this is a classic TMBA journey. That is - Curtis was working like many of us in a J-O-B. In his case, as a nurse where he identified a problem he felt he could solve: doctors losing money and reputation through unfair reviews. Now I gotta say I had no idea. I mean, of course I look at reviews every day. But I hadn't thought that much about reviews outside of sort of the Amazon ecosystem. And I learned so so much about the profound effect that reviews have on commerce on the web in today's interview, so stick around. Coming up, you'll learn about coming up, you'll learn about how Curtis used his hustle muscle to pay off his student debts in one large fell swoop. It's an incredible story. And that gave him the springboard to start on his journey towards building a SaaS at scale that solves this problem and uses AI technology and machine learning, something that many of us are interested in. Stick around to the end of the episode to hear all that and more. I'll pop in at the end with some updates as well as an opportunity to book a call with me if you'd like a podcast strategy session. I started by asking Curtis the core motivation for people who use his service
Curtis: It's pain. I mean, at the end of the day, it comes down to bad reviews causing their business pain, because of the objections that their leads give them when it comes to new sales. Whether it's legitimate or not, when you read a bad review, you become fearful, you wonder is this experience going to happen to me? And I think reviews provide a tremendous amount of reassurance there. So you know, when, when a bad review is written for a business, the business owner, you will, you know, usually seek out ways to see what they can do, especially if they believe that this review is unfair. A lot of business owners, in my opinion, they are, you know, ethical people, they will take ownership if they feel like they screwed up, and they'd be like, ‘You know what, I deserve that review, I'm not gonna try and dispute it’, I'm not gonna try and fight it’. The problem is, is we have more and more entitled consumers who feel like they can go online, destroy a business's reputation, and kind of get away with murder only because they didn't absolutely get what they want. And they will manipulate and they will lie, and they will make stuff up in their reviews to get what they want. And that's when a business owner usually finds me and they'll say, ‘You know what this review is totally out of line, totally untruthful’. And we will systematically take an approach to dispute and contest these reviews until they get taken down.
Dan: Do you have a story that you could share about an instance of a customer using reviews inappropriately for leverage that was particularly memorable?
Curtis: Oh, my gosh, it gets particularly bad as the price point goes up, right? I have customers who run restaurants and they'll get some silly stuff written about the restaurant because they want a $7 gift card for the next time they eat there, right? But then I work with some cosmetic surgeons who charge $15,000. No matter what your surgery, the outcome of a surgery comes, two days after surgery, nothing looks good, right? That takes weeks and weeks to heal. So even like two days post op, you take photos yourself, you upload it to the internet, it's really out of context, because well, you're two days post op. And especially when you know the price of the service goes up, you'll see more and more ridiculous things being posted being used as leverage and manipulation against a business owner to get their funds back. It's all about the money for those people, they want the service for free, and they want their money back. And they're going to use reviews in order to do that.
Dan: Interesting, help us to visualise what your company is all about?
Curtis: So we have five people on staff. We have two full time SDRs, basically appointment setters. We have two full time developers on the back end. And, basically a customer fulfilment director who makes sure all of our customers are happy outreaching, making sure any support queues are taken care of. We right now are running anywhere between 55- 60K MRR and that's over the course of about 18 months of being in business.
Dan: Congrats, that’s exciting.
Curtis: Thank you it's a fun place to be right now. Because we're at the point now, where we actually have a budget and to do experimental channels to find new customers, something that we never really got to do before. But now we fortunately have a little bit of padding and, you know, padding for employees for paychecks and stuff. It's a much better place to be at so thank you.
Dan: Yeah, congrats. And, you know, I want to get back to the company because I'm, I have a lot of questions about your process and how, what we can learn from you, but I want to kind of roll back the tapes really quick, just to get a sense for sort of your core motivation. Were you a kid that knew they were going to own a business someday? Or was it a gradual process for you to become an entrepreneur?
Curtis: I was an entrepreneur kid. I don't know if you I don't know how old you are.
Dan: I’m 39.
Curtis: Yeah, that's what figured. Do you remember ‘Raven's Revenge’ in school? It's like vial candy. It was pure sugar. And they ravaged the schools for six months?
Dan: I know what you're talking about yeah.
Curtis: I was the guy who literally went to the liquor store and was like, whatever boxes are coming in, just like I'm on it. As a 13 year old, I had hundreds just ready to go for that liquor store cuz he was the only guy in town who could get it. And I sold them individually for $5 a vial to these kids who are just fiending for sugar. I was obsessed with multi level marketing stuff when I was super young, you know, when I was 18?
Dan: Really?
Curtis: Oh, yeah, I loved being in a sales spot. I loved just having being able to say like, ‘I have a business, I sell products’. My stepdad, he who took care of me for a really long time. He ran a very successful, very large, HVAC company. And, you know, you always look up to the people above you. I was always trying to figure it out on my own, I knew I didn't want to work for him, I knew I wanted to find my own way. You know, he runs a truck company that runs about 100 trucks. So it's pretty, it's pretty busy here in Los Angeles. Watching his success over the years and seeing what he was able to do, I knew that I needed to just hustle. I didn't know what I was gonna do at the time. I knew I just wanted to do it.
And then on the other side was my mom, my mom was very sensible. She was a nurse and she's like ‘Curtis, you know, you should get a good job, nursing is great. And you should go to nursing school. And don't worry about business, you know, you'll always have a job as a nurse’. And it was for me, it was always a pull between the sensible and the, I don't want to call it unsensible but the entrepreneur pull. And it really, it really is a pull. I mean, he he, you kind of don't have a choice. It's like the subconscious feeling that you need to go out and either build something on your own create a system, create something of something of value and then go sell it, you feel that, that pride in that value of doing something on your own and reselling it so, yeah, I certainly struggled in early college, kind of figuring out where my life was gonna go. And then one day, it just happened and I never looked back.
Dan: What did you go to college for?
Curtis: I went to college for nursing.
Dan: So you took the advice of your mother?
Curtis: Yeah, on the advice of my mom. Absolutely. You know, nursing don't get me wrong is always a stable job, especially now in the health care pandemic. I mean, they're always calling for nurses. I ended up going to school for nursing. I got my bachelor's degree in nursing, I actually had a really hard time graduating, because at the end of it, I was really, really cranking things out. Is this a good time for me to go into my full story?
Dan: Yeah, absolutely. So I love the idea of polarities. You're in university the whole time and thinking, ‘Well, at least I'll have the nursing’.
Curtis: Yeah, but absolutely.
Dan: What changes things,
Curtis: The desire, I think, to solve problems, and then solve my own problems at the same time, right, this I win/you win dilemma came into my life, when I was in nursing school, I was, uh, you know, I was precepting. It was my last semester, I was in the ER, we had a doctor come in, he's a cosmetic surgeon. And, you know, he was doing a consult for a patient who came into the ER, and I was, you know, really just kind of cleaning up the room as a student nurse. He started complaining about his online reviews. This was about 2014- 2015.
Dan: And this is like, four or five years before the conception of your business.
Curtis: For my SaaS company, absolutely. So, I started complaining about my student loans, he’s complaining about a bad review. I’m like, ‘I hear you doctor, you know, I owe $30K in student loans. He kind of rolls his eyes a little bit like chump change, right? He's like, ‘Curtis, you figure out how to remove my bad review, I'll pay off your student loans’. And I'm like, ‘What, what, are you serious?’ And he's like, ‘Yeah, I'm serious’. And I kind of don't follow him around the unit the whole day. But inside, I'm like, I'm gonna do this because I could work an entire year as a student nurse and still not make 30 grand. And as a nurse, you could absolutely but you know, at the time, I was a student nurse working as a CNA.
Dan: How do I know he was serious?
Curtis: That's a great question. Two days later, he has a private practice. I called his office, I talked to his secretary who knew about the issue. She knew about the bad review. And I said, you know, as a student, nurse, my name is Curtis, the doctor, you know, kind of blurted that he was still looking for help in getting rid of his bad review. I talked to him, I was wondering if he was serious about the offer he made? Is he willing to talk and I could hear them talking back and forth. And she was telling me how they hired an attorney, how they hired a consultant and nobody could help them. It had already been about a month, and he was losing about two to three consults a week, which for him, $15-20,000 surgeries, he thinks he had already lost S150K because of this bad review. He was happy to pay my student loans off.
Dan: So you, you have no reason to think that you could solve this problem. You weren't in the online review game at this point. You were just a nurse.
Curtis: No, I was just a nursing student. You know, I played a lot of video games and was on computers and websites, I knew websites pretty well. I knew how to navigate stuff. So when this doctor challenged me, I immediately accepted it, I didn't promise him anything. But I said I was going to look into it and make an attempt to take care of this problem for him. I went to work and I started disputing, contesting and rejected, rejected, rejected just across the board for about two weeks. I had just finished this semester. So I had about a two and a half week break, I had $800 in my bank account. I decided to use that money to fly to San Francisco, to fly and meet people who worked at that company to see if they would tell me how to do this. The doctor promised me if the review was illegitimate, that it was fake.
Dan: Can you remind us what the nature of that review was like and what was illegitimate about it?
Curtis: So this doctor had no idea where the review came from. He believed it was written by a competitor, it included a lot of words that were in his like wheelhouse, you know, in his vernacular. So he didn't he didn't believe that an actual customer wrote it. And he felt like he was being targeted. He was a pretty predominant cosmetic surgeon in the community. Even back in 2015 business owners were still going after other businesses in the more lucrative areas like cosmetic surgery, personal injury attorneys, I could go through a few other industries where even back then that was really still a thing. Today, everyone's after everybody. But this particular review was, the doctor told me, was not a real patient. And he believed it was written by a competitor.
Dan: So you're 22 you show up in San Francisco, and you're like, what are you gonna go down to cafe Java Joe's? And like, what's the game plan?
Curtis: The company, you know, based out of San Francisco, I figured people who work there, I camped out at Starbucks. It was Yelp. I mean, I can't remove Yelp reviews, the only administrators can remove anything online. But I was hoping to meet someone who worked at Yelp. And I camped out at different coffee shops. I kind of staked out, you know, I was 22. So I could get into bars and stuff. And, you know, I stayed at a hostel. It was $40 a night, you know, shared room and stuff. I stayed there and I bar hopped, and I tried to network as much as possible until I met people who worked there who could help me. I almost felt like a homeless person approaching strangers saying, ‘I’m sorry. Hey, do you? Do you happen to work at Yelp? No, do you know anyone who does, I'd love to meet them. I'd love to be introduced to them. I'm trying to solve a problem for a doctor down in South Bay, California, where I'm from. I'm here just to get some information’. And I would repeat that over and over to strangers on the street. At first, it was like, ‘What do you want? Like, do you want money?’. ‘No, I don't want money, do you happen to work for Yelp. Like it's right there’. And finally, I met someone it was, it was on a Sunday afternoon, I met someone at a coffee shop who said they worked for the company and I can't say their name. But they showed me how to essentially take the dispute to the next level and how to get to the right moderators who can take a second look at things. And I did and 48 hours later I had a check for $30,000 from this doctor.
Dan: And now is that a public technique that they shared with you?
Curtis: It is a publicly available technique, absolutely.
Dan: Can you share what they shared with you that day?
Curtis: Sure. So if you Google the phrase ‘questionable content’ on Google, ‘Yelp questionable content’, it should be the first thing that comes up, there's a link where you can click enter the case number and then provide a secondary follow up to your dispute request. On Yelp, once you flag a review, that flag button is removed, and there's kind of a static icon that's there. And you're done. You can't as a business owner, as a user, you can't dispute that review a second time. And that's the problem that doctors were having. They had flagged the review. They said no. And that was it. And they were unaware that you could actually bring up the request, again, in order to get in front of a moderator to have this content evaluated for a second time.
Dan: Wow. You flew to San Francisco for what is a Google search essentially?
Curtis: Yeah 2013, the link was a bit different back then. And harder to find, but she also gave me some really great advice in the content, like the strategy involved in getting the moderator to agree with me. Like to align myself with Yelp because at the end of the day, Yelp wants this content to be useful for consumers, right? Yelp wants to have the highest quality reviews out there so the quality of the site will be enhanced. And at the end of the day when your removal strategy aligns with Yelp goals or with Google's goals, your outcomes are going to be a lot better because you're enhancing the review site, you're enhancing the quality of reviews that are on their site.
Dan: Why do you think she was willing to help you?
Curtis: I don't know, maybe she felt bad for me. Maybe she understood my journey and where I've been thus far, she probably knew that it's hard to communicate with administrators, it's hard to get unfair reviews removed, you know, maybe with her position that where she worked there. I wouldn't say that it's overly difficult to communicate with them. But it's not easy. It's not easy to get your point across and really explain it in a way that, you know, they can they're allowed to take actionable recourse to.
Dan: Right, you can't just go to them and say, I don't like this review. Tell me about then the conversation with your first client, like, how did they respond to this? Did they believe it?
Curtis: He was in total disbelief that this young student nurse out of nowhere, was able to take care of this problem for him. I was just beaming with confidence, I knew I could do something about this. If you really try and you really invest your time and energy into solving a problem, you can solve it. And the reward, the first time I removed a review for someone, I just felt like, wow, I just did, I just did something great for someone. And financially it really helped me out too. But the look on his doctor's face, like the appreciation he had for what I just did for him and his business, the effect that it's going to have on his business for months and months and months gave me that entrepreneur pull I’ll call it. It just bites you and you're like, wow, I want to feel that again, I want to feel that again. And again. And again. This particular doctor was on the board of directors for the entire medical centre. So I struck gold. He introduced me to every doctor within that entire physician network, which was about 575 doctors. He told me how much I needed to charge in order for things to kind of make sense, and introduced me to the CPA. And before I graduated from nursing school, I had a little under 1000 physician customers with my reputation practice at the time called ‘Future Solutions Media’. And that was in 2015/2016. And I had a reputation practice
Dan: So he became something of a mentor for you.
Curtis: Absolutely, all of a sudden, I was making, I don't want to, you know, sound ungrateful. I was making rather than the $50/60K a year a new nurse makes. I was making that at some points every .. Well, a lot, a lot sooner, let's just say that I was making. I was making a lot more money. And I was young.
Dan: Essentially you're running an agency, correct?
Curtis: I was.
Dan: High end services, local services, we're calling you and saying, ‘Hey, Curtis, I need your help’. And, you know, why aren't we talking about that?
Curtis: So after a few years of doing this for not just doctors, but lawyers and contractors, and growing my online brand for my consulting company, articles about AI and computer science just kept coming out right? ‘AI is gonna replace our jobs. I was gonna do this’. So I decided to go back to school and learn how to code. If anyone's gonna take away my consulting job and removing reviews, it's gonna be me.
Dan: So you had a fear that your consulting practice was gonna become obsolete?
Curtis: Yes, absolutely,
Dan: That the big hefty fees you were charging weren't going to be so big and hefty.
Curtis: That's right, I got super lucky with a $30,000 Consulting fee. I may have pulled off a few other one or two or three thousand dollars consulting fees for these doctors. But my prices went down considerably to accommodate hundreds or even 1000s of customers. But I realised that the task of identifying removable content is really easy for computers, I mean, computers can easily identify which reviews are removable. Now looking at a profile and going into the behaviour of profiles is a little bit more complex. But well, we already did that. And, and so what we what we did is we programmed a computer to think like I do, to look at reviews, the way I look at them, so that I don't have to look at reviews all day, manually type up the stuff and send it over to these administrators, on the customers behalf, my computer can do it for me, and I get to focus on growth.
Dan: So wait, you're running a brand new consulting practice, and you're like, you know what I'm gonna do, I'm going to go back to school. So where did you go, what happened?
Curtis: I ended up going, I ended up specialising in AI and machine learning at MIT.
Dan: And how many years did that take?
Curtis: So altogether, all the courses in the classes, about a year and a half. It was arduous. Just in full transparency, I did mostly online courses. And these online courses are available through their Sloan School of Business. And they're available to everyone. It's expensive but it's, in my opinion, it's well worth it. I liked the I, you know, over like Stanford and other online programmes that you could choose. I liked what the professor's were working on. And I felt like it, that type of machine learning really aligned with my goals and what I was trying to do with my company. All the other students there were similar to me. They were, you know, executives, at companies, or to kind of tech founders who, you know, who were all there, but they were all very ambitious. And it was a great community for learning how to code and learning how to, you know, function programme computers to, to function the way you want them to?
Dan: Can you let us know what machine learning is in a nutshell?
Curtis: Sure, machine learning is a way for computers to essentially take data from a data set, and then simplify it and really simplify it all the way down to binary zeros and ones. And then after the computer has its way of looking at the data in zeros and ones, it gets to segment and classify the data the way it wants to. We create these functions where we're trying to essentially tell us what we want using its own segmentation. We, of course, can introduce our biases, and we can introduce what we believe this type of data means to it, and we can express that to the server. It's still very new, and new being 25 years old. We just got access to GTP-3 by open AI.
Dan: What is GTP-3?
Curtis: Yeah, so GTP-3 is the probably the most advanced NLP model in the world. NLP is natural language processing, it looks at words and makes sense out of them in order to perform a function. And whether it's a search for a search engine, or whether you want to write an entire book by this NLP, and there are published articles and books by this AI and you can't distinguish whether it's a human or whether it's AI. And it's fascinating.
Dan: As someone who, you know, is on the forefront of deploying this sort of technology, when do you think AI is going to be smarter and more capable than human beings?
Curtis: I'm a big fan of humans, I'm a big fan of being empathetic, feeling the emotions that we feel, whether it's rage, whether it's even unrealistic stuff. I believe that, unfortunately, a lot of people I don't want to say are gullible, I don't know if it's a fault of our society. This is probably a random topic, but I feel like we're easily misled. And AI is not so easily misled. It really makes its decisions based from data. So in my opinion, I believe that AI has already surpassed a lot if not a large majority of you know, human intelligence at least the supercomputers.Some of these FANG companies, and if you're not familiar with FANG, so F is Facebook A is Apple, N is Netflix and G is Google. Microsoft should be included there, Amazon, of course. They have these giant giant super powerful servers. And those servers, those AI models are way more intelligent than most human beings. Once I went back to school for coding, I definitely want to give a shout out to my AI mentor Alex Panay, he's been amazing. But once we got this built, I realised our service could do so much more than figure out if this review is removable for businesses. We realised that we could use this technology to make a prediction on whether reviews are real or fake, no longer for businesses, but for consumers. And we started working on that project about two years ago. We started the company already, it's called ‘The Transparency Company’. The website is ‘ask for transparency dot com’, but we are able to immediately pull all the reviews in for any business around the world and make a prediction on whether those reviews were paid for online by business owners, or whether they genuinely earned those reviews. And those are genuine experiences. because businesses with fake reviews with fake customer experiences are statistically more likely to give you a bad outcome or worse outcome than a business with genuine reviews online, businesses who have earned their experience with their customer experiences online over the years. Whether it's a doctor who's about to cut into you on the operating table, you want to make sure that their reviews are real before that happens. You want to make sure that the contractor who's about to take your roof off actually did that before with us. consumers' before you pay them ridiculous retainers and have them work on your house, because this is going to impact you for the rest of your life.
Dan: I love this. So you basically were terrified of a competitor that didn't exist yet. So you went to online school to learn about machine learning in order to create the competitor yourself that you were worried about. Have I got this right?
Curtis: Yeah, in a nutshell, I went back to school because I was like, if anyone's gonna figure out how to programme a computer to dispute bad reviews, it's gonna be me. The outcome, though, was me realising that computers can do so much more. In 2019, Google removed 75 million fake reviews, that's 210,000 fake reviews per day. And that's just Google. That's not Yelp. That's not TripAdvisor, both of which disclose the amount of millions of reviews that they remove. But in my opinion, that's just the tip of the iceberg. They're still not identifying all the reviews that need to be removed. Getting access to fake reviews was easy, because now that I have the knowledge, I was like, ‘Okay, well, now I need fake reviews for my AI, my ML’. So what did I do? I googled, ‘buy Google reviews’, I googled that phrase, there's like 30 stores where you can buy them online. They're like five bucks, right? So I went to those guys. And you know, there's Skype IDs at the bottom, and I reached out to them, and I said, ‘Hey, I want to buy all of your reviews’. And he's like, ‘What do you mean? Like, how many reviews do you want?’ And I said, ‘No, I want to buy your entire database of every review you've ever posted in your customer list?’ And he's ‘No’. I'm like, ‘Come on, like, it's no big deal. I need this for my …’, I didn't tell him I needed what I needed it for. But I was just like, ‘I'm obviously gonna pay you what can we negotiate’. And after weeks and weeks of me pressing with these 30 different people, I finally got introduced to the one who does kind of half of them, he runs a giant fake review farm. And I wrote him a very large check to sell out his entire database of fake reviews.
Dan: To basically defeat the point of this product.
Curtis: No, I mean, or Objection dot co, there's always going to be fake reviews coming in from untruthful customers. But I'm right now I'm so much more interested in helping consumers solve this problem of who to choose as a better option to serve my life, like because of the problem of fake reviews. And that's what I'm really I'm really energised about and what I really focus a lot of time on,
Dan: Tell me about building the software from the beginning.
Curtis: I hired, at the time, he was a junior developer. His name's Roman. He's unbelievable. He's been with me throughout my entire journey. So I gotta give him credit. And so we set out to solve a problem. The problem was, ‘I just got a bad review. It's illegitimate, right?’ And I don't know what to do. I don't know. I don't know how to dispute it. I'm not computer literate. I'm a business owner. I'm busy. So we built out the customer log and made it easy to log in, making it easy to tell us the name of your business. making it easy to tell us you know what you want? Yes, dispute it. Okay, done. Once you tell us what you want, it's hands off at that point. And our software takes care of everything.So it took us time to get the SDK is you know, to make sure to connect it to your your Google account, if you want to create an account, right, the login the registration, the billing, setting up Stripe, setting up the merchant account, the subscriptions and debugging all that setting up. But the code, you know the code got more interesting as time goes on, because these websites don't want you to take information from them. So they change their code and it's like a cat and mouse game. Right where we'll just call it web scraping. When you when you are collecting information from a website that has a certain script and a certain code, and then every six months, sometimes three months, sometimes three weeks, it's done at random, they change the way their information is displayed. So you have to change your code in order to relay information to your customer that they're used to. So at first, we had to figure out a way to get the data. And then once we figured out how we are going to get the data, what data makes sense, like what what, what data is relevant to make a prediction about review fraud. So we had to put together metrics that it showed signals, and indications of review fraud. So we had to put together a list of all fake reviews, which we have almost a half a million of fake reviews, and a lot of them are still live, a lot of those reviews are still publicly available, and then we had to put together a data set of all real reviews, a data set that we knew were it's not as big, but it's very large.
Dan: You know, we're talking about 10s of 1000s of reviews here,
Curtis: Over 100,000 real reviews.
Dan: What are some of the signals, if you can pull out of your algorithm or whatever that might surprise people or some that you think are cool.Curtis So my, the one that has the highest weight and the and a model of transformer that uses predictive intelligence. It's a metric called ‘the review pod analysis, just like dolphins and whales, fake reviewers travel in pods. They want the highest margin of profits. So they'll use the same accounts to review the same businesses. And you'll see these accounts used repeatedly across multiple businesses and I could show you links, you know, offline if you want to see what they look like. But you'll see the same 20 reviewers reviewing the same 20 businesses. And, as a consumer, you don't see that because you're just looking at that business's reviews. Most consumers don't click on the profiles to see what else is being reviewed. The curious ones do and good props to you if you are, good for you, you should be curious about that. Which is why a lot of review websites don't even allow you to click on profiles, only the most transparent ones do, only Yelp, only Google, only TripAdvisor allow you to do that type of due diligence. The other review sites? You know, most of them, don't they don't let you research the profile, because, well, let's just say they have a lot more to hide.
Dan: Is there another signal you'd be willing to share? That was really cool.
Curtis: Another one we use is called the Distance Matrix analysis, we measure the distance between each of the reviews that were reviewed by a contributor. If you think about it, like an earthquake, how far away are they travelling? And how far away are they from leaving home? Because the average Googler, the average person, has a certain distance where they travel. And then if you look at that data, versus fake reviews, fake reviewers sell fake reviews all over the world. So that distance in kilometres is way off, right? So we look at that behaviour, that profile behaviour, to make predictions, not just the content. And the content we have our own NLP models. And we look for things like keyword stuffing, to help the SEO of that listing or of that internal search. For example, ‘Best Personal Injury Attorney’. If you see a review that says, ‘This was the best personal injury attorney, and just repeats that phrase. Statistically, it's suspicious, you really need to be careful, because that keyword was stuffed into that review artificially. So we look at all this and that all there's like, we have like 20 other ones that are really cool.We have another fun one called ‘turkey in the straw’. And it's probably my favourite. The turkey is a business category, right? A doctor, a dentist, a lawyer, a service professional, a contractor. So garage door installation moving company, right? That's a turkey. Straw is a restaurant? Or are these nonchalant companies that normally get lots of reviews, maybe a park. And we look, we look to see if there's a turkey hiding in the straw. Because these astroturfers are getting smarter. And what they do is they hide a turkey, a paying customer, in the straw to make it look more organic. So we look at the amount of turkeys in the amount of straw to make a predictive analysis as well as look at that behaviour as different astroturfers.
Dan: What are astroturfers?
Curtis: Yeah, an astroturf, or sells fake reviews to businesses. So when we purchased the data from these astroturfs, we see the different techniques that they use in order to bypass the security detection, in order to make things look less suspicious. And when we started seeing these trends, we're like, ‘Wait a second, that makes total sense’, so we create new metrics all the time. It's almost like an arms race, if you think about it - as these big companies evolved to push out spam, right? Because that's essentially what it is. It's fake review spam, their tech, their tactics strange because they've got a very hungry, they've got a very hungry audience. You know, if they're removing 75 million reviews a year, that means they're paying for 75 million reviews a year. You put a $10 price tag in a lot. Usually it's a lot higher, for you know, the reseller who resells this as a service. But you've got a multibillion dollar industry right there. And that's just Google that's not the 100 other review websites that are out there, including
Dan: So people are optimising Google reviews the same way they optimise the search engines 20 years ago.
Curtis: 100%.
Dan: I'm trying to look at the arc of your entrepreneurial journey like you're studying and stuff online, and then you're like writing big checks to these astroturfers and stuff. Is your consultancy doing so well? You can cash flow this software build.
Curtis: Yeah, well, it's fully bootstrapped, paid for by yours truly, didn't accept $1 from an investor. I'm hoping we're gonna make a consumer product out of it. But we already have an engagement with the FTC with the Federal Trade Commission, and we're working on our first government contract that we're super excited about,
Dan: And what does that mean to have an engagement with the FTC?
Curtis: So the Federal Trade Commission is a federal branch of the government that goes after businesses who engage in deceptive marketing practices. That's the literal exact phrase and that's what they fine companies for. We're hoping to be the technology to help the government find businesses that are participating in deceptive marketing practices. The FTC finds a business $250,000 per fake review. In 2021 one of their initiatives is to get more serious about cracking down on fake reviews online.
Dan: Your journey is so by the books, I love it. This idea that your first step was to solve a problem for money. And then you get good at it, you continue to do it, and you cashflow it and then you build software that can do it more effectively than you once did it. What do you see is like the sort of next step in your journey from here.
Curtis: Yeah, so we're about to integrate into Yext. So Yext is a publicly traded company. They have, and I'm not violating anything with the SEC, because you know, they integrate with a million companies, it doesn't affect anything. And we're a tiny, tiny little company, but I always, you know, worry about liability and stuff. But yeah, it's common knowledge that any company can integrate into a third party, app marketplace, right? So that's essentially what this is, just like Zapier. We're about to integrate into Yext, which we're really excited about. Yext has about, I want to say 2.1 million businesses.
Dan: So this is a way that people can integrate your software into theirs, essentially?
Curtis: With the click of a button. Or a few clicks. But yeah, absolutely, they'll be able to add our product as an add-on product to their, you know, to their reputation portfolio that's already working for their company. So a lot of companies offer reputation, like review monitoring, where you can review your company. A lot of companies offer generating feedback, like texting, emailing customers, trying to generate that feedback. But a lot of these reputation management companies and software companies don't offer disputing. It's kind of ugly in the sense that these big companies have always like discouraged and they said, ‘Don't work with reputation management companies who offer to remove reviews’.
Dan: Because a lot of times it is a scam?
Curtis: It's not necessarily a scam, it's their promises. If you look at what their offer is, if they're guaranteeing a removal, if they guarantee the removal, it's a scam because you can't guarantee jack. If they're guaranteeing customer service that's different. If they guarantee your satisfaction that's totally different because that's what we do. We guarantee customer satisfaction. If a customer is not happy, we’ll refund them, no big deal. But you know, you gotta you got to also look at how they get paid. You know, if you're a reputation consultant, you're charging people upfront for review removal, as a business owner, I'd be a little worried like I'd want to pay you after or at least hold that money in escrow and then pay you after, which is what guaranteed removals do. They do a fantastic job and charge about $1,000 per review removal. It's a pretty interesting thing to see across the internet. How Different reputation consultants charge for it.
Dan: Your company has is doing such an intelligent job at solving this problem that you've recognised, this challenge you're sort of posing to yourself yet again, that you might again, eat yourself that people are going to use your software and sort out their stuff, and then stop paying you for it over and over again, like how are you going to address this potential problem you're building for yourself?
Curtis: We switched from monthly subscriptions to annual subscriptions, which really helped because then they get that value of us taking care of their reviews all at once, and we get the annual subscription.
Dan: That's worth pointing out - this is why pipe exists because SaaS on the internet masquerades as monthly payments are the best thing in the world, but not for all businesses, if you know that value is front loaded and SaaS is designed to keep people on for years. That's how SaaS works. Right. Right. So that's why I just want to point that out strategically, not always the best thing to go month to month.
Curtis: I agree. And we realised that about nine months into our journey, our churn was like, ‘Well, thanks for the results for we don't you know, we don't get that many reviews a month. So we're out. And we're super happy with our results, by the way. So it's not because we're leaving, it's just because like, most months, we don't need you’. And I'm like, ‘I get it. All right, no problem, like thanks for your feedback’. But yeah, we noticed our churn was high. So then what we did is we changed our monthly and annual to just annual, the only way you're working over those now is if you pay for an entire year, because you're going to get that value. So now our customers are with us a lot longer. And the ones who see the full year through, we still looking at that churn rate but it's already better than when we were on that month, a month. We're super excited for some features that we have coming up. And I know this isn't your part of your question but ...
Dan: No, it's okay.
Curtis: It’s another technology we didn't really address is the ability for our NLP to automatically respond to reviews that don't qualify for removal that just can't be removed, you need to respond to them, our AI will respond to them automatically for you.
Dan: I tell you what I think the manager of my condo complex writes worse than AI responses to her reviews. Two more questions. One is, you know, you have such a cool story, man, thanks for sharing with us. You have this moment. You're like cleaning up an emergency room, some doctor offers you 30 grand, a lot of us don't feel like we have that moment yet we have that itch that you had. What's your advice for entrepreneurs that haven't had their emergency room moment yet?
Curtis: I feel like at some points, I have these dreams where my life just came on a silver platter. And while that's true, it's also not true. There's a lot of failures that I'm not telling you about right now that I haven't discussed with you, me losing giant accounts, me not being responsive enough, me failing over and over and over to fix those mistakes that allow me to retain customers like I do today. I don't want to seem like I just figured it out everything overnight. For those people who have that drive, that have that hunger, you need to put yourself in front of them. You need to find those people who are in pain with whatever it is that you solve, whether it's a burger stand, you need to go out and find hungry people and have them try your burger. Whatever your niche is, whatever your product is, you need to find out how to get in front of your target audience. You need to figure out a way, a scalable way to get in front of the people who need what you have. And even if you don't know how they're gonna react, what's the worst that could happen? Don't worry about it, get in front of them.Dan Big up my guy Curtis Boyd, in his LinkedIn profile, he calls himself ‘a fake review sniffer’, love it. It was really fun talking to him, check out his site Objection dot co. And as always, interested in your thoughts about this one. Reviews play an increasingly important role in online businesses, so we'd love to hear what you think. And on that line, some news updates. We're planning a Q&A episode coming up. That's right. We will attempt the A's to your Q's. Email us Jane at Tropicalmba dot com or Dan at Tropicalmba dot com. We're compiling some wonderful questions. And Ian and I will get on the horn and give our hot takes and respond and respond to your queries. Also, feel free to Loom, hat tip to recent guest Sam Hill on that one, we'd absolutely love it.One or two other announcements. We've been sponsoring our own podcast this month, but we do have some openings. This podcast has been downloaded over a million times in the last. This podcast has been downloaded over a million times. And this podcast has been downloaded over a million times in the last 12 months. If you'd like to connect with our audience, feel free to reach out to me Dan at Tropicalmba dot com we'll see if there's a fit there. We're pretty picky about who we partner with. We want to make sure it's a company whose products the audience is going to be really jazzed about. And for those of you who are business owners thinking about getting in front of the audience. It's a pretty smart crew of established business owners. The vast majority of folks who listen to this pod are location independent entrepreneurs. And those of you who aren't yet you found yourself to the right party, you're on the right track.Speaking of getting on the right track, over the next month I am offering a limited speaking of getting on the right track. I am speaking if you're on the right track over the next month I am offering a limited number of podcasts strategy phone calls. Again, our pod has been downloaded millions of times and countries all around the world and we've sold millions of dollars worth of product off of the pod. If you want to brainstorm with me about starting a new show or rebooting an old one, or just anything podcast related a few things that would benefit you. I have a limited number of slots available click through to this episode. I'm tinkering around with our DJ platform, which in part allows people who sell services to post offers. I'm tinkering around with my own offer. So I want to try this podcast strategy call offer with you all here today, see how it works and I'll let you know what the results are. That's it for now. We'll be back as always next Thursday morning. 8am Eastern Time.[/bg_collapse]