Hello everyone and welcome to the Hello Retail winter release. Today we will talk about how e-commerce teams can move on from just good enough email marketing to really well-timed and relevant messaging to their customers. I am joined today by Rasmus, our Chief Marketing Officer, and Philip, our Chief Product and Technology Officer, and later on by Julius, our Product Manager. So here's how the agenda looks like. We will start with some challenges that brands face today and of course we want to introduce a new solution to you all called Product Agents. You will get to see the dashboard or the product for the very first time. We will dive into some real results from first beta customers and at the end we will do a live Q&A. So make sure to write all your questions you have in the chat box right here in the webinar room. I assume it's safe to say that all of you have something to do with emails every day. Either it's confirmation emails to your customers or welcome flows or campaigns and various newsletters. So you know how important it is that all those emails are relevant and that your customers actually get what they want. Today shoppers get tens of emails in their inbox every day. So a lot of them are either not opened or unsubscribed. So that's what we want to talk about today is how to avoid that with really good email marketing that doesn't require much work from you. But before we go too much into that I would like Rasmus to tell us a little bit more about the bigger picture and why we're all here today. Sure. Thanks Katrina. So behind all of our solutions at Hello Retail there is really a promise from us to help e-commerce teams run better stores. That may sound very simple but actually there is a lot behind that statement and that promise to you as our customers and to the partners supporting you. What I mean by that is that it's not just about technology. We also come with a lot of supporting services and a lot of best practices from literally decades of e-commerce experience is baked into every technology solution and the sort of the surrounding supporting consulting around it that we provide to our customer success team. And this new product that we are launching now and presenting today has everything to do with that. We want to enable you to run a better online store. What has happened over the last year if we look at data from for example Klaviyo's state of marketing report for 2025 is that the cost of acquisition, the cost of getting new customers has increased dramatically. So we felt there was a problem space for us to really do something about to just see what we can do from a technology standpoint, from a data standpoint and through AI to help online businesses and e-commerce businesses as a whole to sort of deal with that fact that acquisition cost has become so high. Next from the next from the next from the next from the next from the next from the next from the next from the Klaviyo's state of marketing report we're seeing a lot of respondents mentioned that they're not able to do their job efficiently without AI. At the same time there's also a lot of FOMO or fear of missing out in the industry among these people sort of operating the online stores. So we thought about how can we take our enormous data set our AI advantage in Hello Retail and turn that into sort of a pragmatic way to get started or to advance your AI adoption. So what we're presenting today has a lot to do with sort of sort of here's a way for you to get going. Here's a way for you to sort of get to really level up on your AI adoption in a very pragmatic way. With that said, I want to hand it over to you, Philip, to talk more about the problem in more detail and the solution that we've created. Of course. As Rasmus said, we have quite a bit of a data advantage. We have invested heavily in AI here at Hello Retail for the last more than five years, building our own AI model that is around products. And we do think that's kind of a missing mark when we're sending emails today. So many of you, as Katrina also said, are sending transactional emails. It could be post-conversion flows. It could be various other kind of triggers with your current email service provider. But the issue is that it always takes the user into account instead of products. And that is an issue because most of your customers have not purchased enough to make it clever using AI, only using users. With that, I mean, if you have had like tens of purchases by the same user every month, of course, you can derive a pattern that, hey, every Sunday, Rasmus is buying milk online. But most of our customers are not like that. Most of our customers buy once, maybe twice, and they're sporadic. There's no clear data points for when to reach out in a good way again. And this is because we're only using users right now. We believe that we can do this much smarter using products. Let me give you a few examples. If you look here, we have three different products. We have some wax. We have some mascara. And we have some perfume. If someone buys this, we can see from our AI model that it depends on the product when people come back and buy it again. So when we're looking at the wax, hey, it takes about 60 days before people run out and need to buy it again. 85 days for the mascara and 210 days for the perfume. Of course, with the perfume and others, it depends on the size of the bottle and so on and so forth. But the issue here right now is that when you build the flows in your current solution, you typically build it up. It's static with the timing, so it doesn't really convey the message at the right time. And you end up sending the perfume as a reminder too early or maybe the wax too late. You're never hitting the mark, and it just becomes irrelevant, the message you're sending. But if you're sending emails where you really take the products into focus, this becomes much, much easier, and we can do it in a much smarter way. So in reality, when we talk about timing, timing is super important when we send the emails because I don't want to get the emails at the wrong time. It's going to end up being perceived as spam. Timing is really a product problem and much less of a calendar problem. This is also why I'm so happy that today we can introduce you to product agents. Product agents is our way of solving this problem. We're taking, of course, behavioral data still, so user data if we have enough, but also infuse it with our product intelligence. And with that, we can send automated emails so we decide when we're going to send it based on what customer support. We can decide what we want to put into it, so think of products and how we should talk about it. And this is really unique because first we're using products to decide when we want it to send, but we're also uniquely making a message for each individual user. So each user gets a unique email as well. Well, we are not an email service provider, so we're not replacing what you're sending emails with today. But what we are is an add-on you can add to your existing solution in this type. At this point, it's Clayton you can integrate to to get the most out-of-all solution. Okay, let's talk about how it actually works. So Ekaterina here, she might have purchased some mascara from one of our web shops. And what happens is that we automatically identify that the mascara is a replenishment product. It's something we buy again. We can see that it takes about 85 days before she's going to buy it again. And we also know that, okay, this product goes really well with some lip balm. And at the end, we are scheduling an email 85 days out with a message for the customer, or in this case Ekaterina, to buy this product again. And by the way, this lip balm would look good on you as well. A unique message based on the products she's purchased. We are launching with a handful of new agents. One of these agents is our post-conversion agent. It looks very similar to what you might have known already from post-conversion emails. So, hey, if someone buys some jeans, let's push some products later on to get them to buy more, let them come back. What we're doing is a little bit different, though. It's because we can identify when customers' modes like to come back. So, in this example, we can see that jeans typically are bought with T-shirts, but it takes about 10 days before you come back and buy it. And with the more longer-sleeved shirts, it takes about 15 days. So, in this case, if someone would buy those jeans, they would go about 10 days, would automatically schedule an email for the T-shirts. And about 15 days, you would get an email about the longer-sleeved shirts. So, we're really putting this on autopilot based on the products. So, no need to create very complex rules and flows depending on what people have purchased to get the most out of it and to hit the inbox at the right time. Another agent is our replenishment reminders. So, we talked about, hey, a lot of our products we're selling today, you tend to buy them more. Yes, I'm not buying the flat-screen TV more than once, or at least it's going to take a while before I'm going to buy another one. But the USB cable, at least for me, they tend to break all the time. So, you might want to buy this again. The same way we have an agent that can automatically identify if a product is a replenishment product. This also means that, for example, with the mascara, as we talked about, after 85 days, we can send you an automatically reminder that you might be running out. If you're using, in this case, a facial cream, as you can see here, it might be 30 days. So, we automatically go in and detect which products are replenishment products that they tend to buy again and how long a time doesn't take before they typically buy it again and send emails at the right time. Again, without any kind of complex flows you have to set up. All of you know about price drop emails. Price drop emails is typically an email you get that you looked at a product, you didn't buy it, it got lowered in price, and suddenly an email is shootout. And we know it works because all of us really love a good offer, but one of the key issues is that some platforms did not support this for Klaviyo. With our solution, you can still get the benefits of a good price drop email using Klaviyo, even if you're not on one of those mainly supported platforms. Another key issue for price drop emails is typically that they don't work well with replenishment products. Again, if you're selling dog food, and I looked at some dog food, and I bought it, and it drops in price, most price drop emails won't be sent because you bought it. Again, we don't want the TV again, but those replenishment products, well, I might want to buy it ahead of time just because it's an offer. So, what we can do is, because we understand products so well, we can send an email, even if you bought it, if we identify this product being a replenishment product. So, this means that, hey, if we're buying dog or if you're selling dog foods, we can easily make an email to your customers if their favorite dog food drops in price, of course, with an automatic personalized message at the same time. Another key issue that sometimes we've seen is for price drop is that, hey, we're looking at these kind of products, but the products that got lower in price, we didn't look at. So, we never send many emails because they might not have been the most popular products. And we don't want to send a discount to someone who's not interested in it. Again, it's all about relevancy. Using our proprietary product intelligence model, we really understand what are alternatives when it comes to products. So, let's say Rasmus looked at a pair of white sneakers. He didn't buy them, but another pair of similar sneakers goes on sale. Well, we can send Rasmus a personalized email saying, hey, there's some white sneakers you might be in the market for that has dropped in price. Why don't you get them? So, again, we can send more emails, but to still target the right audience with a good personalized message. Last but not least, another agent that's very soon to be out is also a price drop for upsell products. Again, let's say you bought a PlayStation and you dropped the price for a lot of games. Hey, you don't need to reach out to all customers about it because it really depends on what console they're using. So, if I bought the PlayStation and you dropped all your products in price, let's send an email to me if I'm interested in PlayStation games. You could also see this with clothing if we talked about the jeans before. So, if we purchase the jeans and we know a T-shirt goes well with this and they drop in price, hey, let's send an email to the customer that, hey, this shirt will go well with these jeans you just bought and they're on discount. Why don't you buy it? So, again, many different ways for you to really use products together with behavioral intelligence to send better emails that are personalized. And this is just a first handful. Many more agents are in the works and it's just the first. What do you use? Thanks, Philip. It's really exciting to see all of those agents in action. But let's dive a little bit into what the emails that these agents generate actually contain. Each email is uniquely personalized both to the shopper and the product that it's related to. It contains a generated one-to-one subject line, all of the personalized generated headline and body text, CTAs that fit directly to the product that you're referencing. And then also possibilities for showing other products that are related to the one that you're promoting in that email. And all of that automatically generated for you by the product agent and ready to send out through your channels. So, before diving how the product looks in the Hello Retail dashboard, I want to share a little bit about how the setup of product agents work. There are a few things and prerequisites that need to be in place for you to get started, which includes a regular shop onboarding that we're doing with Hello Retail on a day-to-day basis. But from there, we have designed an easy-to-use onboarding for Klaviyo in this case as the first ESP we will be launching this with. Over time, we, of course, we want to expand the reach of it and include also other platforms. So, if you are working with another platform and are interested to implement this, feel free to reach out to us. But let's have a look at the flow. So, to get started, you just start by creating an API key within Klaviyo, set up all the right permissions, and then you're almost ready to get going. Whenever that is done, it's all about setting up what we call the NoFlow flow, which is kind of a funny naming, but let me say why this is actually the NoFlow flow. In Klaviyo, you oftentimes build long, intricate flows. To run product agents, what is needed is just a trigger and an email object that will be responsible for sending the email. That is all you need to set up to send your first email. Of course, there is all kinds of further setups that you can make, but to get started, you just need to do that. Then, of course, product agents also fit into your designs and your email templates that are already existing. In Klaviyo, you just take all of our predefined blocks, just drag them into your template and the right place, and you will be ready to send emails quickly. Within just a few steps, you already have a setup that looks like a fully designed email with all of the product information that is needed. Finally, of course, you also want to communicate on brand with your brand's tone of voice. You can set up the tone of voice just by an easy slider in the onboarding to get started and test out how the agent emails are sounding. Then, also further down the road, adjust the tone of voice to whatever you need and however you want to sound when you're communicating with your customers. With that, you're all set to start sending one-to-one emails to your customers automated with product agents. When you're through that flow, we will already have processed your receipts, so there will already be mails ready to send out to your customers based on the information that we have. Let's have a look at how this looks in the Hello Retail dashboard. So, as soon as you're finished with the onboarding flow, you'll get to the overview page. On the overview page, you will see that there will already be emails scheduled for you to send out. This page gives you the security that there is actually something happening and your agents are working tirelessly. You can see all of the messages that have been scheduled for the future. Here you can see the different messages that have been sent out in the recent 30 days. And have a look at all of the agents that are enabled on your account. If we're looking at the agents page, the agents page is your control surface over the different agents that are active for your shop. You can see the three post-conversion agents and the three price drop agents that we're launching with here. And learn more about each of the agents. See what kind of products, what kind of flows they are a good fit for. Enable, disable them as well as adjust the prompt if it's specifically needed. So this is really the place where you're at the heart and soul of the agents and control how they work and how you're working with them. But of course, you also want to see the results of the agents. That's what the messages page is for. Within the messages page, you can see all of the emails that have been sent out and that are scheduled for the future. You can see what has been scheduled and on what basis it has been scheduled. Let's, for example, have a look at an email that has been sent. For each email, you will be able to see the context of why this email has been sent out and what kind of agent is responsible for it. And you're also able to see what has been generated. So you can control what is the text that this email has been going out with and what kind of subject line has the agent generated. In this way, you can see how the agents work and how they are creating the results that in the end should lead to higher revenue for you and your shop. I hope you liked this little run through of product agents and I hope you're excited about it as I am. And with that, back to the three of you. Thank you, Julius, for the demo. And now I'm really curious to hear from Rasmus about some real results we've gotten from beta customers. Thanks, Katrina. So in developing this new product, we did not want to launch with an unproven or untested solution. We only wanted to launch after extensive testing where we could really prove together with some of our good customers that were willing to run sort of early testing email flows, but on their production online stores. One of those customers is Yarn Living. They're selling yarn and knitting accessories and other handicraft materials. So Yarn Living, they introduced several post-conversion product agents that started sending out emails to their shoppers on their live site. And these post-conversion emails are set up in their Klaviyo account, which means that we can compare the performance numbers very directly to the kinds of emails they were sending ahead of launching product agents on their live store. So they were able to generate up to 600% higher revenue per email using product agents. And this, again, goes to show how some of our beta customers were immediately able to produce really good short -term revenue results from using product agents. So the beta phase is now over, which means that we're now starting to sell the product to all customers. And therefore, we also have to mention pricing, which Phil will talk a little bit about. Thank you so much, Rasmus. And you're completely correct. Like showing the amount of revenue per email really shows about quality. It shows about us sending the right emails at the right time. And this is all about the product. We're using quite a few AI models internally built here at Hello Retail. And of course, they consume a lot of tokens, so to speak. And this costs some money. What we aim for is to make a pricing model that is as flexible as possible and as fair as possible. So it's a consumption-based model where you only charge for how much you're using. You can decide to pick which agents you want to run, how many, how little. And you can start really small with just a monthly payment for a three-month commitment. So really get started. Try it out. See how it works for you. And then you can, of course, grow and more agents as we go. Thank you, Philip. With this, I think we're ready to wrap up today. Not before, of course, we're going to a live Q&A in a minute. But before we close, I wanted to remind you quickly that today we sort of launched to our customers and partners here in the audience product agents. It's available to our customers in the dashboard to check out, where you can also directly request a demo from us if you'd like to. You can also learn more about product agents on our website, where you can get more resources and more information and details about each agent. I must say I think our team is super excited about this launch and what we've been able to create. And from email relevancy to right timing to one-on-one personalization, I think it shows really that the future is that kind of one-on-one outreach that customers want. So thank you so much for joining us today. Thank you, Rasmus. Thank you, Philip. Thank you, Julius. And we'll see you in a second at the live Q&A. Hello again. I hope you can all can see us and hear us. We are live from Copenhagen, ready to answer your questions. I hope you really enjoyed the session. We already have a few questions lined up, actually. So we're going to take them. And then if you have any more, of course, feel free to write them in the chat. Thanks. Let's maybe start with the first question here, where we have been. The question has been how we ensure that message frequency, that we are not sending too many emails in a short time frame to the same user. For that specifically, we have a rule that one shopper can only receive one email per day from product agents. So there, make sure that only the highest value email gets sent. And beyond that, it will be configurable so that you can say how many emails there should be sent in a certain time frame to a user. Yes. Yes. So there also has been the question about how quick it is actually to set up product agents if you're not already a customer. Philippe, maybe a question for you here. Sure. So there's a bit of data onboarding. And, of course, it depends on the platform. But it's usually a small thing. So for many platforms, we have integrations, too. That makes it quite a breeze to get set up. We need product feeds, auto feeds, and a little bit of tracking. And you can get up and running. So really not a big deal. And as Julia also showed, inside Clavio, of course, you need to design kind of your email template. So that is kind of the complexity there is when it comes to the onboarding. Yes. Then maybe staying on the topic of things coming up. Philippe, is seller retail planning on offering some sort of chatbot service in the future? Right now we are focusing on improving the quality here for the first coming agents. There's a lot of agents in the works. There's a lot more when it comes to the analytics part of it. We do have ideas for how we could expand on this outside of just emails. But a lot of it is right now just on the planning side. Julia, if you have any good ideas, reach out to your customer success manager or email one of us. Gladly be in touch about your thoughts and ideas. Great. Great. Then also one of the questions coming up here is if there will be a more detailed demo to follow exactly how to set this up in Clavio. And there's two ways. One is that within the product, once you have bought product agents, there is a very detailed run through that we have been building that guides you with small videos through the whole setup process. And as well, you can find all of the setup steps in our support portal, which we can also, of course, provide you afterwards. It's at support.heloretail.com slash product minus agents. And there you will also find all of the steps. Great. Yes. Yes. Any more questions? Anything, any remarks from our sides here? Do you have something, Rasmus, to add? I just want to add that there is actually a couple of questions more that we can definitely answer. So, for example, it's being asked, will there be a more detailed demo to follow exactly how to set this up in Clavio? I think one of you can answer that easily. Yeah. It will be. It is part of the support documentation and the flow in the platform. Yes. I think we have, yeah, you're right. There is some more in the chat. Yes. Let's take this question. Yeah. And put it on the screen here. Yes. It's the idea to replace the existing flows like post-purchase or run to separate flows. Yeah. So, basically, there are some flows you still want to keep in Clavio or MailChimp or whatever ESP you're using right now. That's definitely welcome flows and stuff like that. There might be cases for some post -conversion flows you still want to keep, but the idea is actually pretty much to be able to enrich it and basically using ours to take over your existing post-conversion flows. Again, it really depends on the business. It depends on what products you're selling and what use cases you want to get solved, whether we will replace all your post-conversion flows, but definitely something we can talk about. Yes. Yes. Well, so, some more here in the chat. So, would this also work with the built-in emails without Clavio? So, at this point, we haven't really set in stone if you want to do this or not. So, right now, again, the focus is our existing kind of agents to be able to run it through ESPs like Clavio. But there is definitely the opportunity to build it out with more email templates and stuff like that. Right now, we're in the valuation phase again. Love all your feedback and just keep it coming also when you're testing it out. Yeah. I think there is something more here, which I think would be nice to talk about. Yes. So, one question here from the audience is if there is an ETA on an integration with MailChimp. So, to answer that short and quick, no, there isn't an ETA, but we have purposely decided that Clavio will be the first platform to integrate with here. There is also the possibility in the meantime already to use webhooks to send a payload through some sort of middleware to other platforms and make sure that to use it there. But we are continuously evaluating what it makes sense to work on next and see where to expand this. Yes. Yes. Any further questions? Yeah. I think that's it. I think we should be through all of the questions then. If there is nothing more from the audience, I guess you feel free to reach out to us if there's any more things that you feel you hadn't had the chance to ask or just that came up afterwards. And we are happy to help. Sure. Sure. And as also Rasmus put right there in the chat, you'll get a follow-up message with the recording and then a few more links so you can learn more about it. Thank you so much for joining us today. Thank you. Thank you. And have a great rest of your day. And we'll see you again in the next one. Yes. Yeah. Cheers. Bye-bye. Bye.