How is the Ecommerce Industry Actually Using AI?

[00:00:00] Luke Austin: the funny thing is like, I dunno, six months ago, maybe even, everyone was like, AI is coming for production, right? Like creative production and like the and what we've, our experience so far has not not been the case. Like, icon me launched a service offering right on top of their creative production engine.

And so outside of creative production, you can go sort of down the line in terms of, in terms of me media production, like there's. B, creating ads and building out campaigns on meta, building out campaigns on Google somehow, like isn't actually being the thing being affected. It's more like I can be my own lawyer now or whatever the thing, like high level high level, high education you know, sort of skillset that was necessary prior.

So it's coming for the strategy more than the production for sure, in terms of the efficiencies that, that we're seeing and that we're able to. That we're able to engage in. 

This episode of the DTC Hotline is brought to you by charge flow chargebacks cost merchants billions every year in lost product, lost revenue, and lost time. Charge flow automates the entire process, end to end, instead of wasting hours, collecting screenshots, and filling out bank portals, charge flow uses AI to fight disputes for you and wins up to four x more.

Than manual efforts. Join over 15,000 merchants protecting their revenue with charge flow. Right now. You can try it free for 60 days. Check out the link in the show notes to start saving money, time, and sanity today.

[00:01:27] Richard Gaffin: Welcome folks to the DTC Hotline. This is your direct connection to Hot Takes, cold Truths, nothing but real e-comm advice from some of the best in the biz. I'm your host Richard, the Professor Gaffin. Call or text 866-DTC-2263 to get your burning e-comm questions answered. That's 866-382-2263. Leave a voicemail, shoot us a text, ask whatever's on your mind. our

[00:01:52] Luke Austin: our operators.

[00:01:52] Richard Gaffin: by, and

[00:01:54] Luke Austin: And speaking of which.

[00:01:54] Richard Gaffin: introduce them to you right now. Coming in live now from CTC studios in Costa Mesa, California. We got Mr. Luke the weatherman. Austin Luke, how are you doing, man?

[00:02:03] Luke Austin: I'm doing great. We've got some good hot questions today that are gonna spark many, many a good discussion. So we're looking forward to it. As always,

[00:02:13] Richard Gaffin: excited for your piping hot takes. All right. And then of course, coming from his beautiful home in Newport Beach, California, we have Mr. Tony Chop the chopper coming in live. Tony, what's going on man?

[00:02:23] Luke Austin: Flying high, Richard.

[00:02:24] Tony Chopp: To.

[00:02:24] Luke Austin: to elevate.

[00:02:25] Richard Gaffin: Excellent.

[00:02:27] Luke Austin: Alright, cool.

So

[00:02:27] Richard Gaffin: basically like

[00:02:28] Luke Austin: Lin,

[00:02:28] Richard Gaffin: just a couple words about the format here.

We get questions from listeners. We get questions from our members in admission, which is our subscription training platform, and then we let Tony and Luke go nuts on 'em. So

[00:02:39] Luke Austin: so without further do 

[00:02:41] Richard Gaffin: right into our first voicemail here.

Hi DTC Hotline. This is Katelyn. So I've been a part of a handful of conversations about price changes ahead of or in response to the tariffs. So I'm wondering how are you seeing this play out just in the wider e-commerce landscape or you know, from the business perspective as well as consumer sentiment, consumer sentiment, and overall consumer purchasing power?

Thanks.

[00:03:08] Luke Austin: It is a great question. So, Taylor did a poll five days ago. A little under 300 people voted on X. Have you raised your prices due to tariffs? 49.1% said yes. 50.9% said no. So, half of all brands are. Raising prices to some extent due to due to tariff. And that, that pa, that additional cost is gonna be passed down to the consumer ultimately in terms of the price changes and the impact there is not a consensus in terms of there, there's not a strict sort of correlation relationship in terms of price increase, X percentage or X amount is going to lead to.

This specific impact in terms of your conversion rate outcome, and we can just sort of model it out and let's find the optimal point on the graph because similar to spin a MER models, the degradation curve or the impact curve of price increase and what your consumer, what your specific consumer base is going to respond, how they're gonna respond to is.

To a level of price increase is going to be specific to your brand. You have a specific degradation curve for your spin or model. You have a specific curve as it relates to price sensitivity for, for your consumer. And the short answer is until you do an thoughtful. Price testing against what that curve looks like and what the impact is.

It's there's no for, there's no formula or or, or guidelines around what that should, should look like for you. So in, in the context of price changes it's gonna be necessary for many brands getting passed on an additional cost really directly to them. That's going to have to show up somewhere either in the form of more efficient acquisition cost and or in the form of price increase and likely some combination of the two.

And so. The long and short of it is you have a specific de degradation curve for your brand related to acquisition efficiency. The spend a ER model, you have a specific curve related to price sensitivity and how those changes are gonna impact your consumer. And that now you have another dimension added onto your spin, a ER bot model, basically, in terms of finding what that optimal point is and doing thoughtful testing like we do for media channels like we do with incrementality.

Testing, thoughtful testing on price changes and price impact to find what the. Optimal point is based on the needs for margin creation for your business.

[00:05:38] Tony Chopp: Yeah. Yeah. It's one of those things that's like always true, right? Like prices are always increasing. I think a couple things really stand out about this. We, we did some analysis of our portfolio back

[00:05:50] Luke Austin: like 24 hours a day.

[00:05:52] Tony Chopp: Their supply chains when, when the tariff stuff was originally popping up, I was actually a bit surprised by how diverse it was. I, I had expected there to be a really heavy concentration and. In some of the countries that had really high tariffs and, and I was, pleasantly surprised at the, the robustness and diversification of the supply chain of, of a lot of the clients that we work with. And, and I think this is, I don't wanna make this a non-answer, but I think it's really important. For any operator to think about the diversification of their supply chain and, and, and concentration of that in any one place. As, as a risk, there's, there's proportional

[00:06:30] Luke Austin: risk,

[00:06:31] Tony Chopp: associated with the concentration. Just,

[00:06:33] Luke Austin: just like there's proportional risk

[00:06:34] Tony Chopp: If you have a really high concentration of your media investment or your growth engine all into one channel, it can work, but it

[00:06:43] Luke Austin: it represents.

[00:06:45] Tony Chopp: So I think

[00:06:46] Luke Austin: I think that's one of the lessons

[00:06:47] Tony Chopp: to

[00:06:47] Luke Austin: to.

[00:06:48] Tony Chopp: from this.

[00:06:49] Richard Gaffin: sec. I'm getting an echo from what happened with you, Luke. Hold on. Okay. Sorry. Go ahead, Tony.

[00:07:00] Tony Chopp: One of the lessons to be learned from, from this year is keep thinking about how to diversify your supply chain for, for the products that you sell. Which is generally good practice, is good practice in general, just to be considering the variable costs that go in into your cost of delivery. So, so that's point number one. Point number two, I, I think a lot about when you have the conversation around when price, I think it's really important. To not get overly fixated on price and to make sure that you're really paying attention to the question of value. Because tho those two things are not always the same. I read something that Nick, Nick Shackleford posted on LinkedIn recently about his company Breeze and he, he was talking about something about the, forget the term he used, but I want to say like the. The negative space economy or something, something to this effect. And, and effectively, the what, what he was, what he was saying is there's there's all of these various products that are creating value by removing ingredients. So for example, caffeine free beverages as a category are perceived as higher value and thereby higher price. Now that's a double win because. Number one, you have a higher value. So you can charge a higher, you can charge a higher price. But number two, the point that Nick was making is it actually reduces the cost of the product as well by removing ingredients. So I think that's, that's an interesting way to explore the value proposition is, or let's say in a different way, if you're only thinking about price. It's a very dangerous, risky race to the bottom, and you need to make sure that you're thinking about considering your value proposition so that you can support inevitable pathway of rising costs of everything, always all the time. We're just in a little bit of an acute moment here, but it's not, it's not really that much different than the normal trajectory of things.

[00:09:02] Richard Gaffin: So the idea then, just to kind of summarize that last piece, is because there's an in inevitability to prices going up, there has to be a, sort of a, what's the word I'm looking for? Coinciding increase in the value of what you're selling or the way that you're sort of. At least messaging what you're selling so that it kind of holds up to the amount that that price is gonna have to increase.

Right. So if somebody's paying, I dunno, 10, 20, 30 bucks more for whatever, they have to have some understanding that that's a great value and that it's well worth the price going up. That type of thing.

[00:09:35] Tony Chopp: Well, yeah, this is, this has always been true. You know how much you can sell something for

[00:09:40] Richard Gaffin: As much as people will pay for it.

[00:09:41] Tony Chopp: as much as somebody's willing to pay for it. Exactly. that's all about that. That's all about value. And you know, Luke and I have these conversations like almost continuously around, the pricing and the, the concepts on how we deliver CTC services to our clients.

So we're, dealing with our rising costs that happen all the time. It's all about value.

[00:10:03] Richard Gaffin: yeah. Totally. Talk. Let's dive in a little bit. Like what does that conversation between the two of you look like? You don't necessarily have to go into specifics about our product, but like what? How are you? Like where are you looking to increase value? Like how are you going about that?

[00:10:21] Luke Austin: So I would say there's a, in this conversation, all those things being true that Tony talked through will be all. Also all understand is, where we wanna live is at the intersection. We, we want, we want the service that can deliver all the things at the same time. The best product in the and and service in the market at the best price with the best, the best folks being able to deliver, deliver that, that thing, right?

So like once you start operating in one dimension, separate from the other, Hey, we have strong, we have team members with more experience. And we have really cool tools, but we're also the most expensive. Or. We're pretty fairly priced and have some of the best people with best experience and our tools are, are, are less than helpful.

Once you start doing that, then you really have to index in like couple those dimensions to make up for not being able to have a value prop that exists in, in whatever that third, second, or third dimension is. So. The conversations for us orient around how can, how can, how can we deliver all of those, the, those three things at once?

How, how can we just be the, the, the clear choice if you're evaluating against the, the, those three criteria. And what I would say is over the course of the recent months, we've been more and more, interested in helping deliver our service at the best price possible in, in the market. And finding ways to make that to make that possible for our customers through developing an incrementality and measurement offering that we're including as part of our core service when customers work, work with us in our, in our core capacity.

And, and we can kind of go down the line from there, but primarily thinking first about how do we, how do we operate as being the best, the best value against those things in the space. And then, and then work up, work up from there. Because value, the value equation is cost against the return, right?

So the denominator, we wanna, we want to keep that as low as possible always, and find ways to improve that because it's gonna be helpful in anyone assessing that, that value creation. And then what I'd say is on the, on the, on the top side of that formula, in terms of the net new revenue or margin creation, is we are, we are focused on our system operating against forecasting and executing against contribution margin targets on a daily basis. And we are as we are interested in being as closely tied to the financial outcome of the brands that we work with, so that we can clearly show if the financial impact was there or not in a way that we believe is unique unique in the market.

Now, what I'd say to that point is. We, in this exercise, we've also seen ways in which we we're aware of ways in which our system has cracks in it related to the integrity of the data and that aligning with the core the finance team on the customer side and has actually all being able to agree upon contribution margin as the definition and what we're tracking against.

And also cracks that relates to the the, the forecasted outcome versus. Goal or the desired outcome for the brand and bridging those things. So, there, there are, there are gaps that we're looking to address to make the value side of the equation as impactful as it can be, while also thinking about while it's thinking about the cost, but short answer is go going after all the categories as aggressively as we can.

Because once you start losing in one of those categories, then you really have to overcompensate elsewhere to make up the gap. 

If you're still fighting chargebacks manually, you're doing it wrong with charge flow. Merchants automate the entire dispute process from Shopify to stripe. Charge flow plugs in, collects evidence, builds the case and submits it automatically. No more late nights at bank portals, no more disputes slipping through.

Just clean automation and real results with merchants recovering four x more revenue compared to manual handling. See for yourself. Try charge flow free for 60 days at the link in the show notes.

[00:14:05] Richard Gaffin: All right. Let's let's jump into our next one here. 

Team, appreciate the tips so far on DTC Hotline, hoping you're able to help me out, um, with a question around creative performance. Um, looking just to see if you had any tips around measurement. Um, on a more granular level, like how much time are you spending with your, with the clients you work with, looking at things like click through rate or total watch time.

Also, are there any like sneaky metrics that you've slipped into your custom columns that have been really helpful, um, when you're looking, um, a little bit closer at creative performance? Appreciate the help. Thanks.

[00:14:36] Tony Chopp: it's

[00:14:37] Luke Austin: It's sneaky because you can't actually get,

[00:14:39] Tony Chopp: it, but the the creative fatigue measure. Trick is, is I think really, really interesting.

[00:14:44] Richard Gaffin: Mm-hmm.

[00:14:45] Tony Chopp: I, I'm

[00:14:45] Luke Austin: not.

[00:14:46] Tony Chopp: sure why meta hasn't surfaced it more broadly. But we've been able to access it through our, through our various account reps and various accounts. That's like the first thing that jumps out to me is, is the, the sort of measurable, measurable relationship that meta has to how, how tired an ad is.

[00:15:06] Richard Gaffin: Hmm.

[00:15:07] Tony Chopp: And I think it, it probably represents the, probably the biggest area of impact. 

[00:15:12] Richard Gaffin: Yeah.

[00:15:14] Tony Chopp: you know, outside of that, like it's, I think it's really easy to get, to get into this idea of like any one metric being a proxy for like an out, an outcome. And I think it can be. Kind of sketchy. so I'm a little bit cautious to say like, oh, C watch CTRs, or cvs, or CPMs or, or watch time, or really, really any of those things. Because in, in the various analyses that we've done, we see very, very little to no correlation between any individual metric, C-T-R-C-V-R, watch time with, with the actual output. I think the, the biggest, the biggest thing is whether or not, meta is determining a particular creative unit as fatigued or not.

[00:16:06] Richard Gaffin: Interesting. Luke.

[00:16:08] Luke Austin: Yeah. In addition, the, the thing I'm most interested in is where, where is the spend volume getting aggregated against in, in the account? And the principle meaning that the. Majority, 80% of the spend and the volume being driven by the account is gonna come from 20% of the, of the, of the creatives. It's actually, it's, it's typically a lot more exaggerated than that.

Even where it's like out of the 300 ads you have running in your account, five of them are contributing 80% of the spend. And so, finding a winning ad is. Is really hard because there's no, there's no formula to it. When you go into account and you see these are the five hours that contribute 80% of the spend, like, the simplest way to put it's, I'm gonna go and try to find ad number six and ad number seven. How, how am I going to do that? Well, first I'm gonna make 50 to a hundred iterations of all five of those ads that are in there. Different color backgrounds and different, and it's a guy batting underwater, and I'm gonna make a batting in a sort of, dark, dark, deep blue ocean water versus like.

Coastal. I'm in Cancun Water and like just all the variations like sharks swimming by him while he's batting underwater versus there's a man array there, right? Like just a hundred iterations of the top five ads in the account. And then I'm also going to make a hundred ads that are just completely net new based on things that my, are based on my customer review analysis based on.

Competitors in this space and getting inspired by them, but just net new things. And once I do that and create a hundred, a hundred iterations of my top five ads, and I create 500 new ads that are just completely new concepts based on those things, I may have a chance to create a six or seven, maybe even eight.

Probably not add 20. Right? Like, but that's what it's, that's what it's gonna take. So the aggregation of where the spend volume is going, is, is the strongest indication of what's actually gonna deliver meaningful scale for, for the brand in the long term? And it should, should we iterate? Should we create net new concepts?

What's the difference if I change the background from blue to red? Is that a net new concept? Is it iteration like. It doesn't, it doesn't matter. Like create a hundred iterations and create a hundred new concepts. Like you're, you're trying to find ad six or seven, that's going to be the thing that for the next two years accumulates millions of dollars of spend for your, for your ad account.

Like that is, that's the task that we've seen as necessary to be able to actually meaningfully impact the ad account performance based on quote unquote creative insights.

[00:18:38] Richard Gaffin: Yeah. So I mean like the long story short then there, there's no particular like leading metric to indicate creative performance other than did it. Perform at the end of the day, right? Like is it driving conversions? Is it is spend being aggregated towards it? But

[00:18:50] Luke Austin: But I mean,

[00:18:51] Richard Gaffin: know

[00:18:51] Luke Austin: I know we used to put lot of stock,

[00:18:52] Richard Gaffin: of like whatever click through rate, watch time being indicators that maybe there's some magic under the surface that isn't quite working. But

[00:19:00] Luke Austin: really what it sounds like,

[00:19:01] Richard Gaffin: just like elements that

[00:19:03] Luke Austin: that make it,

[00:19:04] Richard Gaffin: are sort of

[00:19:06] Luke Austin: but who knows?

[00:19:06] Richard Gaffin: are really like, like you were saying, there's no formula for it, so you just have to be taking shots as many times as you possibly can.

[00:19:12] Luke Austin: Yeah, I think, I think the metrics are, can be helpful for troubleshooting when there's meaningful, meaningful drop offs in per in performance or, or what the business is seeing. The, the metrics. In our experience are not as helpful in determining what the next best thing is that you are going to be able to unlock for the brand and generate in that new idea.

So said another way you look at the historical ads, you see the click-through rate on this ad is just as good as all the other click-through rates in your account, but the conversion rate is super low. Why is the conversion rate super low? And then you go through. Oh, it's because the shoe that was shown in that ad is sold out of half the sizes, whereas all the other ads that have the click-through rate aren't, okay.

This is, this is helpful. Like we've trouble, we've been able to get down to like a performance related issue based on that. But the click-through rate being lower on average, like what? Why, why is it?

[00:20:04] Richard Gaffin: Mm-hmm.

[00:20:04] Luke Austin: I think we, we all have, we all have ideas. Let's create, let's create ads based on all of those ideas plus, plus more.

Because in creating those new ads, it is just gonna create more, more questions and what the, what the combinations are of the creative elements that lead to the outcome. Create, create a high volume of creative output is gonna be the thing that's gonna lock, unlock more of what that looks like.

[00:20:27] Richard Gaffin: Yeah. I, I wanted to double back to what you were saying, Tony, around the the fatigue metric. So

[00:20:32] Luke Austin: Obviously

[00:20:33] Richard Gaffin: something that people have access to widely, but can you give us a little more insight into like, what, what

[00:20:38] Luke Austin: what does that,

[00:20:39] Richard Gaffin: look like and what, I know meta is a, black box most of the time, but what insight, if any, you have into what the metric itself is indicating.

[00:20:48] Tony Chopp: Yeah,

[00:20:49] Luke Austin: so

[00:20:50] Tony Chopp: I think I, I was thinking

[00:20:51] Luke Austin: thinking about this and talking

[00:20:53] Tony Chopp: the creative fatigue metric gives

[00:20:55] Luke Austin: gives an

[00:20:55] Tony Chopp: ratio measure

[00:20:56] Luke Austin: on how.

[00:20:58] Tony Chopp: In

[00:20:59] Luke Austin: how

[00:21:00] Tony Chopp: meta is

[00:21:01] Luke Austin: is determining the ad is

[00:21:02] Tony Chopp: fatigued. It definitely has

[00:21:04] Luke Austin: has relationships.

[00:21:05] Tony Chopp: They,

[00:21:05] Luke Austin: They'll

[00:21:05] Tony Chopp: won't tell

[00:21:06] Luke Austin: tell us exactly,

[00:21:07] Tony Chopp: what goes into it, but.

[00:21:09] Luke Austin: but

[00:21:10] Tony Chopp: I

[00:21:10] Luke Austin: I gotta

[00:21:11] Tony Chopp: about

[00:21:11] Luke Austin: about it

[00:21:12] Tony Chopp: was

[00:21:12] Luke Austin: talking, it's like

[00:21:13] Tony Chopp: all it

[00:21:14] Luke Austin: it tells you to do

[00:21:15] Tony Chopp: is

[00:21:16] Luke Austin: is go make more ads.

[00:21:17] Tony Chopp: It

[00:21:17] Luke Austin: It might tell you a little bit sooner

[00:21:19] Tony Chopp: Then 

[00:21:20] Luke Austin: an extra losing.

[00:21:21] Tony Chopp: the performance from a historically, a historically

[00:21:24] Luke Austin: High,

[00:21:25] Tony Chopp: high

[00:21:25] Luke Austin: high spending ad.

[00:21:26] Tony Chopp: But the, but the road that

[00:21:28] Luke Austin: That it.

[00:21:28] Tony Chopp: to is no different than, than what Luke was describing. It's worth looking at.

And I think the, the, this is the number one thing that I think is interesting about looking at it across accounts. It gives you an average time to fatigue based on the whole portfolio of ads, which we've seen some pretty wide variances in anything from a week or so to several weeks. And I think maybe there's, maybe there's something to be gathered from that as like, why do some accounts have seem to have more, a longer runway with the creative they put in the account as opposed to a shorter runway.

[00:22:03] Richard Gaffin: Let's let's jump to another one. This is a, a text message question that we got, which is, Hey guys, how are you recommending clients respond? To the new Texas SMS law, should we just exclude Texas from our SMSs or is there a better way to approach.

[00:22:25] Luke Austin: So there's an opportunity cost in any one of these decisions. And and the trade off of the risk versus the reward is, is something that each brand. Is going to have to figure out for themselves in terms of, in terms of what that looks like. So this is, this advice and this perspective is not a specific recommendation or advice to be taken in any way, but a framework for.

Thinking about it, which is it's a risk reward decision. It's a, it's non-binary. There is, there is risk reward on either side. It needs to be, and it needs to be weighed. For,

[00:23:01] Tony Chopp: Yeah.

[00:23:01] Luke Austin: a good number of brands. There's likely a small amount of revenue. Meaningful revenue volume that you're driving from SMS in Texas and the work that's going to go into uncovering what the risk of continuing to send in Texas is what the what the changes to your program need to be to be able to accomplish that.

The additional cost of each SMS send to that market may indicate to you the volume's low enough that like. Maybe you just stop. Maybe it's like there's, it might not be worth the resource and the energy and the potential risk or trade off based on the volume you actually get there. And we don't do that.

And we focus on other states and find ways to increase overall contribution through some net new flows across email and SMS or make up the ground and paid media or whatever, whatever it might be. For the brands, there might be more meaningful volume there, where it's definitely something to get after.

In which case there's also a question about are like, what is the, what is the risk for yourself as a brand continuing to engage there against. When there might be enforcement against that law, and if you're the person that will actually get raised in some of those issues or, or not, like again, there's a risk and reward on either, on either side of those things and, and something that needs to be taken into consideration.

But it's before. Setting out on, okay, before accepting, okay, this is something we need to do and we need to invest the resource and the time and really change this, and it's gonna be a big overhaul project. Think the first step for anyone is do the risk reward. Understand how much volume and actual marginal dollars you're creating from that specific subset of that channel and that.

Geolocation and understand if it's going to be, if it's something that's worth just not engaging in and redirecting the focus and attention elsewhere, do the risk reward trade off on the front end to, to help to decide what course of action you take before you just start engaging in and accepting, oh, well this must be the path because this is what everyone is saying, or this is what I've heard that everyone's doing, or whatever it might be.

Do do the math for your brand specifically on the front end. And then you'll have to make a decision based on the opportunity cost between. Between the two decisions.

[00:25:09] Richard Gaffin: Tony.

[00:25:11] Tony Chopp: Yeah, I, I, I read a little bit about it. I, I don't, I definitely don't fully and deeply understand it, but I, I think what a

[00:25:18] Luke Austin: Part of it's

[00:25:19] Tony Chopp: is requiring businesses to register with their secretary of State so that the business is actually real. So as, as somebody

[00:25:27] Luke Austin: somebody who gets like.

[00:25:28] Tony Chopp: 10 million spam calls a day, like honestly, I'm here for it.

Like, so

[00:25:34] Luke Austin: I

[00:25:34] Tony Chopp: the,

[00:25:34] Luke Austin: the policy is actually not

[00:25:36] Tony Chopp: as, it, it potentially as punitive. I think it's more a matter of

[00:25:40] Luke Austin: like establishing,

[00:25:41] Tony Chopp: a legitimate business and, probably some governmental fees to deal with. And, and I, I

[00:25:47] Luke Austin: I think there's also a.

[00:25:48] Tony Chopp: out on it that it doesn't apply to messaging existing customers, so anybody that you actually have a customer relationship with. so yeah, I don't know. I, I don't feel, I, again, I know I'm not. I'm not super email SMS expert guy, but I think the generally speaking,

[00:26:07] Luke Austin: speaking

[00:26:07] Tony Chopp: I'm

[00:26:07] Luke Austin: I'm happy to receive messages.

[00:26:09] Tony Chopp: companies that I do business with and very unhappy to receive unsolicited texts or phone

[00:26:14] Luke Austin: Call

[00:26:15] Tony Chopp: from people that I don't so hot

[00:26:17] Luke Austin: Take. I'm here for it.

[00:26:18] Tony Chopp: Sorry,

[00:26:19] Luke Austin: Jessica.

[00:26:20] Richard Gaffin: There you go. Love it. All right,

[00:26:22] Luke Austin: Alright, cool. Alright, so let's jump into

[00:26:24] Richard Gaffin: One final voicemail here. This is, I

[00:26:26] Luke Austin: this.

[00:26:28] Richard Gaffin: going to give you the most o opportunity for a hot take here, but we'll see. So I'm gonna play it off right now. 

Hey guys, love the hotline. Um, from what you're seeing across the best e-commerce brands, how is AI actually showing up in their workflows now? Um, is it only the research or creative strategy process, or are you seeing it anywhere else? That's interesting. And also, what's your hottest take on where this is all headed for e-commerce businesses in general?

Thanks.

[00:26:54] Luke Austin: Yeah, so the, this is the funny, the funny thing is like, I dunno, six months ago, maybe even, everyone was like, AI is coming for production, right? Like creative production and like the and what we've, our experience so far has not not been the case. Like, icon me launched a service offering right on top of their creative production engine.

And so outside of creative production, you can go sort of down the line in terms of, in terms of me media production, like there's. B, creating ads and building out campaigns on meta, building out campaigns on Google somehow, like isn't actually being the thing being affected. It's more like I can be my own lawyer now or whatever the thing, like high level high level, high education you know, sort of skillset that was necessary prior.

So it's coming for the strategy more than the production for sure, in terms of the efficiencies that, that we're seeing and that we're able to. That we're able to engage in. Now I think that's that's gonna continue to change. Like this space changes really quickly. Google's nano, nano banana that that was launched recently is doing some wild stuff as it relates to creative production and, and what's able to do so, like there's gonna be, continue to be new tools.

Where, where we see it coming up is more creating efficiencies in in sort of high level thought work and strategy. And supercharging those sorts of, those sorts of sorts of human beings where the like production layer is, is not being affected in the same rate at this point. And I, I think it's worth noting that.

Meta. So like Meta has simplified their ad buying platform, right? Like simplified into ASCs and things like that. So that's like maybe one way they've consolidated the like media buying labor intensive workflow. But outside of that, like you still have campaigns like Meta's Portal isn't like an AI agent that you drop and you're like, build me a campaign.

And it does it

[00:28:50] Richard Gaffin: Mm-hmm.

[00:28:51] Luke Austin: the AI is like in. The targeting and the backend and the algorithm, right? Like the strategy of where the media gets deployed, not we're gonna deploy AI against the labor intensive sort of building out the campaign that which they very much could have done and created a building portal that's like an AI agent and you're just like, Hey, I want you to build this, and you do it.

And they build your campaign and ad set and your ads and they tweak all the settings and it's like your AI media buyer agent versus the, the. Investment has been made into, into, into the algorithm like the, into the media buying. It itself and in terms of where the media gets deployed, which I think is like a parallel to what we're seeing as well, which is like, that's where the value is actually created most and going after, going after that sort of production labor layer in, in the process as it relates to the design of ads related to the building out of campaigns is, is not where it's being focused on right now or the values being captured, which I think is interesting.

[00:29:50] Richard Gaffin: Interesting, Tony.

[00:29:54] Tony Chopp: Well, AI

[00:29:54] Luke Austin: ai,

[00:29:55] Tony Chopp: taken my job yet. so yeah, I mean, I, I sort of echo Luke's sentiment like a year or so ago, six months or so ago. Like, I, I

[00:30:05] Luke Austin: I think

[00:30:06] Tony Chopp: the, the

[00:30:07] Luke Austin: the idea of like ai,

[00:30:08] Tony Chopp: of does the thing end to end was like. Maybe it'll, maybe we all jump the gun just a little bit.

[00:30:15] Richard Gaffin: Yeah.

[00:30:16] Luke Austin: But.

[00:30:20] Tony Chopp: where do I think it's going? So,

[00:30:24] Luke Austin: Well, here's, here's what I, here's what I 

[00:30:26] Tony Chopp: I

[00:30:26] Luke Austin: take the ai

[00:30:28] Tony Chopp: is an

[00:30:29] Luke Austin: enable

[00:30:29] Tony Chopp: It's an

[00:30:30] Luke Austin: enablement

[00:30:30] Tony Chopp: function that

[00:30:32] Luke Austin: function, makes, processing,

[00:30:34] Tony Chopp: of

[00:30:35] Luke Austin: processing,

[00:30:38] Tony Chopp: Better more robust,

[00:30:41] Luke Austin: but

[00:30:43] Tony Chopp: there's

[00:30:43] Luke Austin: very big caveat.

[00:30:44] Tony Chopp: to what

[00:30:47] Luke Austin: What is the implication for you?

[00:30:49] Tony Chopp: So, and I'll

[00:30:50] Luke Austin: I'll give you, for example,

[00:30:52] Tony Chopp: we,

[00:30:53] Luke Austin: we

[00:30:53] Tony Chopp: we

[00:30:54] Luke Austin: work with

[00:30:54] Tony Chopp: different

[00:30:55] Luke Austin: clients. 

[00:30:55] Tony Chopp: And every

[00:30:56] Luke Austin: every client we work with,

[00:30:58] Tony Chopp: everybody

[00:30:58] Luke Austin: has

[00:30:59] Tony Chopp: a source of truth. So we

[00:31:01] Luke Austin: staff

[00:31:02] Tony Chopp: STAs, we use STAs, and we,

[00:31:04] Luke Austin: work really hard.

[00:31:05] Tony Chopp: to get that to be accurate and representative. And we're actually going through an exercise right now to hold ourselves to a really high standard. That status is accurate and representative from a financial perspective for our partners that we work with. But it's really, really hard.

[00:31:22] Luke Austin: So

[00:31:23] Tony Chopp: Everybody

[00:31:24] Luke Austin: has.

[00:31:24] Tony Chopp: source. Everyone has this different source of truth. Is it triple whale? Is it North Beam? Is it Shopify? Is it Google Analytics? Is it status? So being is that the starting point of the information matters. So I, what I believe is context is, is really, really important to the to the subsequent analysis.

The integrity of the data and we're, what we're doing is we're ctc, we're trying to fold the AI layer into. A super clean, super sharp data set that we own, that we have in partnership with our clients, so that everything that it does past that it outputs. Whether it's analysis or, Hey, you should build this campaign or tweak that campaign or create this much creative or, or, or do this, don't do that.

Do the other thing from like a recommendation or a directional standpoint is all built on top of. Inputs that are really, really rock solid and represent a source of truth that a bunch of human beings all agree is the source of truth.

[00:32:38] Richard Gaffin: Yeah. yeah, so in other words that the, the LLM is only as good as what it's being trained on or whatever, or whatever the AI model is, right?

[00:32:46] Tony Chopp: Yeah, sort of, you know, in the, the training data set for GPT and a bunch of others has been like the Reddit comments from all of all of human history, so take it with a grain of salt, I guess.

[00:32:57] Richard Gaffin: Exactly. Okay, so let me get then, well, first, first I wanna know from both of you, like, how does AI show up in your personal workflows? Like, how are you, how are you actually using it? Whether it's GBT or CLO or whatever the case may be. What's, yeah, what is your actual personal use case for it? Let's go look.

[00:33:18] Luke Austin: I talk to it about everything. My best friend, my best friend, and my assistant.

[00:33:24] Richard Gaffin: yeah. There

[00:33:25] Luke Austin: like the simplest, simplest way about putting it, but like I, the the, and in some cases, well, like. It my like memory in some ways too, like helps me to close the loop, loop on things and then keep those conversations open like where?

In my mind, I'll have, I'll have like open conversations or open loops things that I'll be like thinking about. And what I can do is help to close those loops with the assistant of the ai, make a decision at that point in time, but then revisit that decision in a week and the same conclusion and thing that we had analyzed for a week prior, still there.

And then we can build off that based on the new information, which I think is similar to what Tony's saying, which is like. The information there is, but the history is there, and then we can iterate and expand upon that in a way that things don't get lost. I think for outside of personal, like for CTC we, operationalizing insights from across our data set of clients is, is ha, has to be a value prop of what we bring to the table. And so that's something we're thinking through of how we can more effectively do that through ai, by helping to capture the information in the context along the way, across our data set, and then being able to iterate and build upon that in a way that the nothing's, that there's less things falling through the crack cracks or hanging in silos of a two person Zoom meeting or a three person Slack DM message.

[00:34:47] Richard Gaffin: Yeah, Tony.

[00:34:52] Tony Chopp: Yeah, I mean there's, there's nothing that I haven't tried to use it for.

[00:34:55] Richard Gaffin: Mm-hmm.

[00:34:57] Tony Chopp: Personal use cases going to Palm Springs for play some golf this weekend. I was like, Hey, GPT, you know, from my previous conversations, I like this kind of mixture of value and coolness. Like, not the most expensive courses, not the, not the cheap, cheap and dirty ones like, I

[00:35:16] Luke Austin: I got, I'm gonna play two days out there.

[00:35:18] Tony Chopp: courses. I

[00:35:19] Luke Austin: Two question I I don't wanna make

[00:35:20] Tony Chopp: of the decisions at all. You know, you know what I

[00:35:22] Luke Austin: like you,

[00:35:23] Tony Chopp: sort of

[00:35:23] Luke Austin: certain things that help me.

[00:35:25] Tony Chopp: making

[00:35:25] Luke Austin: This process,

[00:35:25] Tony Chopp: Like just gimme the choice. Got an,

[00:35:28] Luke Austin: friend mine college,

[00:35:30] Tony Chopp: moved up to Hermosa Beach and I'm trying, getting 'em going, surfing.

So I was GPT sit. What are the surf

[00:35:36] Luke Austin: between.

[00:35:37] Tony Chopp: midpoint, between here and. in Hermosa Beach, you know, set it up for me. Gimme the, gimme the, the whole setup. So,

[00:35:46] Luke Austin: And I think this sort of mirrors like one of.

[00:35:48] Tony Chopp: that I, I was at a Google performance. Marketing conference earlier this year and the one of the, one of the folks that was speaking was talking about how we, how we learned to Google, you know, we learned how to use keywords and search and sort of speak in that language and how we, how we actually inquire about things is really messy. So I think the, the things that I love about GPT is I can give it my human messiness and have it help me organize and orient and prioritize. et cetera, et cetera. So, and, and ultimately arrive at a decision on the work side. A couple things, like practically, I, I have, I've gotten in the habit of setting up folders for various, various projects that I'm working on, or clients that I'm working on.

So it kind of gives me a space to, feed it the context. I use it a lot for data analysis. If I'm trying to seek out some pattern or something else, something that would've taken me, you know, a bunch of time in a spreadsheet, I, I guess maybe say it this way, I spend less time. Mucking around with spreadsheet formulas than I did two years ago.

A lot of that work is happening in, in GPT. And then to the other point that Luke made, I think is a project that I'm, I'm working on right now is trying to capture and synthesize contextual information. That happens in Zoom meetings and slacks and like in all these various places in, in dropping, dropping it into GPT and helping it have, it, help me synthesize the, the overall what's, what's happening here?

What, what are the takeaways? What are the, what are the big ideas that are, that are recurrent to fulfill this? You know, the, the promise to continue to fulfill the promise that, that Taylor's made forever, which is an agency, one of the bene, one of the value propositions of an agency is it's, it's not just Tony that you're hiring or Luke, you're hiring the experience of all of us together doing this across 60 or 80 clients or whatever, over 10 years.

So I think, I think the AI's ability to like. Take a bunch of information and synthesize it, and organize it and orient it. A bunch of messy human information, synthesize it, squish it, orient it, structure it. It's gonna be I think it's gonna be something that, that becomes more and more prevalent throughout our organization and, and many others.

[00:38:08] Richard Gaffin: Awesome. All right. Well, Luke had to jump, so we're gonna wrap it up here. And just a couple words, if you want your questions answered on the pod, just remember you can call us and leave us a voicemail. 8 6 6 dtc 2 2 6 3. you can send us a text there as well. We might read your question on a subsequent episode.

So for Tony the Chopper shop for Luke, the weatherman, Austin, I'm Richard, the Professor Gaffin signing off. Take care everyone. We'll see you next time.

[00:38:31] Tony Chopp: See y'all.