Connected Commerce: If You Can't Train a Human, You Can't Train AI

Hai Mag 00:09

Hi everybody. Welcome to the Connected Commerce podcast. And your host, Hai Mag. I'm the founder and CEO of Eva Commerce. we provide our Vueria software and connected commerce to TikTok, Amazon, and Shopify customers. Now, today I have Faith as my guest. It's a distinct pleasure to have her on our podcast.

Faith Denniston 00:33

Mm-hmm.

Hai Mag 00:35

What I really, really like about her is besides her name, which I really like, but I mean she is like a real real Amazon expert because now

Faith Denniston 00:46

Mm.

Hai Mag 00:47

on LinkedIn pretty much after two months everybody is like Amazon expert, senior PPC

Faith Denniston 00:53

Yeah.

Hai Mag 00:54

manager. And and here we are talking about someone who almost started like Almost like a ten years ago, right? Faith and now she

Faith Denniston 01:03

Yeah, it's ten years exactly with the Amazon stuff, yeah.

Hai Mag 01:06

Okay, ten years. and maybe a last thing is now she also created a company. She's the co-founder of ROI Renegades. Seems like

Faith Denniston 01:17

Mm-hmm.

Hai Mag 01:18

they help Amazon sellers to get some more ROI out of Amazon. So Faith, it's

Faith Denniston 01:24

We do.

Hai Mag 01:24

please tell us more about yourself and the company.

Faith Denniston 01:27

Yeah, absolutely. I'm Faith, as Hai already mentioned. I have ten years as of twenty twenty six experience in the Amazon space. So I started out as an employee at Amazon. was in their social media department. so I was getting all kinds of cases from the social side of things. It's actually one of the things that I recommend to our clients if you want something done. but they're not as good as they were when I was there. So from there I started working with a small company who sold on Amazon because they offered me better pay than Amazon. and they wanted my Amazon expertise. And then from there I was able to build my career. worked at a couple of agencies, and now my co-founder Jason Mastromatteo and I have ROI Renegades, which primarily focuses on any issue under the account health umbrella. So as a former Amazon employee, I know how to talk to Amazon employees, even if they're even if it seems like they're a robot. you kind of just have to speak to them in their language, right? And that's where my first connected commerce story is gonna come in. I'm just gonna use an example from whenever we had first started. one of the main things that we do under the account health umbrella, one of our I would say it's the top service we offer is a brand name change. Amazon's policy technically says this isn't possible. So a lot of companies will hire us or a lot of businesses will hire us to make that argument to Amazon to change the brand name. And we are successful with that. There and but there's not a secret sauce with the success, or I guess there is, but it's hiding in plain sight. I'm about to tell you what it is. get really comfortable with Amazon's help pages. That's all you have to do. we are successful in our push to get any sort of account health thing fixed. Sorry, I that was a super professional way of putting that. but in pushing to get any sort of issue fixed under the account health umbrella, the number one thing that you want to do is quote Amazon's literature back to them. So their brand name change page technic has like a technicality on it. If there's a legal issue or something then you can do it. So you just that's one example. You just have to find the technicalities on their help page on their own published literature, quote it back to them enough, and you'll get it done. it's still a process, which is why our company is still successful, because Amazon is still unwilling to be helpful to sellers, but that's fine, because as long as they're unwilling to be helpful, then us renegades will come in, help you out and help you recoup that lost ROI from a stupid listing of deactivation or something like that. Because that's another thing we do as well. We have like a lot of listing inks and, you know, working in the Amazon space, I'm sure you've seen a number of stupid reasons that something would get taken down.

Hai Mag 04:27

Absolutely. Like, I mean, and the there there are so many reasons for that. But I like this, like the brand name change that you mentioned. You know, that's like an address change. sometimes it happens, especially

Faith Denniston 04:40

Mm-hmm. Yes.

Hai Mag 04:42

in the beginning when you start like trying to create a company where the owners are outside of US and I mean there will be so many other things and problems, but there's always

Faith Denniston 04:55

Yeah.

Hai Mag 04:55

a way to sort you know solve these. things. w a question that I'm kind of interested like what is the success rate of ROI Renegades to fix Amazon problems?

Faith Denniston 05:08

That is a great question. So I have here, I'm I'm not gonna share my screen or anything, but I'm gonna I've sorted our tasks into what's been completed and what we haven't been able to complete. we have About ten we weren't able to complete. one was or a couple of them were from missing reviews. So whenever we first started out, there were a lot of people who reached out to us because they saw some of their reviews were missing. We've you know, we did our tried and true, like, hey, here's what the hope page says, like where our reviews, but for some reason Amazon, I'll say for some reason, but we all know how Amazon is about their reviews. they're they are really unwilling to budge on that issue. So we do not have a lot of success with review issues. one other one I see that we were able to resolve.

Hai Mag 06:06

Yeah, but tell us more about like, I mean, it seems like you know, Faith, like you're not the the sales guy, because I asked like what is your success rate, and you immediately started with like telling what was not fixed, which is great by the way. So you're already sharing us, but I'm sure you Yeah, like I mean, but most of the salesy people

Faith Denniston 06:21

yeah, no, I start with the bad news. Like bad news is we can't fix this. But as opposed to as opposed to these 10 issues that we could not fix and they were all missing review related, we have fixed hundreds, hundreds of issues. we've been,

Hai Mag 06:36

Exactly.

Faith Denniston 06:38

yeah, we have been a company for three years. and yeah, definitely not the salesperson because that's something that we've heard that our clients like about us is we don't We don't really BS or anything. Like we will tell you straight up if we think we can't fix anything. So that's kind of where my head is at whenever somebody's like, what's your success rate? I'm like, Well, here's what we can't fix, but here's what we can. So I would say if I had to put that into

Hai Mag 07:05

Exactly.

Faith Denniston 07:06

an estimation, I would say that we are we have a success rate of about ninety-seven percent. If I had to give it a number. But now you have inspired me to actually

Hai Mag 07:15

I think that, you know, that's an amazing yeah. I think it's amazing.

Faith Denniston 07:17

Yeah. You have inspired me to actually run the numbers. Actually Jason did once and he put it on his LinkedIn, but we do have a really high success rate. Yeah, he did.

Hai Mag 07:25

he did? Okay. So, like, I mean, and that is I think very important. I you know, here like the success rate is something, of course, that is important. But what I can say to, you know, our you know, listeners at the end of the day, maybe you don't buy a service like this until you hit to the problem. You don't know it. At least you know that there is a team.

Faith Denniston 07:50

Mm.

Hai Mag 07:51

Which can fix it when you hit you know to a problem like

Faith Denniston 07:55

Yeah.

Hai Mag 07:55

that. And when I was running our own pretty much a nine-figure seller account that was five years ago, and at some point the the store was suspended, and that was not

Faith Denniston 08:13

Mm-hmm.

Hai Mag 08:14

a legit reason. Now we fixed that in six days, but Given that it was a huge store in six days, think about how much money we lost. But still

Faith Denniston 08:27

yeah.

Hai Mag 08:27

we were so happy at the end. but I mean that six days probably I lost maybe six months of my life or whatever, you know, like thinking about it. So

Faith Denniston 08:39

Yeah.

Hai Mag 08:39

it is important to know that there is a team and they can solve

Faith Denniston 08:44

Yes.

Hai Mag 08:44

the problems. One of the things also I am also curious about like you know like Amazon is using more and more and everybody you me everybody is using AI meaning that

Faith Denniston 08:57

Mm-hmm.

Hai Mag 08:58

more and more even instead of like normally you you know you talk to a an Amazon guy in India sometimes maybe in Costa Rica I mean there are a couple of locations for that but I'm kind of thinking that more and more these real people are even maybe vanishing and AI is coming. And by the way, maybe it is better. Maybe not. What is your view if you look at the last six months versus the previous years and how this really works?

Faith Denniston 09:32

Yeah, and you can see that too whenever you start a case with Amazon. it's all very automated now. They want to give you an automated answer before they connect you with a quote unquote person. because they also want you to go on live chat. So that could be them training their AI agents. Now if they I'm not worried about AI in the sense that I think it would take over mine and Jason's position, like if we had a competent seller support that could have got that account reinstated in like two hours instead of six days, let's say. I don't think that's ever gonna be possible for Amazon to replace that because there's so many nuances within a business too, right? So that's kind of where Amazon's system fails in the first place, is they try to have AI detect something wrong with the listing. And something might technically be wrong with the listing, but it's not like we need to close down your whole store and shut down the whole listing and all this. Like it might just be a minor typo or something like that. which is something that we have dealt with before. I won't get into it. I know we have limited time, but yeah, so I don't I do think that, you know, AI can be efficient if it is trained. But because Amazon can't even I'm gonna I'm throwing a lot of shade at Amazon, but Amazon can't necessarily train their own employees, in my opinion. So I don't think that they're gonna be able to train an AI agent on how to handle seller issues if they can't teach a real person. And I I hope that doesn't sound old-fashioned or anything like that. But so that's the beauty with AI, right? The reason that we like it. The reason that it's so helpful for us is because it takes something and it makes it simpler. but it's only able to do that because it knows how to do that task. Like it knows how to do the math, it knows how to create the spreadsheet, it knows what metrics to look for to determine if a campaign is a good campaign or not. but it does not have the knowledge or the nuance or the capability to read into a situation like an account of reactivation or a brand name change or something like that. I hope that makes sense. What I'm trying to say.

Hai Mag 11:58

it makes perfect sense by the way. I mean, I am an AI developer for more than 20

Faith Denniston 12:03

Mm-hmm.

Hai Mag 12:03

years, meaning that I was developing AI when AI was not even existing in a way, you know, like in the way that it

Faith Denniston 12:09

Yeah, before it was even cool. Mm-hmm.

Hai Mag 12:12

was it's here. Now, there is absolutely, by the way, no difference, even most of the algorithms were existing 20 years ago. But the thing is, I always sell even my people. People work with people, people Buy from people, people love people. Secondly,

Faith Denniston 12:28

Exactly. Yep.

Hai Mag 12:31

people are the ones which are creative, they create a new way of doing things, but AI shows you what you don't see, but it's already there, but you don't see it, but AI sees it because which means that AI needs to be trained, and AI is not authentic, people are authentic, not all of them, but some of them.

Faith Denniston 12:52

Exactly.

Hai Mag 12:54

And authenticity, creativity, and the customer people connection, the human connection, none of these three can be replaced by today's AI tech. Well, I don't know about tomorrow's AI tech, but today's

Faith Denniston 13:10

Yeah.

Hai Mag 13:11

AI tech is not a replacement. In that sense, back to what you say, if the training cannot even train a human, it would not also train.

Faith Denniston 13:20

it's my mm-hmm.

Hai Mag 13:23

The AI. So it is it is like a logical statement.

Faith Denniston 13:25

Yeah, exactly.

Hai Mag 13:27

I mean, it makes a lot of sense. By the way,

Faith Denniston 13:29

Yeah, like if you cannot train a human, you cannot train AI.

Hai Mag 13:34

Yeah, like I mean, because it's the way it works. You need to have a really good data set and instructions and everything in order to train AI

Faith Denniston 13:43

Exactly.

Hai Mag 13:44

in the right way. By the way, for the audience, both Faith and I being Hai and Faith, we all have AI

Faith Denniston 13:51

Mm-hmm.

Hai Mag 13:51

in our names. So we are ready whenever AI takes over, we're gonna be like, Hey, we are here, no worries. Kill this human, we don't care about that. You know, they don't have AI in their names.

Faith Denniston 14:01

We are the forerunners. Yeah.

Hai Mag 14:05

We are with you. Like, I mean, that's what we're with the minority. I mean, so well, I mean, that's that's all true. And

Faith Denniston 14:14

Mm.

Hai Mag 14:15

I think that you know, like but what is exciting here is if you as a human, if you are creative, if you are authentic, if you can communicate, and on top of it, if you have AI. Then you can do a lot of things on your own. Like you don't need five

Faith Denniston 14:32

Yep.

Hai Mag 14:32

more people anymore. For example, that's what I like. Now, you know, I am like three to four times more effective compared to last year. Now it's a personal effectiveness, you know, kind of measuring measuring that.

Faith Denniston 14:48

Right.

Hai Mag 14:49

I am very happy with that. What about your second story, Faith? Like maybe you can tell us more about your experiences.

Faith Denniston 14:58

Yeah, so my second story is not as riveting as the first one, unfortunately, but it's just kind of taking it back to basics. so we talked about how AI has all the data and can help us make decisions a lot faster. but and we talked about how some things need a human element. So that's Bringing it back to what we talked about earlier, whenever I say that we have a few clients we do full service management for. So like I said, this is not extraordinary, but this is just something that I found works. all I take is a Google Doc and I write down why I did what I did that day for the advertising. I know it's very it sounds it sounds a little silly, but hear me out. You know, like we we can get a report that's like, Hey, this campaign's not doing great. you know, it's got XA costs, no sales, etc. we can go into that campaign. Most of the time, you know, like the AI's right. we don't use AI, by the way. just like we go off like Amazon's AI recommendations. So that's why I said it's not super extraordinary what we do in terms of advertising optimization. It's just us. Amazon will suggest something, we'll look into it. And that's another thing that drove my response for the last thing. If Amazon can't train their tech then or train their people, then they can't necessarily train their tech. I will admit that Amazon's ad console is definitely more refined than other parts of Seller Central. but that being said, it's missing that nuance. And we bring it back to the human connection, right? So in just writing down like, hey, this campaign was not doing great, but also there were like a hundred keywords that were wasting $75 worth of ad spend. We cut that out, and now your campaign's doing great again. some I've seen in the past working with other agencies where a campaign like that would be flagged, which understandably so, of course, but then you don't like look into the nuance of why that campaign was not performing well. Like it's just the number. And that's the number one question I would get as a brand manager as well: is why did you do this? why? was this action taken? What was the thought process behind that? And I didn't always have an answer. I'm not gonna lie to you, because at agencies, that work is typically delegated to another advertising team. Nothing wrong with that. but what we like to do here at our rent ROI Renegades and why again we only have a few clients is we really dive in and we really get in there. So whenever we do get those questions like, why is advertising performing this way? We can say, well we tried this and it didn't work or Well, here's what's going on with this campaign, here's what we did, and here's why it worked. I hope all of that makes sense.

Hai Mag 17:52

Absolutely right. I I love it, you know, in in

Faith Denniston 17:55

Mm-hmm.

Hai Mag 17:56

I'm also thinking that now, you know, you're talking about like writing what you are doing. And one of the practices that I am started having right now in the last thirty days, I talk to AI. So even I do, you know, for example, when I'm driving,

Faith Denniston 18:11

Mm-hmm.

Hai Mag 18:13

I am coding too. Or when I am on something,

Faith Denniston 18:15

really?

Hai Mag 18:16

yeah, like because a lot of things can be done now. Like, you know, I I because I mean I can, you know, basically hold a lot of things in my mind, meaning i while I'm driving, I think about all these things together and I I give some comments to AI to continue the development, you know, like I mean, do this, do that, and I just talk to the AI. And by the way, talking. Is also, isn't it like the

Faith Denniston 18:40

Mm-hmm.

Hai Mag 18:41

natural way? Like if you think about you know the keyboard and the the interaction, it is not exactly what is natural, it's only available for the last hundred and fifty years. Like before that,

Faith Denniston 18:54

Right.

Hai Mag 18:54

people were talking or they were writing,

Faith Denniston 18:56

Yeah.

Hai Mag 18:57

and not typing. Now we can talk to AI, like, think about like you know, you you can already taught tell AI, like tell the system. What are the reasons why you did something and when you need it you can get it back? Like basically I'm trying to do it even in the daily life.

Faith Denniston 19:14

I see what you're saying, like keep that as a log. Yeah. Like instead of typing it down, just tell the AI, Hey, so I looked at this campaign today, here's the actions I took, this is why. Yeah, no, I I see what you're getting at.

Hai Mag 19:27

And even if you want, like what I'm doing is I'm using right now because things change almost every month, but at this moment I'm using the Grok Bot for that reason, which is from SpaceX, right? Tesla

Faith Denniston 19:42

Mm.

Hai Mag 19:43

SpaceX, and it's working amazing, you know, like it's really you know, keeping all the notes and everything about the business. So I know, for example, I fired this guy, but that's the reason. You know, sometimes you forget.

Faith Denniston 19:55

Yeah. It's like, wait, why did I get rid of him? Yeah.

Hai Mag 20:00

The reason is no reason. I didn't like him. No, no, no, that cannot be a reason, you know. So things like

Faith Denniston 20:02

Right.

Hai Mag 20:05

that. But I mean, that is so cool. I mean, what I like about you know, 97% success rate on Amazon is important, but you also know that if Faith is not able to fix your problem probably you your problem is not fixable. So meaning that, you know,

Faith Denniston 20:26

Yep, that's yeah.

Hai Mag 20:28

the last person like you can call and check the account, you know, can be Faith or other ROI Renegades. And and I think it's important to know that. But also like I mean I always appreciate, you know, you know, the you know the the the work that you know

Faith Denniston 20:47

Mm-hmm.

Hai Mag 20:47

you guys are doing for the customers as I also heard it from the customers too. So thank you very much, Faith, for being our guest today and thank you for the two stories.

Faith Denniston 20:58

Well, thank you for having me. Of course. Thank you so much. Hai. It was great being on. And I'll catch you later. Thank you.

Hai Mag 21:06

Thank you.