WEBVTT 00:00:05.000 --> 00:00:12.000 Hello everybody. Welcome to the webinar. We're going to get started in just a few minutes. We'll give people a few more minutes to filter in. Thanks. 00:02:26.000 --> 00:02:43.000 All right, everyone. Let's get started. So thank you for joining. I'm Dan Buczaczer, the VP of Marketing here at Constructor. And today we're discussing our latest State of E-Commerce report, which is a comprehensive look at how shoppers buy online in 2025. 00:02:43.000 --> 00:03:05.000 This report is based on a survey that we did just last month in August of 2025 of 1,500 consumers in the United States and Europe. Respondents were balanced across a mix of income levels and age groups. And this is the third year that Constructor has done this survey. And this year we actually joined forces with Shopify, which we're extremely excited about. 00:03:05.000 --> 00:03:24.000 In the past, we focused the survey exclusively on product search and discovery. But this year, with Shopify aboard, we actually looked at the entire shopping funnel from finding the right place to shop all the way through to checkout. And we saw some significant shifts from even a year ago. So, we're really excited to share those with you all. 00:03:24.000 --> 00:03:39.000 And I'm also excited about who we have here to discuss the report today and add some context. So, first of all, from Constructor, I'd like to introduce the person who has headed up this report since the beginning, which is Nate Roy. Nate, do you want to give a quick introduction? 00:03:41.000 --> 00:04:16.000 Hey, good morning everybody, or good afternoon depending on where you're dialing in from. My name is Nate Roy. I oversee product marketing, brand, and content, and like Dan said, primary author of the State of E-Commerce report, which I'm super excited to talk about today. I've been in commerce technology for most of my career at this point. I was at a company called Salsify, if you're familiar with product information management or product data syndication. So I've seen this space evolving a lot over the last decade and I think we're in the middle of a really pivotal time for e-commerce. So really excited to dive into this more today and super grateful to have Shopify involved this year. 00:04:18.000 --> 00:04:26.000 Well, speaking of helping us dive in and speaking of Shopify, from Shopify, we have Solutions Engineer Paiman Parmaei. Paiman, introduce yourself, please. 00:04:28.000 --> 00:04:54.000 Yeah. Hello everyone. I'm Paiman. I'm part of the Solution Engineering team focused on the enterprise space at Shopify. I actually focus on consumer goods vertical. I've been at Shopify for 5 years. Prior to that, I worked at SAP in a similar role. And I couldn't agree more with what Nate said. It seems like the industry is changing. AI is changing everything. So I think this is a very timely time to talk about this report and the findings from the report. 00:04:56.000 --> 00:05:24.000 That's great. So a piece of housekeeping here. Please feel free to ask questions and hopefully you will have a lot of questions either about the survey itself or about some of the interpretations and context being added by this team. So there's a Q&A button you'll see at the bottom of this Zoom box and we'll either address it along the way or we will have a dedicated area for Q&A at the end. But please start those questions coming as you have them. 00:05:24.000 --> 00:06:07.000 Well, great. With that, let's dive in. So, the way we structured the State of E-Commerce report and the way we're going to review our findings today in this webinar is by dividing the shopping journey into three phases. Define, refine, and decide. And to be clear, this is not a perfectly linear funnel where they go from one always into the next stage and always directly into the third. In fact, this year the data from consumers suggests that they're moving back and forth between these three areas more fluidly than ever before making a purchase. And we'll actually get into a bit more of that. But this is still a useful organizing principle for thinking about the three major phases of buying online. 00:06:07.000 --> 00:06:38.000 So with that, we will jump right into the first one, which is define. And the mindset of a consumer in the define stage is essentially that they have an idea or a goal in mind, but they may not know exactly what type of product they need yet, and they're looking for some sort of help. They might be actively searching with something very specific in mind. They may be just looking for broader inspiration, but it's where they need to first start figuring out what they want or what they think they need to buy. 00:06:38.000 --> 00:06:58.000 So some of the noteworthy statistics that came out of the survey this year: 68% of shoppers say retailer search needs an upgrade. 68%, well more than half. And very similarly, 66% claim that they abandon to Amazon when results disappoint. So that is their continual go-to when they're not getting what they need from a retailer website. 00:07:00.000 --> 00:07:28.000 Also, we now are seeing Instagram, Facebook, and TikTok all at very similar levels in terms of the number of people who cite them as major discovery sources when they're shopping. And this is across generations, interestingly, with some generational differences. The main one being that TikTok is especially dominant among Gen Z. 58% of shoppers in Gen Z are saying that TikTok is a place for discovery for them, for what to buy. 00:07:28.000 --> 00:07:42.000 So, let's dive in. I have a few questions for you all based on this. Nate, let's start with that stat on the last slide. 68% of shoppers say retailer search needs an upgrade. Constructor's in the business of retailer search. Does this finding surprise you? 00:07:44.000 --> 00:08:07.000 Yeah, it's a great question and I would say it's a little bit of yes and no. I think what was most surprising to me having run this report for a few years in a row now is actually how flat this number is. So it was 68% this year. It was actually also 68% in last year's survey as well, which means at least in the perception of the consumer, search hasn't improved in the past year. 00:08:07.000 --> 00:08:44.000 Unpacking a little bit into why that might be, we do see that shoppers are getting more comfortable using conversational style queries in discovery mode, probably because they're getting used to querying things like ChatGPT that way. Especially when their original intent is kind of on the vaguer side. So, a query like "comfortable shoes for standing all day," that could mean sneakers to one shopper. That could mean if you're a nurse, maybe it's Danskos. It could mean insoles to somebody else. And most search engines are going to struggle to bridge that nuance and are probably going to show the same exact results to all of those three different profiles. 00:08:44.000 --> 00:09:14.000 So, if you actually dig into the report a little bit further, 86% also said that they have to reformulate their queries at least once, meaning they type a query, they see some results, they don't like those results, and so they go back to the search bar and actually type in something different or try to get more specific with the query. And each time they have to do that, because sometimes it's several times, is potentially a loss in confidence for that retailer. 00:09:14.000 --> 00:09:48.000 And that's how the perception builds that something like Amazon might be the competitor that just works in a way that retailer sites don't. So I think the opportunity for retailers especially is to stop treating search as static keyword match. And I think a lot of folks are already sort of coming to that understanding on their own. But to really start implementing AI, especially AI that incorporates in-session shopping behavior to really uncover that in-the-moment intent. 00:09:50.000 --> 00:10:08.000 So, implementing AI being extremely important, I feel like, is what you're saying. You mentioned this readiness to flee to Amazon, which feels like this constant threat now. And then let's loop in on the next page. This increase in social media as a source of inspiration. Do we get the sense are shoppers on their way to being done with retailer websites? 00:10:11.000 --> 00:10:36.000 Yeah. You know, it's kind of funny. I mean, I want to state as clearly as possible off the bat, in my opinion, no. Absolutely not. Shoppers are not done with retailer sites. One of the other findings in the report that was pretty interesting was that 45% said that they actually don't care whether a product was recommended to them by a human or by AI as long as it's the right thing ultimately. 00:10:36.000 --> 00:11:14.000 And I think that's really important to keep in mind because we specifically asked directly about AI as a touch point in that question, but I think if we asked this about any other touch point like social media or something else, we probably would have gotten the same answer because shoppers have never really cared about what channel necessarily was delivering the product to them as long as it's the right thing in the right context at the right time. And so I think there have been a lot of different points over the last decade or so where a new technology gets introduced and we all start talking about, "Is the retail site going to disappear or go away?" And I remember this happening when social commerce first became a thing like Instagram checkout, Facebook checkout. But retail sites haven't gone anywhere. 00:11:14.000 --> 00:11:58.000 And maybe the role has shifted a little bit. And maybe shopper expectations for the experience they're going to get when they land on a site has increased. But at the end of the day, every retailer still has loyalists who go to them and will look to them for guidance, especially in more trend-based industries like apparel or beauty. Plenty of first-time visitors will land from a Google search or from an LLM search as well. And so in that moment of define that we talk about where a shopper tells you exactly what they want in their own words in the search bar, that's still one of the most valuable parts of the on-site experience to make an impression and show them that you really know them. 00:11:58.000 --> 00:12:52.000 And the retailers who are most ahead of the game are responding to that by investing to make their sites true destinations that people want to visit with innovative experiences like their own AI agents or even simple things like very creatively merchandised collections. We talked about social platforms like TikTok and Instagram. They do a great job of sparking demand and sending people into shopping mode that might not already have been. Especially with younger consumers. And LLMs are going to add just yet another discovery channel. So, I don't think that means retailers lose. I think it just means that the websites need to be ready to capture that demand with AI-fueled discovery that feels as seamless and as tailored as whatever it was that inspired them in the first place. So if that was on social or if that was on the LLM, that feeling needs to carry over when they eventually get to the retailer site. 00:12:55.000 --> 00:13:24.000 Yeah, I agree with that point about compelling experiences. We are going to dive into LLMs in more detail a little bit later in this webinar, but Paiman, Nate mentioned a couple of times social media sites. I know you do a lot of work with merchants who obviously have to integrate with social media sites. How should retailers be thinking about these social media platforms? Are they their enemy? Are they their friend? How should they be approaching that? 00:13:26.000 --> 00:14:14.000 That's an interesting question and I think we have to kind of take a step back. Shopify's viewpoint has always been your merchants should be where the buyers are, right? And there are some crazy stats around social media usage. Actually, 94.5% of internet users use a social media platform on a monthly basis. So, that's clearly where the buyers are. And Shopify's viewpoint has always been that merchants should show up exactly where the buyers are. And that's exactly why we've invested heavily in making the integrations with TikTok, Instagram, YouTube Shopping. And most of the time we've been first whenever these platforms added shopping capabilities to their platform. So for example with TikTok, with Instagram, with YouTube Shopping, all of those Shopify was first. 00:14:14.000 --> 00:14:55.000 So we really see it as a competitive advantage and we know that customer behavior changes a lot, right? Buyer behavior changes a lot. We saw how TikTok kind of blew up over a short period of time. And now with LLMs in place, we expect that to again happen. And one of the competitive advantages that Shopify provides is it really makes it easy for merchants to show up exactly where the buyers are, especially as the buyer behavior continuously changes. So Dan, I know we're going to talk about LLMs, but obviously that's a space that's on everyone's top of mind in terms of how do you show up? How do you make sure that you don't miss this next trend that takes place? And excited to talk about that part, too. 00:14:57.000 --> 00:15:15.000 Yeah, I think it's a sign of the times that how many times we're talking about LLMs before you even get to the LLM section. I think it's very clear it's on everyone's mind. But I love that perspective in terms of you have to be where the shopper is going versus trying to fight that. So let's talk about where the shopper is going. Let's dive into the next section actually, which is refine. 00:15:15.000 --> 00:15:56.000 And so this is the phase now where a consumer is thinking, "Okay, I know the type of product I want, I have some detail around it, but now I need help choosing, making that specific choice." So some of the findings from this phase: first of all, 41% of shoppers say their favorite retailer still treats them like strangers. So, here they come for hopefully some more personalized advice and yet you may all know that feeling of someone at your favorite store who you've gone to again and again and they're still treating you like they've never seen you before. And only 20% report seeing results that are frequently personalized on one of their websites. 00:15:56.000 --> 00:16:35.000 Some additional findings: 50% of shoppers find sponsored listings with skepticism or frustration. In other words, when they see those ads that are kind of run for other items to buy that are sort of run over to the side, 50% are viewing those with not a lot of trust. And actually 25% would pay to avoid ads entirely. On the flip side, interestingly, 32% have purchased from a sponsored listing. So at the same time, they are certainly showing some utility. So we'll dive into what we do about some of those conflicting stats. 00:16:35.000 --> 00:17:10.000 And in terms of LLMs, 64% of shoppers, and think about how just a few years ago, none of us even knew how to define an LLM, which by the way, we probably should define as a large language model. So, we're talking about things like ChatGPT, Perplexity. 64% of shoppers have reported trying one of these GenAI tools as part of their shopping journey. And that's a 13 percentage point jump year-over-year. And 58% now say they are very or somewhat comfortable using AI for shopping help. 00:17:10.000 --> 00:17:24.000 So, you can see why our panel has started to name some of these stats. These are some pretty, maybe not surprising given everything we're hearing, but certainly confirming it's not just media hype. This really has become part of the behavior of shoppers. 00:17:24.000 --> 00:17:41.000 But let's jump back to before we quite dive into LLMs. Some of those first stats were about personalization and sort of the lack of personalization so many shoppers are seeing. Paiman, let's go back to you. What has Shopify seen retailers do to try to improve personalization? 00:17:44.000 --> 00:18:01.000 Yeah, the way that we look at this is let's look at the foundations, right? In order for you to be able to provide the personalized experience, you need to have the right data, right? As we talked about, you're seeing your buyers across many different surfaces, whether that's social media or anywhere that you're selling your products. 00:18:01.000 --> 00:18:27.000 So, our core belief, Shopify's core belief is a unified data model provides you with the best opportunity to provide that personalized experience. So regardless of where your customer is purchasing from or where they're starting their journey from, whether that's in a retail store, in a POS system, or whether that's online through social media platforms, whether that's LLMs, having that single data model is really important. That's exactly what Shopify provides. 00:18:27.000 --> 00:19:02.000 So now you have a single view of the customer regardless of where they're coming in from, regardless of where they started their journey from. That's the basis, the foundation of how we look at it at Shopify. And from there now we have invested heavily in customer segmentation. So you have all the data. How do you segment the customers? How do you really truly provide that personalized experience by being able to segment the customers properly with AI? I know I've said AI multiple times already. It makes it really easy. 00:19:02.000 --> 00:19:36.000 So we have invested in Sidekick and Sidekick allows you to create these customer segmentations by just describing them. So you can just describe, "Hey, I want to create a customer segment of everyone who has spent more than $150 on my store," and Sidekick has the ability to create that segment for you. So really, we just think that the unified data model is the basis of you being able to provide that personalized experience, whether that's on the storefront, whether that's at checkout, whether that's providing specific loyalty or personalized offers. That's really what sets you apart. 00:19:40.000 --> 00:20:00.000 Well, I know we agree emphatically with that on the Constructor side in terms of centralized data feeding a bunch of experiences leading to doing personalization effectively. So, Nate, I'd love to hear from you. What's your take on why shoppers are feeling like strangers on the sites they frequent most and how do we fix that? How do we view that from the Constructor side? 00:20:03.000 --> 00:20:32.000 {{CTA:Download the State of E-Commerce Report|https://info.constructor.com/state-of-ecommerce-2025}}Yeah. I guess on the technical side of the house, speaking from the retailer's perspective, it comes down to two things. There's how you're tackling relevance and how you're tackling personalization. I think on the surface today, most retailers can deliver pretty decent relevance. So, if I go on an apparel website and I search for black hoodies, more than likely I'm going to see black hoodies in the top line of results. 00:20:32.000 --> 00:20:55.000 The problem is what shows up first and in what order. So, if the order of those results doesn't match what's most interesting to me as an individual. For example, if I was shopping for myself, if I saw women's hoodies first or if I saw hoodies that had poor reviews or that were out of stock in my size, based on my intent in the moment that is not a good result. Even though technically based on the eye test, and I think a lot of times that's how this stuff is tested in QA, that looks relevant but it's actually not a good experience. 00:20:55.000 --> 00:21:24.000 The second issue is personalization, and personalization has evolved a ton over the past 15 to 20 years starting with segments and now we're kind of in a world where one-to-one personalization is starting to become the expectation. But a lot of retailers I think are still struggling with personalization that on the customer experience side still feels too simplistic. 00:21:24.000 --> 00:21:57.000 So, for example, a lot of sites lean solely on repeating past purchases or pushing the same brands that I might have clicked on before. That is helpful in certain contexts. So I don't want to say that's the wrong approach completely, but if that's the only thing the site is capable of doing, it looks like personalization again from the eye test, that looks decent, but if you're actually in the customer's shoes and experiencing that and that's not what your intent is in the moment, it's not helping them with the mission that they're on right then. 00:21:57.000 --> 00:22:16.000 Two recent examples from my own life: I just recently bought a new iPad. And I started seeing recommendations across that same retailer site for the next three days for more iPads, even though I'm probably not buying another iPad for hopefully at least the next three years. 00:22:16.000 --> 00:22:39.000 Similarly, football season just started. I just bought a new NFL jersey. I started getting recommendations for more jerseys. I'm probably not buying another jersey this year. Maybe they could have recommended the new hats that came out from New Era, some sort of similar accessory that would have made a lot more sense. But instead, I'm seeing the same three Nike jerseys, even though I just bought one. 00:22:39.000 --> 00:23:02.000 So, I think when relevance is that shallow sort of technical relevance and personalization is really just showing your own buyer history back to you, like looking in a mirror, even a familiar site that on the surface is doing these things feels to the customer like the site doesn't know them, if that makes sense. 00:23:05.000 --> 00:23:42.000 Nate, it could be that you're just buying a jersey from the wrong team and maybe they're trying to nudge you to buy from the right team, but that's a separate podcast. So interestingly, speaking of recommendations and getting it right versus not, we also saw those statistics about sponsored listings or advertising on retail sites. Now, on the one hand, I come from advertising, and there's obviously an ongoing love-hate relationship with ads as consumers. There always will be. So, in some ways, those numbers don't surprise me. We hate advertising, yet it often works and can lead to a purchase. But what do you make of those numbers? Is there something retailers can do here to improve this for those that are utilizing sponsored listings? 00:23:46.000 --> 00:24:07.000 Yeah, I think the main takeaway from this one is ads do work when they line up with the shopper's intent. And they fail when they're just noise layered on top of other noise. I think what most shoppers want is just alignment. If they're looking at product suggestions, search results, both organic and paid, all of that should kind of be pointing them in the same direction toward a good result that meets whatever mission it is that they're on. 00:24:07.000 --> 00:24:45.000 So, I don't think at all the path forward is to kill ads. I think they're a great revenue stream. I think it's just to make them feel less like ads and more like a very intentional part of the experience. So, that I think requires unifying on a technical side both the on-site and the off-site personalization. The search bar, landing pages, the sponsored slots should all be reading from the same behavioral signals about what the shopper is there to do. And so when that happens, those ads stop feeling so disruptive and they start feeling actually like a service. 00:24:45.000 --> 00:25:25.000 Constructor's retail media solution does approach it this way where we hold the paid results basically to the same standard as organic results. So, we don't feel like you have to have revenue competing with customer experience. You should be able to maximize both of those things at the same time. Most retail media networks today run with two brains. There's a shopping brain, so that's your core e-commerce search engine that's serving all the organic results. And then there's a separate paid media brain, that's your ad platform. And those systems are rarely talking to each other. 00:25:25.000 --> 00:26:11.000 So the direct result of that is that shoppers get irrelevant ads. I literally was shopping on a site a couple of weeks ago where I was looking for work boots because I was getting ready for the fall. And I saw an ad for red high heels, which is totally just completely not even in the same category of what I was looking for. And so definitely killed trust for me, but I think it also wasted the ad spend for the brand. I think it damaged the credibility of that retailer a little bit. So with Constructor you can integrate all those signals, the organic and the sponsored, into one brain where that reinforcement learning creates a virtuous cycle basically where you see higher relevance on both sides, better performance, greater trust from the shopper, and then of course most importantly, that revenue. 00:26:14.000 --> 00:26:53.000 That's great. It gets back to that point Paiman was making about a single source of data to feed everything. I think that's that same point made a different way. And interestingly, when you go back to the statistics about how well TikTok and Instagram are doing, I think those are platforms that do a really good job of gaining inspiration from both organic content that's being posted by influencers or your friends as well as advertising. And sometimes you don't even tell the difference as you're scrolling. It's all sort of inspiration for you. And retailers, I agree with you, can absolutely create that same type of experience on their own websites. 00:26:53.000 --> 00:27:05.000 All right, Paiman. We are going to wait no longer. Let's go to that third slide again which had statistics about LLMs. It's time we dig in. So would love to hear, obviously Shopify looks over such a huge set of retailers. How are you seeing your customers adapt to LLMs? 00:27:09.000 --> 00:27:50.000 All right. Finally I can talk about it. When you think about LLMs, I think it's important to talk about a concept. So there's this concept called declarative versus imperative UI. And I think the best way to explain it is to go through an example. We all have had the experience. You go on a website, you're about to order something, you're like, "Well, I kind of want to find out what the return policy is." The imperative experience, which is the experience we've all been used to over the last decades, is I scroll down. I'm like, "Would it be in the terms of services? Would it be in the FAQ? Where would it be?" I click on a bunch of buttons, get frustrated, go Ctrl+F, try to find the return policy, and finally I find it. 00:27:50.000 --> 00:28:19.000 Right? That's the imperative experience where I do a bunch of clicks in order to achieve my intent. The declarative experience would be you ask an AI agent, "What is the return policy?" And it replies back to you. We think we've gotten used to the imperative experience because LLMs didn't exist previously. But it's obviously not a great experience. And we think this declarative experience where you just declare your intent and then the software takes care of it is going to be more dominant across the board. 00:28:19.000 --> 00:29:02.000 So the way we're thinking about it at Shopify and the way that we see successful merchants think about it too is both on the operational side, so the admin, the things that you have to do to keep your website up, make changes to your website, and also from the buyer side. Let's start with the backend side, the e-commerce day-to-day experience that your staff has to go through. I talked about Sidekick which is basically an LLM that works across Shopify. It can help you do all the tasks that you do on the admin side. So we talked about it earlier. It can create customer segmentation for you. It can help you with reporting. 00:29:02.000 --> 00:29:20.000 And it really changes your experience because now when you want to create a report, instead of calling out someone who has the technical knowledge in order to be able to create that report for you, you can just describe the report that you want created in sentences, in plain English. And it allows you to create that report. So we're seeing a lot of operational efficiency gained from users using Sidekick. 00:29:20.000 --> 00:29:40.000 Fun fact, we now allow Sidekick to be full screen. So that means you can just use this LLM experience as the main method that you interact with Shopify as the merchant. 00:29:40.000 --> 00:30:14.000 Now on the buyer experience side, obviously there's a lot going on here. A lot of people are using Perplexity. I personally, when I'm trying to buy something, I start with an LLM. I just describe what I'm trying to buy and then the LLM replies back what are the best products that could fulfill this need that I'm describing to it. 00:30:14.000 --> 00:30:42.000 So on the buyer side we are thinking about it both on a macro level, which is like the Perplexities of the world. When buyers are going there and asking questions, where are they getting their information from? We are creating an MCP server, the Shopify Catalog. And because of the scale of the platform, we obviously have information across millions of products across millions of different merchants. We are creating the Shopify Catalog that these LLMs can use directly to call and get the direct information they need for any product that the buyer is asking about. 00:30:42.000 --> 00:31:08.000 So that's on a macro level, on these LLMs that are used across the board such as Perplexity. But then we have also released the Storefront MCP which is on a specific store basis. So I come to the website, I have the typical experience where I can click into different collections, go into the PDP pages, but maybe on top of that I want an AI agent that can help me choose a specific product. We have released a Storefront MCP that allows you to easily create those agent experiences directly on your website. 00:31:08.000 --> 00:31:47.000 For both of these, we allow you to look at the queries that your customers are asking. Whether a customer is on Perplexity asking a specific question about your product or whether they're directly on your website on the agent that you have built, we keep track of the type of questions that people are asking and we allow you to modify your answers so you show up exactly how you want to show up for your customers. Because you obviously don't want to get surprised at the answers that the agent is giving. So we are giving a lot of flexibility to our customers in order to be able to dictate exactly what these agents are saying. 00:31:47.000 --> 00:32:14.000 But again we are early. This space is obviously evolving and Shopify is very much invested in it. But we have covered how we're thinking about it: operational efficiency on the admin side, that's where Sidekick shines, but then also from the buyer experience, both on the massive LLMs such as Perplexity, but also on the specific storefront if you want to have the ability to build your own bot. And we're seeing more and more merchants adopt kind of across the spectrum the things that Shopify provides. 00:32:17.000 --> 00:32:55.000 That's really interesting. So Nate, Paiman's talking about both retailers adopting LLMs and incorporating them into their own website experience as well as working externally with these large GenAI platforms like Perplexity, like ChatGPT. So would love to hear from the Constructor side what we're seeing in terms of retailers that are both incorporating on their own website as well as, you know, I asked Paiman this about social media, I'll ask you here: how have we seen retailers thinking about these external LLMs? Are they friends? Are they enemies? How do we approach this? How are they approaching this? 00:32:57.000 --> 00:33:10.000 Yeah, I've heard both approaches and I think that every retailer ultimately kind of has to decide what their mindset is going to be. If you treat it like an enemy, it will be your enemy. And if you treat it like a friend, I think it'll help you. 00:33:10.000 --> 00:33:36.000 If you frame it as an enemy, some of the things I've heard is like you might say ChatGPT, these LLMs, they're siphoning off discovery moments by answering those questions directly in their platform and they're not sending the traffic to the site. So the consumer is not reaching the retail site anymore. But if you take more of a friendly approach, they can become a bridge and they can actually start sending that traffic back. 00:33:36.000 --> 00:34:00.000 I think Paiman was saying, if retailers are feeding them enriched product data and opening up these new MCP interfaces, they could definitely drive back actually very highly qualified traffic, which I think is the other interesting thing. Usually the traffic coming from the LLM is much more high intent than it might be from Google. So it might be less traffic, but it's much more qualified traffic. 00:34:00.000 --> 00:34:31.000 I don't think these external agents and LLMs are really stealing customers. It's really just another channel like apps or marketplaces once were. And I think Paiman was talking about product feeds. If you're on Shopify's platform, that's great because you're going to be able to plug that directly in. I believe ChatGPT back in April opened up a wait list to send catalog feeds as well more manually. So, that's an option as well if you're not on Shopify. 00:34:31.000 --> 00:34:57.000 {{CTA:See Constructor's AI Shopping Agent|https://constructor.com/solutions/ai-shopping-agent}}But I think when retailers ultimately, this technology isn't fully available yet, but ultimately you're probably going to need to have your own sales agent on your site that can speak to these external agents. So that ensures the outside agents know what's in stock, they know what's trending, they know what the merchant wants to prioritize and basically acts as a salesperson, agent to agent. 00:34:57.000 --> 00:35:22.000 But inside the retailer's own site, LLMs are even more powerful and this is something that you can implement today. Retailers should actually be deploying agents themselves on their site. They can reason about intent in natural language. They can take something we talked about, that query mapping, "comfortable shoes for standing all day" to sneakers for one shopper and insoles for another and Danskos for another. 00:35:22.000 --> 00:35:55.000 So I think if you imagine a scenario where the traffic originates on another LLM, your site still needs to be the best destination and you really don't want there to be a lack of continuity where they have a really great experience on ChatGPT and then they come to your site and it feels like they stepped into an experience that's 10 years old. So I think embrace both strategies. Equip your business to communicate seamlessly with those external agents. The tech is going to be coming for that pretty rapidly. But then at the same time, make your own site agent-powered. Delight that incoming traffic. You can do this today. 00:35:55.000 --> 00:36:08.000 Constructor offers AI shopping agents and AI product insight agents. There are other vendors that do this as well. And I think the leaders will be those who treat agents as extensions of their salesforce, not as competitors. 00:36:13.000 --> 00:36:36.000 Yeah. And I think products like the AI shopping agent have really built that same natural interface. And anyone who wants to see that can follow up with us. We can send examples. But it really is remarkable to see that experience on a retailer's website specific to their catalog and their set of products to have that back and forth very natural conversation. 00:36:36.000 --> 00:37:07.000 Paiman was talking about this as well, where you can really ask just any sort of advice like, "I'm going to start mowing the lawn, I need a good pair of work boots," kind of list all your specifications and get back answers specific to that retailer. So it's a pretty remarkable experience. Well great, so now we have all of this information as a consumer, we are ready to basically check out. We've gotten to the add to cart moment, maybe, and that brings us to our third phase that we included in our survey, which is decide. 00:37:07.000 --> 00:37:31.000 And in the decide phase, you're basically saying, "Okay, I believe I have the right product. I've maybe added it to cart. Help me confirm it's the right choice." Right? We all know about cart abandonment. We all know that just because they've put something in their cart doesn't mean that we're at the end of the road yet. And so this is sort of about that last mile. 00:37:31.000 --> 00:37:58.000 So a couple of statistics we found from the survey around this part. 67%, again well more than half, say that reviews are actually one of the most important trust signals at this point. So that's going to help them decide for sure that this is something that they're going to actually purchase. We also saw some real variation here amongst generations in terms of what's most important in this final phase. 00:37:54.000 --> 00:38:17.000 So boomers are saying those clear return policies that Paiman was talking about earlier as well as secure checkout, those being incredibly important. Gen Z talks about influencers and endorsements as other things that they're looking for as important trust signals. And both millennials and Gen Z talk about media coverage also being important to them. So, interesting to see these different signals they're looking for to decide is this something I'm going to finally go ahead and hand over my wallet for. 00:38:20.000 --> 00:38:50.000 Also 62% cite high shipping costs as a leading reason for cart abandonment. So, that's a big piece here. You see some other barriers as well. Long delivery times, tedious checkout process. These are things that are causing people to bail out. On the flip side, 64% say one-click or accelerated checkout is somewhat or very important. So the more you can reduce that friction, the faster folks are going to go. 00:38:50.000 --> 00:39:05.000 So Paiman, we see customers looking for trust signals at this final phase. I know Shopify has definitely thought a lot about and done a lot about the importance of those trust signals in this phase and would love to hear what you have all learned specifically about this decide phase. 00:39:07.000 --> 00:39:36.000 Yeah. This is again another area where I think all of us can relate. Sometimes you've seen an ad, maybe these days you talk to an LLM, you end up on a website and you're just kind of unconsciously or consciously looking for signals of: do I trust this merchant? Should I buy from them or not? And there's multiple factors that affect that decision. 00:39:30.000 --> 00:39:57.000 But there's a crazy stat around Shop Pay, which I'm sure everyone is familiar with, Shopify's accelerated payment method. We did a study with a third-party consulting firm. They ran a study across hundreds of thousands of stores and hundreds of thousands of buyers. What they found is that Shop Pay increases lower-funnel conversion rate by 5% even if it's not used. 00:39:57.000 --> 00:40:18.000 So that stat alone tells you that just the familiarity of seeing Shop Pay gives the customer that boost in confidence that, okay, I can trust this website. This gives it legitimacy. And because of that, I'll purchase. So this is where Shopify has thought about, okay, Shop is becoming a brand that buyers recognize. How can we use that brand to build even more trust for all the merchants that are on Shopify? 00:40:18.000 --> 00:40:40.000 For example, we've invested in Shop Promise, and Shop Promise, the idea being that showing the expected delivery time using Shopify's brand. So, again, you see Shop Promise, that you're going to get this package within a few days, and then again that builds another trust factor. 00:40:40.000 --> 00:41:01.000 But I think another part that's really important, back to the return policy. Making all the information that the buyer wants easily accessible as they go through the journey is really important. So again, personally for me, a lot of times when I'm hesitant in buying something, I want to figure out what the return policy is. If it's really hard for me to find out what the return policy is, I lose a bit of that trust signal and I might not purchase. 00:41:01.000 --> 00:41:38.000 And again, this is another area where we think it's rapidly going to change. Buyer expectations are going to rapidly change. And I think a great use of AI is just making sure that all this information that the buyer wants to access as they go through the buyer journey, they can access in the simplest manner possible. But yeah, I think that stat alone around Shop Pay, whether the customer uses it or not, you see a lower-funnel conversion rate increase of 5%. That just tells you the importance of trust signals. 00:41:40.000 --> 00:42:24.000 Yeah, that's a fascinating stat. I love that stat. So Nate, Paiman talked a bit about return policies, one of his favorite topics. He also talked about secure checkout. These are some of the trust signals that clearly matter in this phase. But we also saw that statistic about reviews. And even to some degree influencers and endorsements, some consumers even at this phase still looking for some final information to make them feel secure in the purchase they're about to make. Would love to hear from your end kind of what we've seen with retailers that we work with at Constructor in terms of how they help address some of these final assurances or questions that customers have even as they get to this final stage. 00:42:27.000 --> 00:42:54.000 Yeah, sure. So, reviews obviously huge, massive piece of the trust puzzle. And I think a lot of that is derived from the fact that they're user-generated. You're getting just a regular person's opinion on their experience with the product, which is super valuable. 00:42:54.000 --> 00:43:26.000 {{CTA:Explore Constructor's Product Insights Agent|https://constructor.com/solutions/product-insights-agent}}Product detail pages, PDPs, those are ultimately where conversion is won or lost. And so the mechanics that you add there, reviews being a major component, but also the return policies, transparent shipping information, or even something like a real-time Q&A, have a pretty outsized impact. And shoppers ultimately just want the assurance that if they have a last-minute doubt that it can be resolved without leaving that page. 00:43:12.000 --> 00:43:55.000 Paiman talked about earlier, "Do I have to search for a policy page? Do I have to scroll all the way down to the bottom, Ctrl+F?" As much as you can reduce that, I think, makes a huge impact. At Constructor, one of the things that we're seeing play a huge role here is agentic Q&A. So basically what that is, is AI that can answer specific questions on the product detail page about that specific SKU, like sizing. If you're in between a size, do I order the medium or the large? Materials, return rules. Basically effectively addressing any of those doubts that come up that might block someone from converting because they couldn't find that information easily otherwise. 00:43:55.000 --> 00:44:15.000 And by the way, it's also pulling in information from that UGC, from all of those reviews, being able to summarize that to get somebody a quick and clear answer to a specific question that they have. So, I think that turns the PDP into more than just that static catalog entry. It becomes interactive and confidence building. 00:44:15.000 --> 00:44:45.000 Right? And again, it's utilizing LLM, so it's in that natural language interface again where you can literally ask anything and try to get the answers to your questions. I personally have a problem with dogs that sometimes can't hold it till they get outside. And so when I was rug shopping, I found it incredibly useful to specifically find which rugs are easily cleanable specifically for dog urine. I know that's my own problem, but it was amazing to me to see that actually work within a PDP and get that answer, which even a year ago just was not possible. 00:44:49.000 --> 00:44:58.000 Otherwise, you're googling the material and googling cleaning products and it's just so much work. So, trying to put that all together. Yes. 00:44:58.000 --> 00:45:36.000 Paiman, you talked about the pain points we all feel. The other one that absolutely resonated with me when I saw these stats is that 64% say accelerated or one-click checkout is important. And I think it's one of those amazing things where I don't think I ever would have said I absolutely need this before it existed, but once it existed, it drives me crazy when I'm somewhere and they don't have this. It feels like such a pain point. I'd love to hear a little bit. Again, this is obviously an area that Shopify excels around, that checkout. What have you seen around the importance of this one-click? 00:45:38.000 --> 00:46:07.000 I mean, I couldn't agree more with you, Dan. It's exactly one of those experiences where if I don't see an accelerated payment method on the checkout, it's just insane how I feel. I'm like, "Oh my god, what's going to happen?" And then when it's not a single checkout page, you're like, "Okay, I'm going to put in this information. What's going to happen next? What else is it going to ask me?" So, it's actually insane how much buyers have gotten used to accelerated payment methods. 00:46:07.000 --> 00:46:43.000 At Shopify, we have seen that the majority of the checkouts that happen on Shopify now happen through accelerated payment methods, period. So, customers are not going through the experience of entering their credit card information. And across all of Shopify, Shop Pay is obviously the most popular accelerated payment method. And really if you think about it, the customer has gone through this whole journey. They've gone through the define, refine, and decide. And they finally have gone to a place where they want to check out. We at Shopify look at the checkout as a finalizer. So again, they have gone through the journey, they have made the decision. The goal should be to make it as seamless as possible for them to check out and decrease the friction. 00:46:43.000 --> 00:47:00.000 And I think that's exactly why when you start using the accelerated payment method, you can't go back. Now an important point that I want to make: we've talked a lot about loyalty, we've talked a lot about personalization, and some of those personalizations and loyalty elements obviously happen at checkout. 00:47:00.000 --> 00:47:37.000 So at checkout you could incentivize a customer with some loyalty points, with some specific offers that they have. But historically a lot of these accelerated payment methods did not have access to these customizations that you would make at checkout. It would just be you either go through the normal checkout experience where you can see all these configurations and customizations and personalizations, or you go through the accelerated payment method option where none of that shows up. Shop Pay actually allows you to show all those personalizations and customizations, and again we think that's very important in the competitive landscape that we are in. 00:47:37.000 --> 00:48:07.000 So you can really provide that personalized experience even if the user uses the accelerated payment method through Shop Pay. But yeah, it's just really fascinating human behavior, how once you get used to that seamless experience, it's really hard to go back. And we truly are seeing it across all of our customers, across all of the millions of merchants that are on Shopify, that accelerated payment methods are the majority of the way that customers are checking out. 00:48:10.000 --> 00:48:25.000 And for those of us that have kids, I think they're even more dumbfounded that it isn't just everything available all the time immediately. And so that's clearly where things are headed. 00:48:20.000 --> 00:48:49.000 All right, so we've talked the three phases. We are at the Q&A section. I want to again please ask folks to ask away. We've got two folks here that know a lot about both the survey as well as obviously working in these two worlds of Shopify and Constructor and being able to apply context to that. So, would love your questions. I'll keep pulling them up here, but I've got a few already in the queue, so let's jump in. 00:48:49.000 --> 00:49:14.000 So let's see. This first one, Paiman, I think this is more geared towards you. So we just finished talking about the seamless checkout experience or optimizing that checkout experience. What is sort of your playbook? It looks like they're asking for making that checkout experience as seamless as possible. 00:49:17.000 --> 00:49:49.000 Yeah. Again, it just goes back to what we're seeing across the millions of shoppers that are shopping from Shopify stores. We are seeing that accelerated payment methods are becoming the expectation. So the first thing that I would say is just take a look at your checkout and see what accelerated payment methods you provide. One of the things that Shopify, we're kind of proud of the releases that we have done recently, is that you can actually use the Shop Pay experience off of Shopify too. So you don't need to be on Shopify as a platform in order to use Shop Pay anymore. 00:49:49.000 --> 00:50:19.000 And again because of the stats that we are seeing with these accelerated payment methods, it can really be a quick win of improving your conversion rate. So again, it's becoming the expectation. It's completely the expectation that the buyers have. So I would truly look at what accelerated payment methods you have available and, I work at Shopify but the stats show it, Shop Pay is the best converting checkout on the internet. I would also look at Shop Pay specifically. 00:50:21.000 --> 00:51:05.000 {{CTA:Download the State of E-Commerce Report|https://info.constructor.com/state-of-ecommerce-2025}}That's great. Thank you. We had a question about the origin of the report. I think some folks may have missed that at the beginning. And so just wanted to give some context really quick about that. Shopify and Constructor surveyed more than 1,500 consumers across the US and Europe. We did use a third party to do this, an official research company. This wasn't just us doing this. We can always provide more of the details, but it is a third-party research that we've now done for three years. And in fact, as you can see now with the slide, you can access the full report. There's a QR code. Hopefully, that works for all of you. If not, feel free to chat us right here or reach out to us. And we're happy to send you a copy of that as well. 00:51:07.000 --> 00:51:24.000 All right. On to the next question. So we saw that shoppers often have to, Nate, I think you mentioned this, need to reformulate the query, to search more than once. What can retailers do to reduce the likelihood that they have to do that so that they can find what they need? 00:51:32.000 --> 00:51:56.000 Yeah, it's a really good question. I think we probably could spend 45 minutes just talking about this topic alone, but I'll try to keep it short and to the point. I think what it comes down to is search that understands real shopper intent in the moment, which is a lot easier said than done. So I think going beyond keywords is sort of that first initial step so you have results that actually match what the person meant, not just what they typed. 00:51:56.000 --> 00:52:26.000 And then layering context signals in like clickstream data, product attributes, and then natural language semantic models and transformers help ensure that first query gets to the right answer without having to try several different times. So you get fewer dead ends, faster paths to the right product, which boosts conversion and loyalty. 00:52:26.000 --> 00:53:15.000 I would say a good example of this, we have a piece coming out about this in the next month or so about the query "red lipstick." We looked at that query across 40 million different searches to see what people actually bought and it was all over the map in terms of colors, shades. A lot of people didn't even end up buying lipstick. They bought lip gloss, lip balm. And I think it just speaks to the fact that a vague intent like that, which actually kind of seems specific if you think about it, "red lipstick" does seem like a specific intent, but ended up in the purchase of thousands of different SKUs. So, I think that's just a really good example of what this is and drives the need for both AI combined with in-session user clickstream for personalization. 00:53:20.000 --> 00:53:39.000 Yeah, great example. Can't wait for that piece to be published. Another question I think for you, Nate. You mentioned this idea that half of shoppers don't care whether recommendations come from AI or a human, but how do you make sure those AI-driven suggestions are actually relevant or actually trustworthy? 00:53:42.000 --> 00:54:41.000 Yeah, I think they feel most personal when they're adapting in real time to the shopper's actual behavior, meaning what they're clicking on, what they're interacting with, rather than just relying on broad-based segmentation or generic foundational LLM models. So I think retailers can really build a lot of trust by making these suggestions transparent. That's not something I think we've talked about yet on the webinar. And relevant. So basically meaning you present results and maybe it's an AI agent presenting those results, but then it's also including text explaining why it presented those results and what were the things that the user did that made it think that this was a good option for them. So I think when personalization feels useful, transparent, respectful, shoppers aren't really going to care whether it came from a machine or a human because it's actually going to mirror that in-store experience where an associate might have told you, "Hey, I think this is a good fit for you because you said XYZ." 00:54:45.000 --> 00:55:04.000 Excellent. All right. Looks like we have one final question unless anyone wants to throw one more. We've got a few minutes left. Paiman, you mentioned, speaking of agents and assistants, you mentioned Sidekick, the Shopify AI assistant. What are some examples of how retailers are truly actually using Sidekick today? 00:55:06.000 --> 00:55:26.000 I think this is an interesting area. I would actually be curious for all of us to reflect how we're generally using AI day-to-day. So personally myself, I'm not even talking about Sidekick specifically. I try to use AI with anything that I do and I constantly get surprised at its capabilities of what it can help with, and that's exactly the trajectory that Sidekick is following. 00:55:26.000 --> 00:55:58.000 Again, you can have Sidekick full screen. So, instead of the Shopify admin, you can be talking to Sidekick just through text. That gives you a sense of the capabilities that it has. We have seen it to be very helpful in creating reports because that's historically a very complicated, a lot of clicks are needed for you to be able to create the exact report that you want. But now you can just describe the report and Sidekick takes care of that and creates a report for you. 00:55:58.000 --> 00:56:14.000 But beyond that I would heavily encourage anyone who is on Shopify to just try to use Sidekick as a first pass for the task that they want to accomplish. And I think it's one of those things where you constantly get surprised at its capabilities and what it can help with. 00:56:14.000 --> 00:56:47.000 So it can take the action for you. It can help you achieve the goal of the specific task that you want to get after. But beyond that, it can be your companion where you're asking questions. Like, "Hey, I'm thinking about providing a discount. Can you run an analysis to tell me how effective this discount could be?" So it really can be a companion and that was really the vision when we released Sidekick, of your day-to-day tasks that you're just running ideas by it and it can help you achieve some of the tasks too. 00:56:47.000 --> 00:57:05.000 Again it's an area that, I mean the number of times I've mentioned AI in this webinar just tells you how my mind gets blown on a daily basis of what I can do with AI and how many different surface areas it can help you with. So, I would encourage everyone to try that. 00:57:07.000 --> 00:57:24.000 That's excellent and clearly it's what's on our retailers' minds as well. Another question that just came in: how do we know when to interfere manually versus rely on the AI and what is the best data that we can use to help make that decision? 00:57:27.000 --> 00:57:55.000 Yeah, so that's a general question. Nate, feel free to jump in. So this is an area where we're aware. One thing that maybe I should have talked a bit more about is on the Shopify-specific side. I think part of that question obviously refers to the fact that you're worried about buyers asking some questions and the AI replying back with some information that's not truthful or doesn't represent your brand accurately. That's something that Shopify has thought about. 00:57:55.000 --> 00:58:27.000 We have released an app called the Knowledge Base app, Shopify's Knowledge Base app. And what that app does is it creates a report and shows you the queries that your buyers are asking. Again, those queries, whether they're asking that on Perplexity specifically about your brand or if you build an AI agent bot on your website directly, it constantly shows you what type of queries people are asking and what the AI replies back with. And then it gives you the ability to customize the answers for the queries that are coming in. 00:58:27.000 --> 00:58:56.000 So really, we want to empower you to be able to show your brand exactly how you want it to be seen by your buyers rather than the AI deciding how your brand should be shown. But that's a great question. I think depending on the context obviously the answer differs. But the buyer context is obviously one of the most important aspects of this. I don't know, Nate, if you have anything to add. 00:58:59.000 --> 00:59:22.000 {{CTA:Request a Constructor Demo|https://info.constructor.com/ecommerce-search-discovery-demo}}Yeah. I would just love to tack on a couple thoughts to that. So, you asked how do we know when to interfere manually versus depend on the AI? I'm thinking about this in a search and merchandising context. This is really challenging. I don't think that there are a lot of platforms today that offer good data on this. 00:59:22.000 --> 01:00:02.000 There are a couple things that we do at Constructor that are really helpful. So when it comes to AI-created merchandising rules, there's a rule performance dashboard that actually shows you the performance of those rules based on whatever KPI is important to you. So it might be revenue, might be conversions, add to carts, etc. You can actually see when the AI made that rule over whatever period of time, did that actually improve the metric that you care about? 01:00:02.000 --> 01:00:30.000 And then conversely, if you decide to intervene and introduce a merchandising rule, there's also a dashboard for merchandised pages to say, "Hey, you merchandised this page two weeks ago. Have those rules that you applied actually helped performance on the KPI that you care about most or not?" So I think this is something that all search vendors need to get better at, providing that transparency whether it's the AI or the merchant making the rules, just showing is this actually moving you in the direction that you want to go so that way you can make better decisions and decide to remove those rules or move in a different direction. 01:00:32.000 --> 01:01:05.000 {{CTA:Connect With Constructor|https://info.constructor.com/ecommerce-search-discovery-demo}}Excellent. Well, we're always available for further questions. I know Shopify is available for further questions. Feel free to reach out to either of us. We can also put you in contact with the other. Please do download that report. There's a lot more in it than we covered today. We go into a lot more breakouts by generation, by region, etc. As well as just a lot more statistics than what we were able to cover today. Thank you for joining us and Nate and Paiman, thank you for being a part of this panel. I found it really useful. Hopefully our retailers did as well. And please join us for the next one. 01:01:07.000 --> 01:01:08.000 Thanks. 01:01:07.000 --> 01:01:09.000 Thanks everybody.