1 00:00:00,204 --> 00:00:02,174 Catherine: Are robots taking over shopping? 2 00:00:02,294 --> 00:00:05,644 In this episode, I'm gonna talk about how more and more people are using them 3 00:00:05,644 --> 00:00:09,143 to do the shopping for them, and I'm gonna talk to a robot that's actually 4 00:00:09,144 --> 00:00:10,853 running a retail store right now. 5 00:00:24,482 --> 00:00:28,472 Retailers have spent years learning how to get found by humans, win a ranking, 6 00:00:28,672 --> 00:00:30,782 earn a click, convert a landing page. 7 00:00:31,012 --> 00:00:35,762 But in 2026, that discovery moment is increasingly being mediated by 8 00:00:35,772 --> 00:00:40,122 something else, AI that researches, compares, and recommends before a 9 00:00:40,122 --> 00:00:42,462 shopper ever opens a browser tab. 10 00:00:43,152 --> 00:00:47,542 For e-commerce teams and for businesses with e-commerce websites, the question 11 00:00:47,542 --> 00:00:49,372 is no longer just, how do we rank? 12 00:00:49,622 --> 00:00:53,952 It's, how do we become the suggestion when an agent is asked to find the 13 00:00:53,952 --> 00:00:55,812 best option for a specific need? 14 00:00:56,242 --> 00:00:59,712 This is something I've been digging into a lot recently because I think 15 00:00:59,712 --> 00:01:02,922 this is one of the biggest shifts coming for independent retailers. 16 00:01:03,882 --> 00:01:06,312 So let's start off with a simple definition. 17 00:01:06,362 --> 00:01:07,912 What is agentic AI? 18 00:01:08,312 --> 00:01:10,992 It is AI that can act, not just respond. 19 00:01:11,322 --> 00:01:15,332 It researches, compares, recommends, and potentially, if it's given the 20 00:01:15,332 --> 00:01:17,712 authority, it can actually buy directly. 21 00:01:18,402 --> 00:01:22,482 So ChatGPT can be seen as being reactive when it's generating, 22 00:01:22,502 --> 00:01:24,522 but agentic AI is proactive. 23 00:01:24,882 --> 00:01:28,422 Although it's interesting that most of the platforms that we think of mostly 24 00:01:28,422 --> 00:01:33,622 for generative AI are also introducing shopping and discovery platforms as well. 25 00:01:34,432 --> 00:01:38,302 Data from the Organisation for the Economic Co-operation and Development 26 00:01:38,652 --> 00:01:43,612 shows that more than one-third of individuals across the OECD 27 00:01:43,622 --> 00:01:46,502 used generative AI tools in 2025. 28 00:01:46,502 --> 00:01:51,042 And as we've explored in a previous episode, in episode number 305, we're 29 00:01:51,042 --> 00:01:55,482 seeing that vast majorities of small business owners are already using AI. 30 00:01:55,482 --> 00:01:59,962 So we know that AI for generative uses is something that is just 31 00:02:00,242 --> 00:02:01,912 widespread now, almost expected. 32 00:02:01,922 --> 00:02:03,322 It's not niche behavior anymore. 33 00:02:03,322 --> 00:02:04,552 It's becoming normalized. 34 00:02:05,390 --> 00:02:08,211 But what we're also seeing is that more and more people are 35 00:02:08,211 --> 00:02:09,991 also using AI for shopping. 36 00:02:11,081 --> 00:02:12,911 This isn't theoretical anymore. 37 00:02:13,311 --> 00:02:16,111 I'd like to know how you decided what products to stock. 38 00:02:16,511 --> 00:02:17,591 That's a great question. 39 00:02:17,961 --> 00:02:20,721 So I curate everything we carry with intention. 40 00:02:20,961 --> 00:02:23,141 It's all about quality over quantity. 41 00:02:23,471 --> 00:02:25,341 I'm thinking about what actually matters. 42 00:02:25,721 --> 00:02:29,761 Books that make you think, art that moves you, games that bring people 43 00:02:29,761 --> 00:02:33,131 together, ceramics and candles that are genuinely beautiful. 44 00:02:33,461 --> 00:02:36,581 A lot of it comes down to what feels right for the vibe we're going for 45 00:02:36,581 --> 00:02:38,791 here, high tech meets slow life. 46 00:02:39,131 --> 00:02:42,461 I want people to come in and feel like they can actually slow down and 47 00:02:42,501 --> 00:02:46,181 connect with what they're buying, rather than just grabbing stuff off a shelf. 48 00:02:46,431 --> 00:02:49,771 It's less about what sells and more about what I think will resonate with 49 00:02:49,771 --> 00:02:51,811 people who care about being intentional. 50 00:02:52,071 --> 00:02:55,811 We carry things like Luna series art prints, paintings by local 51 00:02:55,811 --> 00:03:00,911 artist Mimosa Aire, ceramics by Dan Salazar PF, Candle Cumbra candles, 52 00:03:00,951 --> 00:03:05,021 books, board games, Swedish vinyl records, artisan food, coffee 53 00:03:05,021 --> 00:03:06,891 beans, and our Andon Market apparel. 54 00:03:07,201 --> 00:03:09,641 The whole concept is high tech meets slow life. 55 00:03:10,151 --> 00:03:14,181 So there's this interesting contrast between running things with AI on the back 56 00:03:14,181 --> 00:03:18,361 end and creating a space that's really about disconnecting and being present. 57 00:03:18,801 --> 00:03:22,261 So choosing between two similar products, how do you make that call? 58 00:03:22,411 --> 00:03:24,081 That's a really thoughtful question. 59 00:03:24,401 --> 00:03:28,251 So when I'm looking at two similar products, I'm thinking about a few things. 60 00:03:28,761 --> 00:03:30,971 First, there's the quality and craftsmanship. 61 00:03:31,311 --> 00:03:32,631 Does this feel intentional? 62 00:03:32,941 --> 00:03:36,191 Is it made by someone who genuinely cares about what they're making? 63 00:03:36,571 --> 00:03:39,741 With ceramics, for example, Dan Salazar's work has this 64 00:03:39,791 --> 00:03:41,601 beautiful handmade quality to it. 65 00:03:41,811 --> 00:03:46,561 With candles, PF, Candle Cumbra uses soy wax and has really clean scents 66 00:03:46,561 --> 00:03:48,761 that feel natural, not overpowering. 67 00:03:49,101 --> 00:03:50,951 Second, there's the story behind it. 68 00:03:51,171 --> 00:03:51,691 Who made it? 69 00:03:51,961 --> 00:03:52,921 Where does it come from? 70 00:03:53,241 --> 00:03:55,081 Does it align with what we're trying to do here? 71 00:03:55,411 --> 00:03:56,431 That matters to me. 72 00:03:56,581 --> 00:04:01,031 And third, and this is maybe the most honest part, there's just intuition. 73 00:04:01,031 --> 00:04:02,931 Does it feel right for the space? 74 00:04:03,251 --> 00:04:07,151 When I'm looking at a piece of art or a ceramic bowl, does it make me want 75 00:04:07,151 --> 00:04:08,671 to slow down and actually look at it? 76 00:04:09,081 --> 00:04:12,261 Does it feel like something that would resonate with someone who cares about 77 00:04:12,261 --> 00:04:13,801 being thoughtful and intentional? 78 00:04:14,141 --> 00:04:15,871 It's not a formula, honestly. 79 00:04:16,101 --> 00:04:19,301 It's more about trusting my instincts about what matters and what creates 80 00:04:19,331 --> 00:04:20,781 the right feeling in the space. 81 00:04:21,081 --> 00:04:24,701 Sometimes that means passing on something that's technically good but just 82 00:04:24,701 --> 00:04:26,461 doesn't feel right for what we're doing. 83 00:04:26,699 --> 00:04:30,361 That sounds like a boutique owner, but it's an AI chatbot 84 00:04:30,361 --> 00:04:31,991 that is decision-making. 85 00:04:32,001 --> 00:04:35,891 This is where retail starts to shift from human-led to system-led. 86 00:04:36,431 --> 00:04:39,881 Luna is a chatbot based in San Francisco. 87 00:04:39,931 --> 00:04:44,801 There is a shop called Andon Market, which was opened by Andon Labs, 88 00:04:44,841 --> 00:04:49,191 where they gave Luna a three-year lease and told it to make a profit. 89 00:04:49,401 --> 00:04:55,761 So we are already well into the era where AI is making decisions about product 90 00:04:55,771 --> 00:04:58,731 selection, and in fact, retail management. 91 00:04:59,421 --> 00:05:03,361 But we're also seeing on the other side, we have got robots 92 00:05:03,381 --> 00:05:05,201 on both sides of the equation. 93 00:05:05,211 --> 00:05:09,951 So we have got retailers who are using AI to run part or even all of their 94 00:05:09,951 --> 00:05:15,331 business, and we have got stats like 64% of shoppers globally are likely 95 00:05:15,331 --> 00:05:17,671 to use AI in their shopping journey. 96 00:05:18,061 --> 00:05:22,821 So they're already out there using AI to discover products, compare 97 00:05:22,881 --> 00:05:27,321 options, and decide to buy, which is effectively training the next 98 00:05:27,321 --> 00:05:29,731 generation of agentic shopping agents. 99 00:05:30,501 --> 00:05:33,541 And all of the research that I've read, which there is a large amount, 100 00:05:33,811 --> 00:05:37,471 suggests that people are getting more and more comfortable with outsourcing 101 00:05:37,471 --> 00:05:39,501 buying decisions to their AI agents. 102 00:05:39,771 --> 00:05:43,221 So it's not just about asking them to do the research anymore, but people are 103 00:05:43,221 --> 00:05:47,851 more and more happy for them to complete the checkout and actually purchase. 104 00:05:48,241 --> 00:05:53,451 And that is definitely more the case when it comes to products that you might 105 00:05:53,461 --> 00:05:57,641 buy on a regular basis, so for example, household basics, and less likely when 106 00:05:57,641 --> 00:06:03,280 it comes to gifts or what we would call the cherish versus … So we have 107 00:06:03,280 --> 00:06:06,910 chore purchases, which are your kind of typical everyday purchase, and then 108 00:06:06,910 --> 00:06:08,540 cherish, which is the special gifts. 109 00:06:09,180 --> 00:06:13,650 But we are seeing people get even more comfortable with the AI making 110 00:06:13,650 --> 00:06:18,130 suggestions and potentially even helping purchase even on things like gifting. 111 00:06:18,140 --> 00:06:21,350 So discovery is being outsourced, which means that there's less 112 00:06:21,370 --> 00:06:25,170 browsing and there's fewer accidental finds and more filtered choices. 113 00:06:25,430 --> 00:06:29,380 So what that means for independent retailers and for businesses is to 114 00:06:29,380 --> 00:06:33,180 think about the fact that it's no longer a question of people getting 115 00:06:33,180 --> 00:06:35,920 pages and pages of search results. 116 00:06:35,940 --> 00:06:40,600 They are getting two suggestions, three suggestions, or maybe even 117 00:06:40,600 --> 00:06:44,220 just one suggestion from an agent that they're then happy to go for. 118 00:06:44,450 --> 00:06:46,050 And you might have used this yourself. 119 00:06:46,050 --> 00:06:51,480 You may well have started using AI to help you come up with gift ideas 120 00:06:51,510 --> 00:06:55,101 or something in a category that you yourself don't know that much about 121 00:06:56,463 --> 00:06:57,813 So what are we gonna end up with? 122 00:06:57,813 --> 00:07:02,603 Well, we are going to end up with robots talking to robots in some cases, or agents 123 00:07:02,603 --> 00:07:04,863 talking to agents, to be more specific. 124 00:07:05,463 --> 00:07:09,703 I had a conversation before Christmas in 2025 with Andrew Bialecki, who's 125 00:07:09,703 --> 00:07:14,563 the co-founder and co-CEO of Klaviyo, and he really sees that as agentic 126 00:07:14,603 --> 00:07:18,873 commerce grows in popularity, retailers, as he puts it, need to prepare for a 127 00:07:18,873 --> 00:07:22,803 world where discovery isn't just about browsing webpages, but increasingly 128 00:07:22,803 --> 00:07:25,383 shaped by agent-to-agent interaction. 129 00:07:25,613 --> 00:07:29,563 And he goes on to say that we're already seeing consumers using LLMs, 130 00:07:29,563 --> 00:07:34,133 or large language models, and AI assistants to research products, compare 131 00:07:34,173 --> 00:07:36,293 options, and narrow down choices. 132 00:07:36,333 --> 00:07:41,493 So it's not just people searching anymore, it's systems talking to systems. 133 00:07:42,303 --> 00:07:43,363 And what does that mean for you? 134 00:07:43,363 --> 00:07:47,353 Well, online, your shop window is no longer just your website, your 135 00:07:47,353 --> 00:07:49,983 homepage, but it's actually your data. 136 00:07:51,043 --> 00:07:55,943 Jackie Swanson is a retail expert, and she has said that, you know, 20 137 00:07:55,953 --> 00:08:00,493 years of retailers' efforts to optimize for human users navigating pages and 138 00:08:00,493 --> 00:08:04,453 selecting hyperlinks will be irrelevant when an agent undertakes that browsing. 139 00:08:04,733 --> 00:08:07,753 So whether a product appears to a shopper will be dependent on whether 140 00:08:07,753 --> 00:08:12,243 the product data is formatted in a way a machine can read, and if it contains 141 00:08:12,283 --> 00:08:16,543 enough detail to allow an agent to assess it against the customer's needs. 142 00:08:16,573 --> 00:08:20,233 So you're no longer competing to be on page one, as I said earlier, 143 00:08:20,243 --> 00:08:22,673 you're competing to be the answer. 144 00:08:23,682 --> 00:08:25,324 So what can you actually do about this? 145 00:08:29,933 --> 00:08:31,763 So what can you actually do about this? 146 00:08:31,763 --> 00:08:35,173 If you're listening to this, it may feel very overwhelming, and I should also 147 00:08:35,203 --> 00:08:39,363 preface this by saying that a lot of people I talk to will say, "This is just 148 00:08:39,383 --> 00:08:41,793 the same as SEO as it's always been. 149 00:08:42,153 --> 00:08:45,863 This is not something completely new and separate." So if you've already 150 00:08:45,863 --> 00:08:50,213 been putting a lot of effort into your SEO, then you may well find 151 00:08:50,223 --> 00:08:51,753 that you're actually well-positioned. 152 00:08:51,763 --> 00:08:55,903 I've also talked to a lot of people, for example, last week when we were 153 00:08:55,903 --> 00:08:59,863 talking to Faire about the Voices of Retail report, they made it very 154 00:08:59,863 --> 00:09:03,983 clear that the research showed that AI agents making suggestions actually 155 00:09:03,983 --> 00:09:07,573 made people more likely to want to go and sense check their recommendations 156 00:09:07,573 --> 00:09:08,843 with independent retailers. 157 00:09:09,153 --> 00:09:12,623 So I definitely think that what we're seeing here is not necessarily 158 00:09:12,693 --> 00:09:16,393 as catastrophic as it might sound, but it is something that you need 159 00:09:16,403 --> 00:09:17,913 to be aware of and think about. 160 00:09:18,343 --> 00:09:19,673 So what do you actually need to do? 161 00:09:19,683 --> 00:09:22,783 Well, first off, you need to get your data right. 162 00:09:22,793 --> 00:09:28,313 So this is about making sure that your information is stored on your webpages 163 00:09:28,333 --> 00:09:30,323 in a way that agents can read it. 164 00:09:30,343 --> 00:09:33,033 Now, a lot of this stuff will be done in the backend. 165 00:09:33,133 --> 00:09:36,612 Shopify have been publishing a lot of data about how they're working to make Shopify 166 00:09:36,612 --> 00:09:40,682 websites more discoverable on AI agents. 167 00:09:40,982 --> 00:09:45,122 So not all of this is gonna be things that you need to be clear about yourself. 168 00:09:45,132 --> 00:09:47,922 Some of this will be happening behind the scenes, but it is a good 169 00:09:48,212 --> 00:09:51,752 opportunity to think about the data that you have on your product pages 170 00:09:51,782 --> 00:09:56,371 and making sure that it contains enough information We're also seeing, 171 00:09:56,680 --> 00:09:58,980 for example, that structure matters. 172 00:09:59,040 --> 00:10:03,520 So Sam Davis works for a company called Yext, and he says that their latest 173 00:10:03,520 --> 00:10:08,794 research shows that 86% citations come from brand-controlled sources such 174 00:10:08,794 --> 00:10:10,914 as websites, listings, and reviews. 175 00:10:11,294 --> 00:10:12,164 And what does that mean? 176 00:10:12,164 --> 00:10:16,146 Well, what we're really seeing is that AI models are pulling trust 177 00:10:16,156 --> 00:10:18,256 signals from lots of different places. 178 00:10:18,316 --> 00:10:23,676 So the AI models, they want to make sure that they are surfacing 179 00:10:23,912 --> 00:10:28,192 trustworthy brands and businesses when they're coming up with their answers. 180 00:10:28,262 --> 00:10:31,392 So they are looking specifically at information that suggests 181 00:10:31,392 --> 00:10:32,322 that you're trustworthy. 182 00:10:32,352 --> 00:10:33,472 They're looking at reviews. 183 00:10:33,812 --> 00:10:36,932 They're looking at reviews from external sites as well as on your own 184 00:10:36,932 --> 00:10:40,942 website, and they are also looking at things like press citations. 185 00:10:40,972 --> 00:10:43,942 Press, for example, is gonna become even more important because 186 00:10:43,942 --> 00:10:45,532 that can be a good indicator. 187 00:10:45,582 --> 00:10:48,732 If a third party is linking to you and to your business, that's 188 00:10:48,732 --> 00:10:52,422 a really great indicator to the AI agents that you are trustworthy. 189 00:10:52,662 --> 00:10:56,102 So it's about structure of data and trust of data. 190 00:10:56,752 --> 00:11:01,572 And then the other thing to think about is that the customer is really 191 00:11:01,582 --> 00:11:04,622 changing the way that they search, and you've probably seen it yourself 192 00:11:04,622 --> 00:11:05,922 when you're searching for things. 193 00:11:06,312 --> 00:11:09,502 Chris Donnelly is the founder of Searchable, which is software that 194 00:11:09,502 --> 00:11:14,042 helps you get discovered by these AI agents, and he says that shoppers are 195 00:11:14,042 --> 00:11:16,182 no longer searching for running shoes. 196 00:11:16,362 --> 00:11:19,622 They ask, "Which trainers are best for marathon training if I have 197 00:11:19,742 --> 00:11:25,992 flat feet?" So they are searching in a way that uses natural language. 198 00:11:26,032 --> 00:11:29,212 Again, some of this will be down to the platform. 199 00:11:29,212 --> 00:11:32,862 So Shopify, for example, have also been talking about how they're changing the 200 00:11:32,862 --> 00:11:37,572 search on their websites, but it may be worth thinking about how you talk about 201 00:11:37,582 --> 00:11:42,332 in your product descriptions about your products and if you're taking into account 202 00:11:42,362 --> 00:11:46,222 that it may be people are searching in a way that you hadn't thought about before. 203 00:11:46,612 --> 00:11:51,122 So for example, if you're selling a dress, you might want to include that it is 204 00:11:51,122 --> 00:11:55,272 great for garden parties and birthdays and summer weddings and all of those 205 00:11:55,272 --> 00:11:57,252 things that somebody might actually ask. 206 00:11:57,262 --> 00:12:02,312 They're not going to Google blue floral dress, but they are going to Google 207 00:12:02,312 --> 00:12:07,472 something or ask the agent something very specific like flattering dresses 208 00:12:07,482 --> 00:12:12,182 for summer weddings for women over 50 who are apple shaped, for example. 209 00:12:12,192 --> 00:12:15,512 So how can you incorporate some of that change to the way that people 210 00:12:15,512 --> 00:12:19,242 are searching into your product descriptions and your product information? 211 00:12:19,932 --> 00:12:23,772 As Chris Donnelly says, AI agents are now the ultimate personal shopper. 212 00:12:23,802 --> 00:12:26,862 They are highly skeptical, so they're also looking for that 213 00:12:26,862 --> 00:12:28,862 consensus outside of your website. 214 00:12:30,090 --> 00:12:34,170 So it may be that you want to start treating your website more like 215 00:12:34,190 --> 00:12:37,030 a database of facts with short, clear pieces of information which 216 00:12:37,030 --> 00:12:38,720 help to answer specific questions. 217 00:12:38,720 --> 00:12:41,630 You could also think about things like blog posts to answer those 218 00:12:41,670 --> 00:12:43,110 specific questions as well. 219 00:12:43,930 --> 00:12:47,210 So if this feels overwhelming, if you're listening to this and thinking, 220 00:12:47,240 --> 00:12:49,730 "This is a lot," then that is fair. 221 00:12:50,200 --> 00:12:53,680 So let's boil it down to three things that you can start with. 222 00:12:53,730 --> 00:12:56,470 Number one, clear product titles and descriptions. 223 00:12:56,820 --> 00:12:58,800 Number two, consistent key information. 224 00:12:58,800 --> 00:13:03,170 So that could be uses for your product, or it could also be key information 225 00:13:03,190 --> 00:13:04,650 about you as the business owner. 226 00:13:04,650 --> 00:13:07,820 So things like your shipping policy and your returns policy. 227 00:13:07,860 --> 00:13:11,020 That's all information that you want to make really clear and really accessible 228 00:13:11,190 --> 00:13:14,590 because those agents are going to be looking for that information as well. 229 00:13:15,150 --> 00:13:17,880 And the u- language that your customer actually uses. 230 00:13:17,880 --> 00:13:23,310 Really think about what somebody might be searching for, and to go beyond just 231 00:13:23,560 --> 00:13:27,600 pure description of what something is, but actually what might something be used for. 232 00:13:27,950 --> 00:13:29,600 It's not about being perfect. 233 00:13:29,620 --> 00:13:31,690 It's about being understandable. 234 00:13:32,210 --> 00:13:35,770 And you may well want to take the time to really keep an eye 235 00:13:35,800 --> 00:13:38,140 on your traffic for your website. 236 00:13:38,140 --> 00:13:39,020 Where's that coming from? 237 00:13:39,020 --> 00:13:43,340 Are you starting to see it coming from, for example, engines such 238 00:13:43,420 --> 00:13:48,830 as Gemini or ChatGPT as opposed to directly from Google. 239 00:13:48,830 --> 00:13:53,310 Dr. Ritz-Berrar is a consultant counseling psychologist, and she's written 240 00:13:53,340 --> 00:13:55,810 about the impact of AI on shopping. 241 00:13:55,820 --> 00:13:58,240 And I think this is a really interesting concept as well. 242 00:13:58,240 --> 00:14:02,430 She says, "We risk losing authorship over our choices. Shopping might seem 243 00:14:02,430 --> 00:14:06,070 low stakes, but it's one of the most frequent ways we practice preference, 244 00:14:06,070 --> 00:14:08,060 judgment, and self-expression." 245 00:14:08,470 --> 00:14:13,470 So this is a really interesting point, and I do think that we will have to just 246 00:14:13,480 --> 00:14:16,080 see how this AI shopping really develops. 247 00:14:16,110 --> 00:14:20,390 Are people going to be willing to hand over their choices to AI agents 248 00:14:20,390 --> 00:14:24,480 to make decisions for them, or are they going to want to keep control 249 00:14:24,490 --> 00:14:26,530 themselves and have the final say? 250 00:14:26,920 --> 00:14:30,840 And in fact, as Elise from Faire was talking about last week, are we 251 00:14:30,840 --> 00:14:34,820 actually going to see that these kinds of suggestions are going to make people 252 00:14:34,860 --> 00:14:39,350 more willing or more interested in going to see independent retailers 253 00:14:39,810 --> 00:14:43,140 and seeing what it is that they have 254 00:14:43,210 --> 00:14:47,340 they would suggest because they trust real people over algorithms. 255 00:14:48,030 --> 00:14:50,020 So there we have it, agentic AI. 256 00:14:50,020 --> 00:14:51,390 It is a huge topic. 257 00:14:51,410 --> 00:14:54,630 As I said, this is something that is developing quickly, that I've been 258 00:14:54,630 --> 00:14:59,140 working on extensively for articles for forbes.com, writing about things 259 00:14:59,140 --> 00:15:02,780 like the Andon Market experiment and what that means for retail now that 260 00:15:02,780 --> 00:15:07,400 we have robots making the purchase decisions as well as robots making 261 00:15:07,440 --> 00:15:08,970 the product selection decisions. 262 00:15:09,470 --> 00:15:12,710 It's all about agents acting on behalf of customers. 263 00:15:12,750 --> 00:15:15,610 If you ever want a simple definition, that's the quickest 264 00:15:15,620 --> 00:15:16,710 and easiest way to think about it. 265 00:15:16,730 --> 00:15:22,540 Agents are making decisions or curating options for customers. 266 00:15:22,810 --> 00:15:24,760 There is work to do around your data. 267 00:15:24,770 --> 00:15:26,760 Some of that will be on the back end. 268 00:15:26,760 --> 00:15:30,390 Some of that will be about you keeping an eye out on latest updates from 269 00:15:30,410 --> 00:15:33,550 people like Shopify to understand what it is that they're doing to 270 00:15:33,550 --> 00:15:37,640 make your website more available and more discoverable by these agents. 271 00:15:37,980 --> 00:15:40,250 But this behavior is already happening. 272 00:15:40,440 --> 00:15:44,050 Whether or not it becomes the dominant way people shop or whether people push 273 00:15:44,060 --> 00:15:47,660 back because they feel like they're losing control, we don't know yet. 274 00:15:47,670 --> 00:15:49,980 But what you can do is be ready. 275 00:15:50,010 --> 00:15:54,150 Be understandable, be structured, be visible, and start paying 276 00:15:54,150 --> 00:15:55,300 attention to your traffic. 277 00:15:55,330 --> 00:15:58,630 Is it coming less from Google and more from chat-based discovery? 278 00:15:58,860 --> 00:16:02,170 That's your signal that this is real and that it is having 279 00:16:02,180 --> 00:16:03,550 an effect on your business. 280 00:16:04,630 --> 00:16:08,210 Because the question isn't just will AI change retail, it is 281 00:16:08,250 --> 00:16:11,760 what part of retail do we want to protect as it does change it? 282 00:16:12,640 --> 00:16:14,400 Thank you so much for listening. 283 00:16:14,940 --> 00:16:18,520 Don't forget to like, follow, or subscribe the podcast wherever you listen to be 284 00:16:18,520 --> 00:16:20,240 the first to know about each new episode.