← DTC Podcast
DTC Podcast
Ep 615: Prepare Your Brand for Agentic Commerce (How LLMs Are Collapsing the Consideration Phase)
00:36:26
Loading summary
Transcript62 lines
- [00:00]
A
Before we jump into today's all killer no Filler episode, a quick word about who makes this show possible. The DTC Podcast is brought to you by Pilot House, the performance agency behind some of the fastest growing DTC brands in the world. Creative Media and Customer Journey all under one roof, performance and brand without the trade off. Every Friday we hand the mic to a Pilothouse operator to break down what's actually working in their space right now. Want a team that treats your growth like their own? That's Pilothouse. Head to Pilothouse Co and now on with the show.
- [00:31]
B
Really important conversation that I think a lot of people are not aware of right now is how do I sell to an agent?
- [00:39]
C
We haven't had to change our behavior as marketers. Now this presupposes like we're actually going to have to change what it is that we do. No longer am I heading to Amazon to compare products or read reviews. I'm allowing an LLM to narrow the shortlist for me.
- [00:57]
B
If you make that wrong call and you're too committed to a Persona that doesn't exist or is uninterested in your product, it's not just a strategic miss anymore. It's like your brand's a flop. Goodbye. You're not gonna be able to run a business.
- [01:17]
A
This episode is brought to you by Triple Whale. If you're still making decisions based on messy scattered data bouncing between ad platform spreadsheets and dashboards, you're not alone. Most e commerce teams are stuck piecing together a story they don't fully trust. Triple Whale changes that It's a complete intelligence platform built for modern operators who are done guessing. Triple Whale centralizes all your data, equips you with measurement tools you can actually trust, and layers in best in class AI to turn that data into clear recommendations, showing you what's driving growth, what's hurting profit, and exactly what to do next. It even takes action for you. With Triple Whale, you can quickly spot trends, catch issues before they impact revenue, and uncover opportunities you would have otherwise missed. It's about moving faster, making smarter decisions, and operating with real confidence. That's why brands like True Classic, Ouai Press, Juicery, and thousands of the fastest growing DTC companies rely on Triple Whale every day. They're saving time, increasing efficiency and unlocking insights that simply weren't possible before. If you're ready to stop guessing and start growing with clarity, now's the time to check it out. Visit triplewhale.com that's T R-I P L-E-W-H-A-L-E.com to learn more. It's all killer, no filler. And I'm Eric here with Daniel and Aves from Pilothouse. I just came off my trip to LA where I was at the Whaleys and this will all be, I think, disseminated out through our content. But the big announcement there was triple whale launch of mobi2, which is their agentic movie media buying assistant or media buyer, essentially that allows you to advertise agentically 247 with these agents, which was, which was really. There's a lot of interesting opinions around it, but that's not exactly what we're talking about today. The topic that you guys brought to me today was more about the agentic shopper and like what actually what our customers, how our customers are using these platforms and kind of what we can do about it. Aves, do you wanna like kick us off with a bit of your preface that you were talking to me about?
- [03:20]
B
Yeah, for sure. I think I am someone who, like Charli xcx, I love to be early to things. And so something that happens a lot in digital marketing is we all get really swept up in a singular topic. And for the past two years that's been how are we as marketers using AI? But it's a really easy question to answer from my perspective and there are very simple solutions. And Eric, you were saying we'll seminate some of this in our content, but people are actively using it. Big brands are using AI to do marketing. But the interesting thing for me has always been how does the average consumer scrolling their feed, first of all, feel about AI? Second of all, how do they use AI? That is a much more interesting and valuable conversation because at the end of the day, as marketers, we're trying to influence human behavior and decision making, but humans are outsourcing their decision making to agentic AI. So the question stops being how can I prompt most efficiently and what can I automate versus what can I use AI for in more of a thinking capacity? And the really important conversation that I think a lot of people are not aware of right now is how do I sell to an agent properly? Because coming into Q4, like this year, just the amount of people who are going to use agentic AI to say, find me a really nice gift for my mom who likes to stay inside, or my mom who's like a big wine drinker. Like that's the kind of question that people are asking already of their, their LLM. So it's only going to Increase, like, we'll talk a little bit more in the episode, but the amount of trust that people have in their agents is far deeper than the amount of trust that they have in a brand showing them an ad on Facebook. Right? So there's so much opportunity there, and I'm just really excited to talk about it and start working on it because it's cool and it actually does shift the landscape of digital marketing. And I think the way that we all thought using the tools was gonna shift the landscape. This is actually what's, what's shifting the landscape.
- [05:30]
C
We haven't had to change our behavior as marketers. Like, we're only doing the same thing that we've always done, better and faster. And like, what AVES is talking about is like, now this presupposes, like, we're actually going to have to change what it is that we do because, like, yeah, AI is about to collapse a whole bunch of what we're used to into a lot of a muddier, more murky path for a consumer. Right? For a marketer rather than a consumer.
- [05:57]
A
For a marketer. For the consumer, it's cleaner because they have this agent who knows so much. Every time I'm talking to my ChatGPT and I'll ask it a question, I'll be like, well, because I know you're this, this, and this and this and this, and you have this problem and this problem and this problem. I'm like, whoa, whoa, whoa. I'm just like, I forgot that you know all of this st about me. Have you either of you bought something agentically yet?
- [06:17]
C
I mean, like, define agentically, Eric. Like, are we talking about like autonomous checkout or are we talking about like using recommended product?
- [06:26]
A
Recommended? I would say, I would say, yeah, 100%.
- [06:29]
C
I've recently done it to, for some vitamins that were recommended to me by a physician. And like comparing, like, what was, what was best. And that's the really interesting point of that kind of agentic trade off here is we're not saying, AI, go buy this for me, but no longer am I heading to Amazon to compare products or read reviews or even start to look for recommendations. I'm allowing an LLM to narrow the shortlist for me and then I find that I'll research maybe the top two or three choices that are recommended. But I've stopped deciding what is worth buying and like having like abdicated a lot. At least, you know, much of the consideration to ChatGPT and likewise, you know,
- [07:20]
A
a couple of the big proof points. I was just Doing some research before this, in 2025, during Q4 holiday season, I for. I forget what the source was, but it was about 20% of people were estimated to have used an LLM in their purchase path, which is already crazy. Thinking back to the holiday season last year, that it's already 20%. The other proof point that's interesting is from an agency I always look at who's driving the leads to Pilot House kind of thing, and I think it was last month that ChatGPT was both the top referral source and the top closing rate. We're used to when the lead comes from, like, DTC or, you know, from the newsletter, from the podcast, we have a really great close rate because people are already kind of warmed up to us. But I think we're seeing as of last month, ChatGPT is even got a higher conversion rate, higher price point of. Of brands who are kind of looking for our services. So it's definitely happening.
- [08:13]
B
Yeah, like, the intent is there, right? Like, it's such a different mindset you're in. I keep using because Daniel's always wearing one of two hoodies. So I keep using this example, but
- [08:25]
C
it's like you zoom out on the
- [08:27]
A
Maduro hoodie you got on.
- [08:29]
C
I just realized I am literally wearing the same hoodie I was in the last time I appeared on All Killer, no filler. That's sad.
- [08:36]
B
We do love him. But I always, I've been using this example for Daniel is like, if Daniel decides I want a third hoodie, that's like, maybe we go beige this time. We've got like, dark gray, light gray. We're going beige. If he, like asks Chat what is a great brand for a, like, beige hoodie, he's in a much different mindset than just like, kind of fooling around on YouTube or Google and searching with kind of 50% intent to buy. But mostly you're just researching. But people are, like, already at the point of spending money when they're asking an LLM. Like, you got, I would say, like 50% intent. When you're googling and you're researching and you're looking for brands for vitamins or that new hoodie. But if you're asking your LLM, it's like, you're pretty ready to buy. And you also, because you trust it, you're pretty ready to buy what it recommends.
- [09:29]
C
It's interesting you mentioned Eric, like, the funnel changing a little bit for brands like Pilot House. And like, it doesn't surprise me with B2B at all. Like, the More like a category like makes you think or compare or like like doubt your purchase or even ask for recommendations and advice. Like the more exposed it is to to AI shopping. So like vitamins obvious one like I but like health and wellness like we're going to see it with like health care supplements are going to be huge. Those are the primary use cases where like there's a massive amount of choice and a lot of opinions and like that AI just helps us mediate those things.
- [10:08]
A
Can we just talk real quickly? Was going doing a little research on like how this has kind of been rolled out and when it was first rolled out, I guess in like early March, ChatGPT had launched a way that you could just one click buy from the the Chat GPT interface. And I feel like they've rolled that back a little bit and pushed the checkout experience due to like maybe some hallucination or not really understanding or being able to translate the Shopify data into Chat GBT properly. But now it's more of a. It's pulling the information in from Shopify. But you're still checking out on Shopify and the actual storefronts, correct?
- [10:41]
B
Yeah, I think rolling it back is a great way to put it because I think this is happening a lot with most OpenAI products where we put it into the market and then it's not good and we take it back and then we have a better version of it. I do think right now it's like recommendations and you are manually checking out. But some of the work that I am seeing my friends who work on more of the Shopify dev side starting to do is prepping checkouts for. For completely agentic checkout. I don't know. That does make sense. There's a lot of checkout but it's like that's kind of the direction it's heading. I think it's like premature right now, but I think in less than a year most people will be really comfortable giving chat their entire financial information. To make that checkout itself, Google I
- [11:27]
C
O just demoed a feature in the upcoming Android set where user walked up to a concert poster, took an image, was prompted. Would you like two tickets for this concert and press? Yes. And that was the extent of the transaction. So like talk about like abdicate. Like I didn't even see a price confirmation in there.
- [11:48]
A
Right.
- [11:48]
C
Like I'm sure they edited it out for effect but like that's closer. Like that's being built into the next operating systems that are all going to be rolled out to our phones soon enough. Right.
- [11:59]
A
I think your point, Abe's, about this being quite often for problem aware people. It's, it's probably cannibalizing Google maybe a little bit more than it's cannibalizing Meta, which is more of a true top of funnel play where you're, I go on Meta or Facebook pretty much just to shop at this point, just to see the clothing brands that I like in the ads. But what does this do to the funnel? What, what, what, what do these LLMs, it's basically kind of like eating the middle out of the funnel in a way. Right? Where exactly is that? Right.
- [12:27]
B
Yeah, we've talked about, like Daniel and I, we've been talking about the way that it crushes a consideration window because I can like again the hoodie example, I can hit Daniel with like 10 ads for a cream hoodie and that is going to cost me a lot of money as a marketer. But if I figure out how to put together a agentic centric strategy and we are showing up, when Daniel asked that question that we were asked earlier, like, where do I get a cream hoodie? The consideration window is just squashed completely. And so it's can be, I think, very efficient for brands to start thinking about this because then you don't have to be hitting a singular person with, you know, 10 touch points before they buy. It's a couple true top of funnel awareness ads on Meta and then that person's prime, they're interested, they're problem aware and then they go for it.
- [13:17]
A
So what do we think we do here? Like, what is the play? I think of Pilothouse and I think of like, we've sort of blasted Pilothouse into the ether for like 1200, you know, newsletter issues. All these podcasts that are kind of leading back to Pilot House as this source of authority. We haven't done almost anything to optimize our pages in some sort of SEO kind of capability. But how should brands think about structuring their content in a way to make themselves more discoverable?
- [13:46]
C
It's interesting that you asked that, Eric. Like that. Like, as marketers we were rewarded in the past for like, like traditional search rewarded us for visibility. So we were doing okay because like in, in that sense, like think all of the tendrils that we can send out into the ether, they're all roads back to the brand. Right. The change happens now is the agentix search rewards not just visibility, but more so clarity, credibility and evidence. And so just by establishing the experience and the expertise of a podcast, like dtc that's gone a long way in the pursuit of visibility. We've also clarity built credibility in the pursuit of visibility and that's going to ultimately be what drives agentic search and will be rewarded through agentic search. So it's not altogether that different from SEO, but it is kind of like an evolution of it, right? Really falling back to the old kind of like Google's rewarding experience, expertise, authority and trust. Like that's traditional SEO, but it's also extends into AEO and GEO or whatever you want to call the next frontier of marketing.
- [15:08]
A
How should we think about that Venn diagram? Because I understand SEO and AEO do have a have overlap. I know, I know SEO for years has gone away from this idea of like technical SEO that you're trying to talk to the bots essentially who will serve you to really being as useful as you can for the customer. And I know AEO is about that as well. But what, what parts of the process do we need to be thinking about that we weren't thinking about as much for SEO? Where do the Venn diagrams diverge?
- [15:37]
B
One thing that I think is very interesting is a large difference is almost reps. So continuing to enforce the same brand identity and truth. The thing that is maybe like this is the most comfortable hoodie for working from home to like. I think what we used to sort of assume is putting up a couple blog posts that were like SEO specific on our site. It was going to kind of flip the bill when it came to background SEO work. But now what LLMs are crawling for is recent artifacts of that being the truth. And so really simple. But adding dates to these blog posts to make sure that they're current and relevant, really helpful and finding ways to reinforce that truth or your sort of brand identity across different mediums as well and searching for ways that you can create either earned mentions or consumer kind of customer LED mentions. So a site on its own is not really enough to have you treated favorably by these LLMs. Because I could make a website right now that was like I am the best boxing coach in the world and I can really like rig it up to hopefully platform myself on LLMs. But if I don't have any earned media, so any PR placements, anyone else saying that what I'm saying is credible, then an LLM is not going to treat me as a credible source with only a website. So your actual consumer sentiment starts to matter a lot and then real old fashioned earned media starts to come back in a really important way because LLMs are like sort of amalgamating all of these different touch points to really put in front of you what they feel is a accurate answer to your question.
- [17:33]
A
And it kind of goes back to what we have been talking about, about Andromeda for years as well, which is this idea of not just understanding kind of what the customer wants, but like who the customer is that wants it and why they want it. Kind of creating these, these arcs for like the customer journey. And it's like because these limbs they know, like I was saying, they know more about me than is comfortable even having Claude being a desktop. How I use Claude. And every time I come back, it's back at it. Eric, you're. It's just building, it's building more trust with me that it knows me and knows me better. So it's like the conversion rates are only going to improve, but it's also looking. So when it's looking for like the best I was actually the one thing I've looked for is a I want to wear a ring. I want to like get a nice like pinky ring kind of thing. And so. And it knows me. And it, and it like. And it starts. It gave me all these options that were. Because it knows me so well and it's looking for content that's from a voice like mine. Not not just a customer, but a customer voice like mine. Right?
- [18:29]
B
Yeah. Yeah, that's a really good way of putting it. Like an important part of formulating how you want to start showing up and like what type of person or what type of use case you want to have LLMs recommend you for. You really have to deeply understand who you're servicing and then understand. Is that delusional of you to think that you can do that? I think for some brands it's like you might have an idea that you're Persona is somebody and that they. What's a good example? The make believe mug company. As per usual, it's like you might have an idea that you're talking to young professionals and they like your mugs because they are aesthetically pleasing. But if you dig into your actual consumer insights and see that it's middle aged women and they like that mug because it's leak proof and they throw it in a baby bag. If I start producing content on my site and through organic channels and I'm really pushing this idea of young people and aesthetics, I'm not going to rank favorably because what people are actually saying in real life is that the mug is leak proof and it's Great for a baby bag. So that kind of tension split. Previously it was just unstrategic and you were kind of getting yourself stuck in a tactical spin cycle. But now you are actively losing money in a way because you're not going to be surfaced. And it's like if you make that wrong call and you're too committed to a Persona that doesn't exist or is uninterested in your product, it's not just a strategic miss anymore. It's like your brand's a flop. Goodbye. You're not going to be able to run a business.
- [20:03]
C
Yeah. In this area, like, it's not just about what you're. Like you're publishing about your brand. Like, like, so the brand is no longer what you publish. It is like what gets summarized about you. And that summary is, is generated not just from what you're in control of. Right. But it's what an LLM is learning about the audience that is attracted to you as well. So you abdicate a little bit of like, control of the brand in that sense. So you do have to recognize that like brand storytelling is not only about what you want people to believe, but what people are believing about you a little bit. Not to put too clean a wrapper on it, but yeah, it's interesting.
- [20:43]
A
What do we do? So with clients that come in and they want to understand this world, like, what are the practical steps that brands can take to I guess first understand how they're showing up and then to start putting together a plan about how they want to show up and how to get there. What do we advise brands?
- [21:01]
B
I love this and I feel like Daniel might too, because it's really like kind of old school real brand work. But it's really about going up a little bit and answering those questions that we were sort of just discussing. Almost separating your thought process. Now when you think about brand and separating, what is my human focused brand storytelling piece of the pie and then what is my agentic lead shopping, brand enforcement side of the pie and marrying those two together. And this is something I know, Eric, you've talked to lots of our E commerce strategists about and I've talked about on Adventurous time and time again, which is doing the work first to understand who your target Personas are and understanding your segmentation, understanding how brand storytelling sits there and then marrying that with work around social listening, cultural insights, what are people actually saying and where is the need? And then Daniel mentioned this earlier, like the activities aren't that different, but the higher level thought has to be a little bit different.
- [22:02]
C
To that end, like it's a good leading question to ask a partner, a brand that you're working with. It's like what gets you excited about creator marketing, for example, like the old answer would have been like, oh, I get access to new audiences and like I get reach. But like what really excites us in the agentic age is like, oh, now we're like, we get to lean into experience, we get to lean into and borrow some authority from that creator. It's not just about reach and impressions, it's about third party comparisons that we can start to pull into the net. So I love working with creators, I love working with creators because they recognize that what they're providing their audience is like that earned experience, that authority, that trust. And as a brand, if we're welcomed into that partnership, we can tap into it to mutual benefit. Like that's, that's the exciting part about the creator relationship and how it's kind of like evolving, you know, to use one one channel as an example.
- [23:04]
A
What are categories that are going to be most affected by agentic shopping?
- [23:10]
B
Like everyone is, which is really interesting and fun and an opportunity. It's not scary, but particularly Daniel was mentioning kind of CPG vitamins. They rank highly amongst what people are querying. Also consumer electronics. So what version of the iPhone should I get? Should I get an iPhone at all? What headphones should I purchase? That also ranks, I think at the very top. A study based out of Boston on how people are using a gentic AI to shop and consumer electronics was by far the top. But. But I think we'll start to see more uptake in apparel, more subjective verticals as well. As we continue to kind of commodify community in this digital space, I think we'll start to see more people use it for those subjective verticals. But right now it's a lot of CPG and a lot of consumer electronics
- [24:03]
C
travel too, for the reasons that AVES has cited. There's just a lot of reviews, a lot of comparisons and a lot of word of mouth. Right?
- [24:13]
B
Yeah, Anything where there's like a ton of information that you could take into account when making a decision that tends to be leading the charge in a way. But I also know I've talked to like our Amazon team and I did an adventurous episode about it as well. Like Rufus, their sort of magentic AI, only on Amazon is really interesting as well because people are almost being trained, if you're an Amazon user, to use Rufus in a very specific way which is asking for product details, asking for like more general recommendations. And I do think Amazon rolling out Rufus is going to really sort of shift and open people's eyes to how they can be using, you know, Manus, Claude, whatever they're using in a different capacity because they're already doing it with Rufus.
- [24:59]
C
As of last week, Rufus is dead and it is now Alexa for shopping. And like typical Amazon way they've gone back and forth on the name. But I think technically it's it's Alexa for shopping now instead of Rufus.
- [25:13]
B
Long live Rufus.
- [25:14]
A
Rufus collapsed into Alexa Aves. You mentioned something early on and I think it is really worth highlighting, which is the gift buying aspect. I think so many people this coming holiday season are going to be using LMS to buy the best, you know, most unique gifts for their friends and family. What what are some thoughts here in whatever we're in Q2 about really gearing up for the holiday gift buying season with these LLMs.
- [25:41]
B
I mean I feel like you guys were boots on the ground, but like do you remember those? I only remember the concept of them. Like the buzzfeed, like here are the top 10 funny mustache t shirts you can buy. I think we'll see a resurgence in those pay to play kind of article styles. Not that they ever went away, but I do think the listicles will come back in a big way. It's like very dead Internet theory but starting to think about if you've never been placed in a listicle like that and you don't have the experience to do it, that's going to be a huge one. I also think making sure that you're like including gifting language or highlighting gifting related reviews on site will be really key as well. So trying to balance from here to Q4, both opportunities that are external to your business. So that's kind of the listicle dark posting creator stuff where you're looking for earned trust and authority and then also starting to look at well on my site, do I anywhere speak of gifting or who this is a good gift for and understanding before you start doing that work like who is it a good gift for? Because that's like based on how people search on Amazon it's usually person forward, right? Like gift for dad gift for a teenage girl. So really digging into like the psychology of like who's buying for who and who is that final who and starting to put information in context on your own site and then also looking to those outward outlets.
- [27:12]
A
What happens? We've talked since the beginning of this podcast about the value of landing pages and funnel congruency. So when you're advertising to one of your avatars, you have an ad and then you go to a landing page that speaks to that avatar a little bit and then, and then you go to checkout with the middle part of the funnel being collapsed. Does that make the landing page or the post click experience more important or less important?
- [27:34]
B
That is a really interesting question. It makes it in my experience so far more important in a different way. Because your site experience now has to be cohesive to an agent operating it. Right. So a lot of people aren't, I don't think thinking about that right now. We're not like all talking about WebMCP. It's not a super sexy topic to dive into, but it's like your site needs to be optimized for agent traffic immediately. Because if your site can't be navigated and an LLM can't find the information that it needs, then you're not going to be surfaced. Right. It's like super simple. So the stuff that we used to talk about all the time, Eric, where it was like, oh this ads all about this product doing X, Y or Z. So you're going to put them on a landing page that continues that story in a way that is becoming less and less of the like name of the game is people stop doing that work and they don't want to go through. Like I remember at some point I was probably talking on this podcast about I was running people through like a three stage funnel for high consideration products. It was like ad advertorial, fake PR PDP and people were happy to do that. And now that would be absurd to pitch to any client to be like, oh, it's going to be added and there's three stages. So I think it's more about like back end work and making sure this is a big one. I've been talking to people about too. If you rely a lot on in cart upsells, be prepared to lose that incremental revenue. If you're not thinking about how you can lace that into a agentic process. If you are like we talked about travel, an experience focused travel agent, Expedia or something. If an LLM in the back end can't operate all the options that you have on your site, it's like what's going to happen to you? It's just going to realize it can't and then leave and float someone else.
- [29:25]
C
What about all of us out there who use the in cart as their wish list. Right. Like it's just going to fundamentally change all the way that we shop. Eh, that's interesting.
- [29:34]
B
Yeah, yeah. Like abandoned cart emails. Another piece of the puzzle that's so deeply impacted by this because you're not adding to cart yourself, you're not doing that like here's my little secret wish list. So you're not getting the opportunity to talk to that human and say here's an offer or it's almost sold out. So you have to make the content so much more compelling and focused and targeted to a Persona and then yeah, you've got to make sure you're thinking about how to add those in card upsells. That's a big part of your business.
- [30:05]
A
Can we end here on some thoughts about what Pilot House can do? Like what, what, what when brands come in today, what are first of all, some ways they can start future proofing their business and what are some ways that as our bolt on growth services that we would help brands prepare for this future?
- [30:22]
C
In my opinion, it begins and ends with kind of like that user intent. We've discussed that before. That kind of like search intent really shows us what people are confused about, like what they compare what they fear and what they like they need to believe before they buy. So for us that approach can take like that intelligence and like basically turn it into like an overall creative system. And that can be paid media. It could be content, it could be site experience, it could be proof, it could be performance learning. So like Pilothouse, on behalf of the partners and the brands with whom we work, we will help them become legible not only to humans, which we've been very good at up to this point in time, but also like increasingly to the AI systems that are like going to like increasingly shape what they're trusting and what they're choosing.
- [31:11]
B
Another way of putting it how I think about engaging with partners and helping them navigate this is almost cutting out the noise at this point in time and really figuring out some of these topics we've talked about diving in deep to set that strategy from a high level for humans versus agentic led strategy in terms of content and how we communicate. What's really nice about working at Pilot House and with the large team that we have is it's not siloed to okay, here's the brand work and then I'm going to hand it to someone that I rarely speak to and they're going to do the translation and disseminate this across multiple channels. It's like I will hop On Huddle with Joe, go another podcaster and we'll like talk immediately about how we can translate that higher level strategy to content across, you know, whether it's life cycle, larger marketing or like site experiences as well, and actual Shopify work. So I think the real benefit that we've been bringing to partners in navigating this ultimately huge shift in E commerce is being able to set the strategy from the high level, but then operate on a tactical way across every medium to make sure that that brand truth is not getting diluted out there in the world and making sure that it's consistent because consistency is the way you're going to win. With LLMs and Agent 2, how do
- [32:42]
A
we audit how we're doing? How do we audit whether brands are showing up? Is it a matter of just going to like an invisible chat so they don't know it's you and then really just digging deep on like trying to get it to prompt, Like I'm looking for a jewelry brand that does this kind of ring and blah blah blah and just seeing if you can get it to come up. Like how do we actually audit whether we are benefiting from LLMs?
- [33:04]
B
That's a good one. What I've been doing is like I guess lying to an LLM and creating a version of a target Persona and then asking from there. Because I find the blank slate approach both to therapy and also research doesn't always work when it comes to LLMs or humans I guess. So it's like really thinking about crafting that Persona and saying I'm this type of person and then asking those questions and figuring out who's being ranked well and then really looking at like what are the activities that a brand's doing. This is how I've been sort of auditing it. On the side is how brands currently operating and then just are they thinking about it. And something I've always looked for but is more important now is I try to find through lines in their content of like a singular perspective. And if I can't, if every ad is a different creator talking about a different angle that pertains to a different person, that's also a really good sign that a brand needs assistance with a lot. Let's say if I look at an ad library and that's going on, they can help with a lot. But it's a good way to tell if there's. I've been using this a lot but like brand reinforcement happening. If everything's really scattered, then you know, there's not like a singular truth that they're consistently putting out there in a very reductive way.
- [34:24]
C
Eric what I like to do is like, I like to take a look at like what the queries are inbound from from Google. Like essentially understanding like traditional search queries. Kind of like evolve those into prompts and then like build out a little matrix so that if I were a brand, big or small, I'd come up with a small list of prompts, build a matrix across LLMs like ChatGPT, Gemini, Claude to begin with, and then literally just pump those prompts in understand what competitors are mentioned, understand what sources are being cited and if we're mentioned at all. And that is the foundation. Not everyone needs a massive investment of a huge tool to understand if they're visible, if they're being cited accurately, and if they actually have some recommendation share I would say to your point, just go out and start plugging your brand into these answer engines and understand what they're outputting. And it all just starts with measurement in that sense.
- [35:20]
A
Well, we are one step closer to my Everyone has a butler theory that I talk about on this podcast all the time. And our own personal AI butler who knows us better than we know ourselves. Sometimes to a scary extent. I I'm interested to see like, yeah, my first experience with this ring. I think I. I've thought about this ring a lot. It found a really great one. I'm going to find it, gonna probably buy it and then, and then we can talk about it on the next one.
- [35:41]
B
But do an update on whether or not it like worked well.
- [35:44]
A
Pinky ring. But yeah, look forward to kind of keeping in touch with you guys as this revolution continues to roll on. Because it's definitely happening.
- [35:52]
B
Yeah.
- [35:52]
A
Nice.
- [35:53]
B
I'm excited.
- [35:53]
A
Thanks guys.
- [35:54]
C
Thank you so much, Eric.
- [36:01]
A
Thanks for listening to today's episode. If you're not getting the DTC newsletter, you can subscribe for free at directtoconsumer co. And if you want to learn more about Pilothouse's all killer no filler services, take off to Pilothouse Co. I'm Eric Dick and this has been the DTC podcast. We'll see you next time.