Loading summary
Transcript48 lines
- [00:00]
A
Hey, quick heads up. If you're marketing to marketers, this is where you want to be. Sponsor Perpetual Traffic and get seen and heard by thousands of seasoned marketers, CMOs and agency owners. We get hundreds of thousands in downloads every single month, all to marketers. So go to perpetualtraffic.com to apply for a spot on the show. Check out the link in the show notes to apply.
- [00:22]
B
We have a jewelry brand client that, that specializes in secondhand sales of really expensive pieces of jewelry. As an example, a lot of times their transactions are 50, $100,000 or more. And one of the things that we noticed is that their abandonment rate was unusually high.
- [00:40]
A
So now in the AI realm, obviously there's probably going to be some skepticism. How would you overcome that?
- [00:46]
B
So what we did is this is something we would have dreamed of doing years ago, but we couldn't. You're listening to Perpetual Traffic.
- [00:56]
A
Hello and welcome to the Perpetual Traffic podcast. This is your host, Ralph burns, founder and CEO of Tier 11, alongside for show number three with the goat of CRO conversion rate optimization. When I first met this guy, I was like, I have never seen anyone do CRO and conversion rate optimization and how he thinks like he does when it comes to this kind of stuff. So if you are stuck with landing pages, websites, collection pages, PDPs that are just not converting. Well, today's show is all about you and also the future of CRO, which is something that Ned's been keeping this as a secret from me for a little while now, which I'm kind of pissed off about. But anyway, it's going to be reviewing it here today. So welcome back to the show. Head of CRO over tier 11, Ned McPherson. Welcome back.
- [01:57]
B
Thanks for having me. Yeah, exciting stuff. We'll talk through today. So some pretty cutting edge things.
- [02:01]
A
The biggest question that everyone sort of gets is like, how are you incorporating AI into your business? Oh, we're an AI centric business. Oh, we're this and that. Well, we're going to be revealing some things today in the latter part of the today's show that I've never seen anyone do, which is insanely great. So anyway, just to take us back, we've already done two episodes here. The first episode was all about ideation, sort of design, design centered CRO versus what you do, which is really just analytics. It's analytical. The real big difference to that. So we're going to leave links in the show notes for that original show. So if you haven't watched that or haven't listened to it, definitely Check that out in the show notes over perpetual traffic.com Our second show we went through the audit process which, you know, we can actually do this in a very, very automated way. But also we are offering free audits for your website over@tier11.com CRO this is going to be an in depth analysis. We do a screen share of that and the second show that we did here today. So definitely check that out if you want us to give you a hand on what we think is the best way to optimize your site. Tour11.com/forward/CRO so today we're going to be getting into some of that predictive AI agentic stuff. But anything you want to add to shows number one or two. I know we kind of, both of us are talk fast and we went through a lot of stuff there. But anything to add to those two shows which I think were pretty chunky to begin with.
- [03:31]
B
Yeah, Maybe not to add, but just to reiterate because they're so critical. Right. And again, maybe give folks that the TLDR if they weren't able to watch the first or second which is, you know, starting with just kind of the premise like what is CRO? Obviously we all know it as conversion rate optimization. Candidly, it's kind of a catch all for all things optimization on whatever the landing event is that's typically a website, right? Like an E commerce website, maybe a landing page, it might be a form capture, right. For more like B2B brands or service oriented brands, could be a web application. But any sort of like digital landing event, which is everyone spends so much time and effort and money and rightly so out there trying to kind of wave the proverbial digital hand to be like, hey, we exist. This is our value proposition. Would you like to take a look? And for those people who say, okay, I'll give you a moment, I'll take a look. Where do they land? Right? Wherever they land, to the degree of which you can optimize that experience can have massive impacts on the business, right. Not the least of which is obviously a higher conversion rate means just straight more revenue, right? Better margins. But you also get this great feedback mechanism that goes to ad engines, et cetera, that goes to affiliates, that goes to partners, whatever, where when it converts better, everybody up and down the chain kind of sees the value chain, kind of sees the implications of that, right? The benefits of that. And it's kind of one of the tides that lifts all ships. It also has a big play on aov, candidly and that's specific to E commerce brands, right? So I know it's conversion rate optimization, but candidly we spend pretty much just as much time focusing on how we can increase the average order value, right. And then as kind of a tertiary focus, things like viral coefficiency. How, how can you turn your, your web asset, your website, your landing pages into a referral mechanism, right. Or into something that has an element of virality to it or how can you also encourage repeat rates using automations and typical lifecycle? And we're going to talk a little bit about that with AI and such. So we talk about CRO. It's really more of this all encompassing post click optimization. Once you've got someone to click through and they say, hey, I'm willing to at least hear you out, brand as to what it is you're offering to see if it's. For me, that is where CRO begins and everything relating to that is under the auspices of, of a CRO team. So that's kind of point one. Point two is what are the fundamentals of CRO like the how to run it. And unfortunately the majority of the industry has done it the wrong way for quite some time. Which is, you know, the kind of adage I always use is most folks in the CR industry will look at a site and they'll look at, you know, the structure of the homepage and the navigation and the PDPs in the cart and what have you, and then they'll look at competitors, right? Or they'll look at sites that they think convert well because they'll be like, oh, it's really pretty, it's really well done, it's a leader in their industry and they're kind of just making the blind assumption that that site does really well analytically. Side note, a lot of the brands you may be thinking of, I've had as clients or have as clients and they're not always as converting as well as you may think, right. Despite some of their expert designs. But that being said, they'll, they'll do a design centric approach, right? So they'll be like, okay, this competitor of ours, or maybe it's not a competitor, this brand that we look up, it's adjacent in a different industry, but somewhat similar, has a really unique PDP architecture, a really unique cart experience. Let's just try copying that, right, As a shortcut to get better conversion rate. That's a really dangerous approach, right? And it's not even dangerous because it's just inherently wrong and not empirical. But it's also dangerous because if you do that enough, you're actually just like blindfolded, eventually going to find a win. Not because it was the right idea, just because you, you know, you put a blindfold on and shoot enough times, you're eventually going to hit the target just out of sheer luck and the power of numbers. And that creates a feedback loop that you did it right. And then you kind of embed yourself into this process of just copying what others are doing, which is not a good decision. There's also the element of like, I've seen many brands who come to me and be like, hey, we want to run some CRO, let's start with the homepage. That's what we want to do first. It's like, okay, let's just take a look under the hood. Only for us to realize that their homepage is arguably, you know, I'm using this as an example, the best performing asset on the site. It's like all the other areas need work, that's the last place you want to be working on. Interesting. And that all kind of boils down to the main thing here, which is to do CRO properly, you have to be just as invested into data analysis as you are ideation, design and development. So the crux of what I built out over the year and the, you know, the years and the great teams that I've worked with is everything is empirical, everything is data driven. So, you know, people come up with ideas and the answer is, yeah, you might be right, might not at all, it's hard to say. But let's dig into the data and let the actual empiricals of the intra funnel rates, the index rates, how certain people are interacting with the site, how does that translate to what we need to focus on? Like where in the site is weak in terms of its performance, where is it strong? And then why? And then we create hypotheses to effectively challenge that. So, like, this is the classic example in statistics of having like a null and an alternative hypothesis, right? So we're creating alternative hypotheses and we're measuring those up against a null hypothesis to determine that this hypothesis that we have as to why users are or are not doing a certain thing is correct. And then as we prove those out, it's really just rinse and repeat, right? Just keep doing that. So we get a lot of credit in the industry for some of the unique ideas that we have. And don't get me wrong, I don't want to disparage the ideas that we have But I'll be honest, the ideation is not that hard. It actually becomes almost obvious when you do the data analysis. Right. So based on that, and again, we went in this and the second podcast that we did where we started to actually go through some almost like case study examples with the audit to call out some interesting things there. Right. So have a couple of anecdotes as well. And so all that kind of boils down to, you know, the third stage here, which is, where is CRO going? Now, obviously, AI is the talk of the town. Everything that artificial intelligence can do, agents can do, agentic architecture can do. And so it's like, where is CRO going from here? We're going to talk a little bit about what that is. As a quick preface to it, AI is magnifying the level of ideation and execution that we can do. Right. So it's bringing to the forefront, we can speed up analysis, we can find new interesting data points quicker. We can also ideate more broadly, we can open the aperture on ideation. And most importantly, some of those really interesting ideas that we came up with, like yesteryear, if you will, that we couldn't do because we're like, man, how do we even pull that off with AI, we actually can. And I'll give the example of the, the, the jewelry and watch brand to, to maybe start that off. But that's the genesis of where we've come and kind of where we're going
- [09:41]
A
now makes a lot of sense. I mean, I guess CRO for me is. I mean, I introduced it as CRO, but I really look at this as after the click. It's like whatever happens after the click, it's not necessarily conversion rate optimization. At the end of the day, you're really doing revenue optimization. I mean, that's really it. So it's anything that happens after. And it's using those analytics, like you said. And we went through this and obviously we'll leave links in the show notes. If you missed the second episode of this, where we're going to using empirical data, it's like it's data analysis. It's not like, oh, well, Ford Motor Company uses an orange button. You know, when somebody wants more information. That's not necessarily the best premise, but it's actually what's happening on your site. And to your point, it might be the homepage, it might be the PDP page, it might be the collection page. And this is a question that we've gotten a fair amount since the first couple of episodes come in. People are asking. There's a lot of E commerce focus here. I just, I think it's important for you to talk about this maybe a little bit further before we get into the AI stuff. It's not just E comm. E comm is kind of the easier examples that we give. But the breadth of experience on after the click is far greater than just E com alone. So maybe you can expound on that a bit.
- [10:58]
B
Yeah, no, of course. So I mean let's take some like, you know, lead gen brands as an example. Even some B2B style brands. I mean one of the classic things that I'll see is just on form optimization. So this is, this is nothing to do necessarily with like an E commerce experience. It's just a simple. How do you reduce the friction of getting somebody to fill in their information so that they can quote, learn more. Right. Or book a virtual appointment? Some very, you know, kind of just benign thing. It's not, it's not really that big of a commitment for them. But like how do we do that? And like, here's just an example. One of the classic things that I see is just this very clear inverse relationship on a form field with the number of things you ask of a user and the conversion rate of that form. And so a lot of times as an example, I'll see people be like, hey, what's your first name? What's your last name? What company are you with? What's your phone number? How did you hear about us? When's a good time to talk? Who referred you? You know, and it's like, I know those are all things you want to know. It's way too much to ask all at once when someone's just trying to say tentatively, keyword there, tentatively, I'm willing to learn a little bit more. It's like the fundamentals you need, like their name, their email and that's probably it, you know, in certain industries maybe need a little bit more. Another classic one is where people have like a required text. Whether you force the user to type out, you know, what are you interested in learning about? And it's like they're interested in learning just what you have to say. Like don't force them into typing this answer out. It's like, why would you know? That's that's such a high friction experience. And again, you know, these things we can get to. That's the point is when you're getting into like form fields and such as opposed to E commerce, which is a singularity of a transaction in this Case it's like you're going to have a follow up call. You can ask those questions on the phone, like telephonically. You can have like and build a rapport. Um, don't, don't bog down the conversion there or like another example is I've seen brands be like, okay, step one, let's ask them their name, step two, let's ask them their email. Super basic. Combine those two steps, you'll see a boost in conversion rate. Really basic, right? Really simple, really straightforward. But a lot of this again is just fundamentals. And then another one is to your point that's beyond E commerce is, you know like getting a form capture is often just the first part of the battle. Like we've had a number of clients in like the healthcare space as an example where you're, you're booking an appointment to like some medical appointment or something like that or even a virtual whatever. And the forms are okay, they do a decent job, they have decent conversion rate but in between when someone booked it and when the appointment is, there's no follow up, there's no automation, there's nothing there. And because of that the gains that we get on the conversion rate of the page are lost because of the lack of communication. And then because if people don't show up to the call, it's like who cares how many leads on the website you got? Right. And so to your point, like revenue optimization is it and like this is just one small piece. So, so the whole concept, you nailed the head. It's like post click optimization from A the way to Z where like revenue actually exchanges hands.
- [13:43]
A
Right, Right. I always bring this up, I brought it up I think on this show, maybe not in the last year or so. One of my first clients when back when I used to do after the click and CRO, what we got, we got this client, virtual bookkeeping company. Now there's tons of them. Back then there was only a few. We got them 3200 leads and we were fired after the third month. And I'm like, well why? Like we've given you so many leads. And then I actually filled out the form and I realized nobody was following up. Once I actually put in my name, my email and my phone number compulsory like three form fields. Right, right there. And that was the reason why so oftentimes that's the issue. One of the things, before we get into the AI stuff, a pressing question I want to ask you is I was just at a $17 billion home builder doing a full day session and we looked at all their forms on all their individual sites. There was no consistency to it. Some just had first name email, some had phone number, some said what's your price range? What are you looking for? What town are you in? All of that. I asked them when was the last time you actually split, tested your form? They all blank stares for everyone in the room. Sometimes when you're in a service based business and your lead gen is your first step, is that oftentimes maybe the first question that you might ask. I mean obviously the, the data is going to draw you to that but I mean I find so many businesses have just never even tested the most important thing on their site, especially if you're a service based business or a local business. Is that.
- [15:15]
B
Yeah.
- [15:16]
A
What's your thoughts on that?
- [15:17]
B
Yeah, well a lot of people fall into this fallacy that it's like the utilitarian steps are not to be optimized with because it's like, well if you have the intention to fill it out, like just fill it out. Right. And like here's a real world example of that is that I, I think it was, I may be giving them unfair credit but I want to say it was Sephora that made this famous which is most checkout counters in most retail stores. You go, you browse around the store, you grab your things and then there's a sign somewhere that says checkouts over here. You go up to a counter and you do the utilitarian transaction. And that's pretty much it. But if you look at like Sephora and again I don't, I don't know if it was them. I can't remember if they were the first ones to do this. But if you ever go into a store like that or like any beauty store does this now they walk you through this like maze gross to get. Which is all.
- [16:00]
A
Yeah. Greatest that they're all sample sizes which lead to future purchase. It's absolutely brilliant.
- [16:06]
B
I was just with my wife about
- [16:08]
A
this this weekend as a matter of fact.
- [16:09]
B
Exactly. And so it's one of those things where like some enterprising individual and again somewhere at some point was like can we improve upon this otherwise highly transactional experience? Right. And again that's the fallacy is most folks just fall into like oh, the job is done at like the add to cart level or the pull it off the shelf level. And once they've done that, that's the trigger that intent is there and they're going to follow through unless there's some weird stimulating circumstance and it's like actually did you know most forms have like an 80 to 90% minimum abandonment rate? So it's like quite the opposite. It's like most people getting to the form is, is a small win, don't get me wrong. But like that's just the beginning. Getting them to complete the form and is a big undertaking and a lot of people just abandon it. You know, they procrastinate, they're lazy, don't want to do it, they forgot the value propositions. Why am I even doing this? They question the form, maybe the privacy factor. They don't want to give all the information. Da da da da da. So, so, you know, back to it. It's like, yeah, I mean even the utilitarian steps that seem innocuous, that seem kind of just transactional really require a very keen eye to look at and make sure that you're doing the right things.
- [17:15]
A
Yeah, it's so true. I mean it's just man, I think a lot of this, and maybe we can get to it in the next section here is like a lot of it does come back to sort of asking those simple questions of your audience, like, you know, why are they here, what are they looking for, that kind of thing. So let's get into the AI part of this, which is really exciting. Now everyone's talking about AI these days, but some of the stuff that you and the team are doing here is absolutely insane when it comes to like the future of after the click, the future of CRO. Let's talk about some of this. Some of we can't really show, I believe some of this is a bit more confidential so we can't do this on a screen share. Although we'd love to. Maybe we'll have you back and actually show it in in person at a later date. But this stuff is pretty mind blowing. So let's talk about that a bit.
- [18:03]
B
Yeah, you bet, you bet. So a couple things we want to talk to and I'm going to give you an example of one that's, that's really exciting. So we have a, as you know, a Julia brand client that, that specializes in second basically retail sales, secondhand sales of really expensive pieces of jewelry. As an example, high end watches, you know, the Rolexes, the AP tanks, Cartiers, Pateks, etc. Right. And so a lot of times their transactions are 30, 50, $100,000 or more. Right. And so their average order value is huge. Right. These are big ticket item purchases. And one of the things that we noticed looking at the brand when we first had taken them on is that their abandonment rate was unusually high, their exit rate kind of mid funnel and you could see that because you know, add to cart rates were really low, which was going to be expected. But interestingly the add to cart rate was much better on like secondary and tertiary visits, not on primary. So all these data points are to add up to like what is a kind of obvious hypothesis which is people leave the site, they find an item they like and they leave the site once they found it to see if they can find a better price shop. Yeah, exactly. On a different online e tailer, right, that sells this high end item. And again these aren't things that they just buy on a whim, these are things that they've been thinking about. These are like serious investments, high consideration purchase. Yeah, big time, Big time. And so we basically validated this talking to some of their users and customers as well. And what those folks are doing is exactly that they were going to shop around a price point. So what we did is this is something we would have dreamed of doing years ago but we couldn't is we used a AI feature that effectively every like intermittent hour or two hours will scour the web, right, for all available public listings that match that item and that variant and will determine the price point and say there's 10 compet competing variations of that exact watch, right, that they're selling and it'll look at all the price points of it and then it'll state what they are on this site, it'll state the low, the high and then where they're priced at and what percentage above the mean or the median that it is. So basically it's a feature to be like hey look, I know you're already thinking this, you don't need to go do your own research, we've already done it for you and packaged it nicely here. We are not the cheapest, we're not the most expensive, we're X here. And then that lets the user determine if this is a good buy for them or not. So yeah, pretty interesting feature, right? That again, without A.I. you just, I mean would be no way. Yeah, I, I borderline say it's impossible but like at minimum would be excruciatingly difficult and take forever to build. But like with AI we can do that now and it, you know, didn't, it didn't take as long as you think to build that feature out.
- [20:35]
A
So yeah, there was something like that in the past I would imagine the biggest. First off, way cool the fact that it's actually using real data. But being a former affiliate marketer, I used to do things like this in the past based upon made up data. So now in the AI, no, now in the AI realm obviously there's probably going to be some skepticism. But the point is, is your secondary, your tertiary visits, as far as conversion rates go, indicate that these guys are very competitive in the market. So therefore this is a feature that's worth having. How would you overcome that? There's also, this is the used watch market. So there's also a quality element. Like you might be able to get it for $10,000 cheaper on some other site, but ours is flawless or at a perfect condition. Like how would you even incorporate those kind of things to sort of skepticism? Which is immediately what I sort of go to.
- [21:30]
B
Yeah, yeah, so you bring up a really good point. And like the brand ownership basically stated that like they didn't ever want to compete on people price per se. They want to be known as like the cheapest option. They want to be known as we curate and only list items that we've bought that are now reselling that are of the highest quality. So the hypothesis is if we can get the price barrier out of the way and we can show people, look, Here are the 10 listings are going to find worldwide. Here's the lowest, here's the highest. And oh, by the way, here's where they're being listed. Like here it is, you don't have to go do your research for you, we've done it for you. The idea is that not only would they avoid leaving the site to do that, but they would then move to the next stage of the consideration factor, which is this beautiful section the site has as to exactly what you just mentioned. In other words, okay, they're going to focus more on the quality because we removed the whole pricing objection and they now intellectually know that. Right. Because if they're sitting there reading all this jargon about like we only curate the best, it's of the highest quality. A lot of that is like, yeah, yeah, yeah, I got it. But like how much, like where is this? Are you way over everybody else? Because everyone wants to pay for high quality but not like way over market price. Right. So that's the logic behind it. We're still running this experiment right now. So far it's looking really good. But like exactly to your point, that's like the second order effect of the feature.
- [22:42]
A
Yeah. So you're qualifying, you're just knocking those people out and you're trying to overcome some of those objections before they sort of get to phase two, which ultimately is sort of more of a quality question. It's a, it's a, you know, it's a. Yeah, it's a quality play as opposed to just a pricing play.
- [22:58]
B
Exactly right, exactly right. And with the assumption that if we can address all the, the pricing bits up front that they will then shift into that quality, which seems to be a secondary consideration. And I mean secondary like less important. It's more of just an order of operation. Right. It's like price first, you know, quality, secondary.
- [23:13]
A
Yeah. You know, knowing the tools right now, especially with the mcps that are out there, like I. This actually, even though it's so cool, I mean and can be like, this could be like a mind blowing sort of lift that you make here. Probably like you said, it wasn't as hard as you would think because the
- [23:30]
B
technology exists now, can dip it every two hours. Yeah, I mean super, super specific.
- [23:37]
A
So.
- [23:37]
B
Yeah, exactly.
- [23:39]
A
Very cool. I mean that's, that's like the future of after the click right there.
- [23:43]
B
Yeah, totally, totally. And like I'm talking about with AI, there's a bunch of fundamental stuff too. Like we're working with a lot of clients in the respect of like GEO right now is, is. Is the talk of the town. Right. Generative engine, which basically like LLM optimization, aio, whatever, you know, initialism or accurate aogeo. Right, exactly. But it's. Well that's, you know, for those that maybe that's. That's foreign to is is basically how do you make your site optimized to be discoverable. Right. So to speak, by LLMs in the sense that it might use you as a reference or as a source. So if you go talk to Claude or chat GPT or what have you as to the best X or Y or something you're looking for, you want your brand to be in the consideration. So hopefully you might get a tip of the tip of the cap, you might get a direct link to your site, you know, whatever. So that also spills just like all things into the world of CRO, which is, is your page optimized not just from like an SEO consideration, but from like a CRO for LLMs. And like here's a cool thing that we found is one of the classic tenets of web behavioral psychology is, is a aversion to scroll down. And what I mean by that is if you look at any website, any website out there and you take a cohort of new users, one of the first trends you're going to see is their unwillingness to scroll down on the landing. Doesn't matter if it's a homepage landing, page collection, page product, page service, it doesn't matter what it is. So in other words, the bottom threshold, the footer area, etc. Right. That whole area generally gets like very minimal human visibility. So what we've done is on a lot of sites we've actually built, I'll admit, not very attractive, but very LLM friendly. Tables, us versus that. Comparison charts, data charts, things that. Things like LLM Gobs loves that. Right. Whereas a human might look at it, be like, this is overwhelming. You're throwing too much at me. Not for LLM. So we kind of conveniently and strategically put it where we know humans aren't really scrolling anyways, but LLMs will still pick it up. And so that way we can kind of have our cake and eat it too. Where, if you will, the top 50%, so to speak, of the page is more human optimized. And like the bot and I'm generalizing, But the bottom 50% of the page is more LLM optimized. Right. So as an example, we spill heavily over now into that world as well. And again on that note, like, you know, tables, for instance, are critical. And it's funny because like this is so advanced. A lot of this stuff is super advanced, but it's also so rudimentary and foundational, which is like, yeah, the LLM just wants to know, who are you? Do you have any credibility behind you? Who do you sell to? And what are the value propositions? Which is exactly how the human should speak too. We see all these sites of all this fluff and like hyperbole and it's just, just tell us who you are. How is it going to help the user? What makes you unique? Like clear value propositions, clear data points. Right. Front and center, you know, and a lot of sites have this fluff you got to like dig through to get to like, what is it that they're actually offering? Jargon. Yeah, you know, we're the industry leaders to optimize your outcomes and it's platitudes. Yeah, yeah.
- [26:41]
A
Thing to make.
- [26:42]
B
Yeah, yeah, exactly.
- [26:43]
A
So, yeah, pretty cool. I mean, above the fold is still like, it's the oldest expression, but still like that's. I mean, any heat map that you look at, which I know you guys use heat maps on this, it's like it just turns cold like after, like. Well, now it's thumb scroll, it goes cold really, really fast. You gotta captivate everything. That, that is sort of a true tenet of everything. After the click still, it's like anything that's above the fold, you've gotta grab their attention in some way, shape or form. And I know that's something that you all obviously look at totally, but that's analytical and a bit of design too,
- [27:21]
B
to a certain degree. Yeah. And it mirrors a human sales. So this is always a fun exercise if you can do it. Is you. If, if you ever try to sell something, watch what your first verse, second sentence is. It's like the proverbial above verse. Second, if you don't nail the first sentence of what you're doing and you say something that's confusing, just watch the person you're talking to. They're like, oh, okay. And their eyes kind of gloss over and the chances of you winning them back to paying attention on the second sentence is going to be dramatically lower. So it's like it almost mirrors like
- [27:48]
A
the lines up, right?
- [27:50]
B
Yeah, exactly. Well, like your best pitch has to be right out of the gates to get their attention where they're like, huh? And then you've got their attention. Same exact thing on a website. Same exact concept. But again, back to the whole AI conversation. Not with LLMs. LLMs in theory can pick up all that ugly table data that might live at the bottom of the page, which, you know, again, humans aren't going to necessarily really see anyways.
- [28:12]
A
Yeah, yeah, tremendous. So give me one more example. I know there's an example on the E commerce space of utilizing AI in order to determine which products to potentially sell at which any given point in time.
- [28:28]
B
Yeah, of course. So this one was born out of some of our clients that have really, really heavy SKU counts. We have a, we have, we actually have one client that believe it or not in their ERP system has over 2 million unique SKU number. Yeah, I know it's in the automotive space, but the point being is some of our clients that have heavy, heavy SKU counts, one of the challenge, or let me put it this way, one of the longtime wins that we often find is products that are priced a certain way and have varying degrees of sales velocity. Some sell really quickly, some sell lower, and oftentimes we just fall into the trap to be like, oh well, product, you know,327 doesn't sell as well because it's just not as in demand or it's just not as attractive a product, you know, Whereas product number 13 sells really well because it's just a high demand product. And whereas that is generally true, a lot of times people may mistake product price market fit, which is like that product actually might be more in demand than you think. You just didn't price it right. So features that we have now is a internal basically forecasting tool. It's like a sales velocity measurement tool. It's all driven by AI, which basically is real time measuring the threshold of sales across a huge amount of SKUs. And it has a threshold that's predictive to be like we should be selling as an example, 100 units per month of this product. Let's assume we're tracking towards 40%. Right. Only 40 units per month are selling. What the AI does now is it automatically can inject its own experiment to price experiment using discounts. So without us even having to pick us pick it up, it allows basically price experimentation to go in and inject like a 15% off to see if that will increase the price point and vice versa. If certain products are selling, dare I say, like so fast that it's like, are we charging enough for this? It can actually do the same thing. So it has this permission as an agent to create price elasticity experiments. And what does that do? It basically speeds up the whole process. You don't need this feature if you got like four SKUs, I'm just telling you right now. But if you are a company that's got dozens and dozens or hundreds or thousands of SKUs, this is a really interesting model because usually we focus on like the top 10% or 5% of SKUs. There's a huge amount of optimization that lives in that iceberg below the water, so to speak. Right. Of products that maybe just aren't price product optimized. And so there's another feature that we're testing out right now. It's in like a beta right now with a couple clients to watch how we can increase the sales velocity on the holistic amount of products, not just like the top performers. So pretty, pretty interesting.
- [30:57]
A
That's so cool. I mean this is for thousands of SKUs, obviously, but in the test that you now, are you running it on a subset of individual SKUs, or is it throughout the entire catalog, which would obviously be a major undertaking, but I assume it's like in a certain subset, you know, top 25% thereabouts, like maybe bestsellers, that kind of thing. Like how would you go about doing the methodology on that?
- [31:24]
B
Yeah, we have one client that's on the entirety. Just because they know they have, I don't like 100 SKUs. Or something like that. We have others where it is on a subset of like a specific part of the site in terms of like a niche offering. It just kind of depends. But again, it's all through the inspiration that if we can measure real time demand and measure that up against predictive demand and determine how low at or above that threshold, we are allow AI to, to autonomously inject itself to adjust and can adjust quite literally the price via injecting a coupon code, via applying an automatic discount as an example. So it's not like you have to change the price point itself, you can just inject coupon codes. But again that kind of plays into the narrative that maybe the pricing was a little too high or a little too low as an example. So. Right, that's, that's the genesis of it. But you can kind of use your imagination as to where this could go. So like in AI, it still took the human to ideate on this concept and to pick up the data and to come up with it and to engineer it, but now the AI can execute on that at a speed no human could. So that's, that's kind of, you know, where, where the human and the AI intersect.
- [32:28]
A
That is amazing. It's like a gentic pricing adjustment. Like it's not even discount like, like you said. I mean if you got something that's selling like hotcakes and you're at risk of maybe running out of inventory and hey, let's check whether or not we can increase our aov. It can increase it the price on those items just as easily as lower the price of the ones maybe aren't moving the fast that you have more inventory of.
- [32:51]
B
Yep, exactly right. Exactly right.
- [32:54]
A
That is insane. Well, like I said, we're gonna have to have you come back once all this stuff is live and we can show it, divulge it. This is the future of, of conversion rate optimization, the future of after the click. Like I said, Ned, amazing to have you on this three part series here. Obviously with the new format here in perpetual traffic, you know, having tier 11 folks coming on, on, on a regular basis and sharing their expertise is something that we're going to be doing and would love to have you back on. Especially this AI stuff, which is just absolutely insane. In the meantime, if you do want our expertise. Where do you get that?
- [33:36]
B
Yep, exactly. So tier 11 CRO. I said that right, for CRO. Yes. Y. In which case, yeah, we're happy to take a look, look under the hood, do an audit. Exactly. I'll. I'LL chat with you directly. I, I usually lead all those. So, yeah, again, we, we'd love to take a look at the brand and give you some, some pointers and, you know, at minimum you get a real great free value add, you know, and potentially we could find a way to work together. So.
- [33:57]
A
Yeah, absolutely. And this isn't something that like, you're, you're involved. I mean, you're technically doing like CRO and tier 11 as our client. Like, you're involved. It's. Yeah, this is not something that you outsource, even though, like, the team that you have is absolutely amazing and a huge amount of value in that CRO audit. So definitely check that out over at 211.com forward/CRO. So, Ned McPherson. So on for episode number three. We'll have you back on in a few months or so. But yeah, it's been a pleasure.
- [34:29]
B
Awesome, brother. Thanks so much. Appreciate it.
- [34:30]
A
You bet. All right, well, we're going to leave all the links, everything that we mentioned here on today's show over on the show notes over@perpetualtraffic.com of course, you can check us out on the video version of this@perpetualtraffic.com YouTube and of course, if you'd like a free CRO or after the click audit, check that out at tour11.com forward/CRO. So until next show, everyone. See ya.
- [34:58]
B
You've been listening to Perpetual Traffic.