Cool, all right. I will start things off now. Thanks everyone for joining. This is recorded, so if you want to share this with anyone else in your organisation, then please do. It will be sent out after the call today. So we're excited to tell you what's going on at the moment. The product office hours are quarterly, so the objective here is to increase communication and shorten feedback loops. The session will be interactive as possible, and we'll send out polls where we can and a survey at the end, so please do give all the feedback that you can and we'll improve as we go on. So the agenda today, first thing we're going to give you an update on the recent and imminent releases and beaters. Then what's next? So road map stuff, including an update on alerting and some AI improvements that have come out recently. And then a couple of interesting spotlights for you today. So we'll run you through what we've been doing in preparation for peak and also some information on how you can drive success through this peak time as well. So the host today will be myself and Laure on the product side. So I am the principal product manager for product discovery and then Lor is the head of product management for our AI side. Then we've got Ernst who runs our operations and support and Andreas who runs our experience consulting. So everyone here is kind of a subject matter expert in different bits and we'll run you through all of those. So come releases and beta updates to begin with. Let's have a look. If you have any questions about how you want to adopt these, please reach out to your CSM that you'll get some information today, but there'll be more coming as well. So the first one is the Java upgrade. So this is actually out at the moment and pending rollout. And so if you do want to get involved with this, then please do reach out to your CSMs and we can factor that into the rollout process. The goal here is twofold. So we wanted to make sure that we are always improving on performance. So across the board, we've seen about a 15% improvement on query response times. So that's quicker page load basically for you guys, which obviously has knock on effects to many KPIs that we all know about. So that's great news. Additionally, there are multiple security updates in him. So we need to make sure obviously that we're staying on top of those, but almost more importantly than that, that you guys are getting them in a timely manner. So that's why at the bottom there I've mentioned that we are also changing the way that we roll out our Fred Hopper upgrades into more of a scheduled manner. So this will be the first time that we're doing this, pushing from our side and scheduling in upgrades rather than requested by the customer. And this will be the way moving forward because then you can take advantage of all new features, the security updates, performance updates. So you'll see those timelines basically getting shorter and shorter. The first one will have huge manual oversight and then over time we'll add more and more automation so that you guys get these upgrades as quickly as possible. So that's the main part of the current Fredhopper upgrade. But we did also include some other bits in this release as well. So there's now a safeguard to stop redirects going live without a trigger. So that can be have a huge effect on your website. It would technically be a configuration error, I suppose, but we've blocked that from even being possible. And then multiple efficiency items as well. So Fred Hopper is now linked to XO. So our advanced recommendation solution. If you do use both, you can now fit between the two at the click of a button rather than logging into two separate things. There's also a scroll bar and preview pages. You can switch between locales easier and you can search within that. So that should make previewing your content much quicker. And preview page load times have improved as well. We've minified a lot of code in there so that we've optimized those loading times. So overall, you guys should be able to preview things much more quickly. And then relational search as well has had an improvement. A couple of our beauty customers were approaching a limit. Typically, beauty customers have a lot of variants for a specific product. So relational search is where, for example, if I type in red dress within the search, it would show the red item rather than the blue variant or whatever. And that now is unlimited. So we have no limitation then. And there's obviously always with these things, multiple fixes and improvements that we've released with this. If you are waiting on anything, please reach out to your CSM. We can make sure that your improvement is actually included in this release. Then an XO deliverable. So the account portal is actually already in beta for onboarding tooling and checklists that any new customers can implement the tool much more easily. But the goal here with the second part is to actually enable you to know exactly what's going on at all times and take corrective action with regards to your activity tracking. So if your views, ads, purchases, things like that within our tool, if you do use our activity tracking, then this would tell you if suddenly there are changes that you've made to your website, has knocked that out and you can go and resolve that. So this was released last month, but it's still included in that beta. So please do reach out to CSM if you want to be included in that. This is XO for now, but as I will go on to show you in the roadmap portion, this is moving into a thread hopper tool in the future as well. Then if I hand over to Laura. Thanks Tom and hi everyone. So today I'm going to talk about the latest updates from the AI team and that's going to be about the visually similar recommendation and AI search improvements. So something to bear in mind is that whenever we deliver, we push production a new model, a new algorithm, we don't just leave it there. We keep improving those models to include the latest technological advancements because everything is going very fast and we want to make sure that we provide you the best results, the most relevant results, whether it's for recommendations or for the search. So here one of the foundations of visually similar recommendations is to create a vector representation or if you wish, a mathematical representation of images so that we can then compare them with each other and provide with relevant recommendations. Here the team has been working on leveraging an even more powerful model to do so, which has been trained on millions of data, which makes it even more powerful. So the impact, the positive impact on visually similar recommendations is straightforward, but it also impacts AI search because AI search is based on queries data, analytics data, but also on images and products catalog. So you can see, for example, on the right, I will let you read that, but you can see some AI metric improvements for one of our fashion customers. So for example, for those who are familiar with that, you can see the hit rate that's 100% which has increased by plus 14%. Now I want to show you an example of what the visually similar recommendations look like on one of our customers. I'm going to share my screen. Here we go. So you should see very soon, here we go, the website of one of our fashion customers, which is based in Australia and which is mainly for women's fashion. I put like four random items, mostly dresses and tops because that's mostly what you can see on the website. Bear in mind that the customer has added some metronizing rules as well, so that might impact the results, but let's see how it looks like. So here you have a dress, like a black and white dress with these kind of flowerish patterns, if I may say, with short sleeves and kind of white at the bottom. And if you look at the recommendations here, you can see very similar dresses with these three with the same shape and the same patterns, but then in the catalogue we run out of these kind of dresses, so the model tries to detect all the things in the catalogue that are not dresses anymore because they are not more, but in other categories with this kind of same patterns. So you can see here the same prints and here as well the same prints. Another example here would be with this pale yellow top, kind of summerish as well with short sleeves. And here you can see, it's not obvious on the screen, but it's also like a light pale yellow with short sleeves and very similar style all over. Third one, here you have a denim shirt, long sleeves and you can see here recommendations of similar denim shirts with the long sleeves. And then there's no more on the catalogue to propose, so the model detects similar-ish fashion pieces, whether it's the same fabrics or the same color and the same shape. So here we keep on going with denim, then we keep on going with tops, and then we take another category but keep the denim fabrics. Okay, our last one for the road. That's a mini dress, black as you can see, with like this kind of formal style, short sleeves and then you see the recommendations. So there's no other items in the catalogue with this kind of black and white style, so the model tries to find the ones that are the most similar in the catalogue. So you can see this steel formal style, mini dresses and in black. So that's a few examples of what you can obtain. Of course, there's the influence of mesh lighting rules here so that you can have some sort of control over it. Thanks very much. Now stay on webcam, Lor, because we've got you back on, I think, in one sec, but let me just check that I'm sharing the right thing. Yep, so we've got three distinct roadmaps here because there's three areas of focus that we need to make sure that we're keeping focus on. One on the hopper side, one on the XO side, the advanced recommendations, and another on AI. So we'll start with AI as you are on webcam already. Okay, so me again. So yeah, if you can go on the next, then we can. Perfect. So if you look at the first column, so the things we're working on at the moment, the first item you will see is LLM Power Search. So it goes back to what I've just said before. We keep improving our models until we can deliver the most updated models in terms of relevancy. So here we're using the latest technologies that now everyone kind of knows about is LLM Large Language Model. So we're changing, we're testing new architecture based on LLMs for our search AI search model. Why? Because it will increase very likely the relevancy even more. It will make the model understand even better the context. And the third reason is that we're preparing as well the foundations for what's going to come next, which is conversational experience that you can see in the next column. The second item you can see is also related to AI Search and Vision Similar recommendations. Here is really about the, let's say the switch on. We want for you to be, to make it easier for you, sorry, to have a faster and simplified switching of those models. So we're working on internal toolings, internal processes to make it even better than it is now. I'm not going to spend too much time on the other columns. You can see what it is next visual search. You can search with an image a bit like what you could see on Google Lens, but apply to your website. In the future, you can see AI data enrichment. So it's really to increase, to have a gain in productivity. Instead of manually tagging, creating attributes, you could have this automation in data enrichment. For the rest, I will leave you to read the slide and conscious of time, I will give you back the slides. Tom. Thank you, Laura. So I'll move on to the Fred Hopper part. So there's a lot on this slide to run through. The bold things have their own independent slides, so I will skip over them while on this slide. And the future column, just because of time, we won't be running through in detail today. There is a poll at the end, though, so if you are interested to talk to us about any of those topics, then please do hang on and just click on that poll and say, yep, these three for me, please. Because we'd love to talk to you, get your thoughts. Obviously, we've got a plan on most of these, but we need to make sure that that fits with all of your guys' plans as well. All right. So when it comes to this, we've got the Java upgrade, which I've run through. That one is out and pending rollout. Linking XO to Fred Hopper is a big one as well, which I kind of touched upon. So at the clicker button, we'll be able to switch between those two. The insights, UI and visualization update, I'll run through on the next slide, but that's what I'm really excited about, all about making sure that you guys are supported, so you know exactly what's driving success and what opportunities you have remaining. Then the bottom two here, the GA4 one and the provisioning upgrade. So that's because Google and Amazon made a change. We then had to accommodate that because we get a lot of data from Google, and Amazon is where we store basically where all of our infrastructure is served from. So the provisioning tool that we used to use needed to be replaced. They're two relatively sizable deliverables that we had to do that we were forced to do. Luckily, those are over now, so we can go full steam ahead on that middle column, which is the next bit. So the tagging, again, we've got a slide coming up on that, so that's two slides away, but that is a huge one to make sure that you guys have all of the data that you need to use the tool in the best way possible. Automatic version upgrades, again, already spoken about this one, but just to make sure you're getting everything in a timely manner. The block facet values one, this was a huge one requested by customers. So we do listen to you guys and make sure that we're prioritizing things that are most relevant to you, but this would enable you to on one page, block a certain facet value and still show on another page or block it site wide. So lots of customers are struggling with this within their PIM to do that, or maybe as I say, they might want it visible on half the website, but not on the other half. So you can do it with specific values within that so you don't show what you don't want to on certain pages. Then there's customer alerting, items and activities. So this is based on an XO deliverable that I will be running through as well. So while this is starting on XO, we are combining both of our features. So you'll see a running theme of more and more of our items being delivered on both sides. Then live preview to match pre-published. So there's a few features on our live preview that on our pre-published preview that aren't available on live preview at the moment. And we're going to be changing that so you can basically see what result modifications and where those result modifications are, start to hide them from certain pages. So you can troubleshoot all of the things that you're used to on your pre-published preview. You'll now see on live preview. So if you have an issue with your live website, you can go on there and check if it is actually a problem on your side or if it's something that we can help with. Then insights internationalization and filtering update. So this is at the moment you would have different insights dashboards for different websites. Moving forward, that will change. So we can see a global view of all of your websites on one page. And you want to see if one ranking is performing well worldwide. But then the filtering part is then you can still break that down per country. But in addition to that, we're building the filtering in such a way that we can filter on anything that we're fed. So it could be stuff from your front end. So if you can pass us, for example, device type, or if this is a premium customer or this is a logged in customer, whatever that is, and then we can start to use that to filter the reports. You can see how those different customers behave on your site. Additionally, we could break down per category, per search path. There's lots of things that we can do to split up those reports. So while this had an initial focus on internationalization, the filtering element we're building in an open way so we can use anything from any place to split those reports. You can slice and dice and make sure that you get the best understanding possible and also make sure that you can actually action those changes. You know exactly where that money's coming from or where it's not coming from anymore. And then on the right there, I did say I was going to skip over these, but yeah, there's five really interesting topics there that I'd love to talk to you guys about. So please do stick around for that survey at the end. So the insights visualization update, I promised another slide on this. The idea here is making sure that we're surfacing information in the way that merchandisers can use to best improve your website. So making sure it's as actionable as possible. So I'm looking forward to demonstrating this one next time because this one's really close now. I've seen an early demo and there's only a couple of bits that we're still waiting on. So as I said, this one's for Fred Hopper and within Fred Hopper you can make hundreds, probably thousands of changes and config tweaks. This is to make sure that you can report on the successes from those, identify trends either before or after you've made those changes, and spot new opportunities to make sure that you can get additional success from the tool. The idea here is to make sure that they're easily attainable. So some of the reports today, while the information's there, it can be quite clunky to get the information out. So we plan on overcoming that and visualizations here at Key. So on the right, while there'll be trend graphs on all of our pages so that you can see any line item over time, this one here is the slightly more advanced one for A-B testing. This is a little preview of what that will look like. So for example, if you have variant A and that beats variant B, until now, you'd say, okay, A is the winner and you would make whatever that percentage was, more click throughs or more revenue or whatever. But it could be that before payday A was winning and after payday B was winning, but you didn't have that visibility before. Now that we can see over time and that would potentially give you another hypothesis to go test and go, okay, why was that? Is it because we're showing a certain set of items that are more popular when people have got a bit more cash in their pocket? Whatever it would be. That's potentially a silly example, but you can see that that would change. It could be that on weekends, A was winning and B was losing. Whatever those trends are, you'll never be able to see them at a glance and then take the next action. Also within this, we'll be looking as a next step into the internationalization part and how we will best present that to you guys. We've got rule-based and semantic and visual tagging. So the rule-based tagging, this is actually already in use in our XO tool. So we'll be making this general availability for the Fred Hopper customers. And this allows for a few things. You can do data mapping. So if you've got 10 different types of green and you just want green for your facet, fine, we can handle that. There's also rule-based changes to tags. So if you're over a certain amount of sales, market is a best seller or a new in, or even if it says within the description, this is a party product or a summer product, you could extract that information from the description, put that in as an attribute and wait that in search without having to wait your whole description because that can sometimes lead to a bit of noise in your search. And finally, the mass tag assignment as well in here. So you can just tag any list of products with any tag. So for example, if it's Black Friday and you want to tag a load of products for that, or if there's a new edit page that you want to make for a short amount of time or for a long amount of time, you can tag those and then we can group those products all together really easily. Then on the AI side, there's a few different ways of doing this. It can be trained with your data and our models. It could be trained completely by us and we just roll out new tags to you guys. Or it could be via search enrichment. So these all look at your images, your product data, your searches, and the navigation on the site to come up with the best tags possible. So the idea here is that you could homogenize your catalogue. So if you've got varying tags across the site, you want to make sure that they're more consistent because they're from different suppliers or whatever the problem is, we can help with that. Or amplifying your data. So if you've got 20% of your products tagged with a certain attribute, we could roll it out to the extra 80% using these models. Or just generate new tags based on your search data. So it really captures the new trends of how shoppers describe your products and identifies what they're looking for. I think I said a little early there, but I was ready to move on to this. So we've got a few things in here. The account portal and the monitoring have got their own slides. So I'll run through those two already in the now column. There's also some monitoring and alerting on the right that I'll run through in just a sec. But the linking to Fred Hopper, obviously that's on both sides because we need to make sure that we can do that both ways. The feed processing scalability improvement, this is to make sure that we have, we can basically batch up all of your full feeds together so that they're passed through our pipeline as quickly as possible. We have a streaming API so that throughout the day we can receive updates and present them back to you almost instantaneously. But for the full feeds, we needed to make sure that we'd optimize for those because you're sending so much data at once. So this is that to make sure that those full feeds are processing as quickly as possible. So when you send them at night, you process them straight away. And if you ever have to send them throughout the day, you're not waiting for that for too long. And this is the new version of the recommendation API. So this is a unified data pipeline for product discovery. And the idea here is that this will help with easier onboarding and again, to help with that data processing speed. Then if I skip over to the right column, we've got improved internationalization capabilities. So this is something a bit inspired by the Fred Hopper side. This is huge strength on that side that we're hoping to bring over there so that you can publish things to multiple countries all at once in the easiest manner possible. The UX improvement for advanced widget configuration. This is basically just to make sure that when you are deploying new widgets, new configuration, that you have an improved view to make sure that merchandise have the clearest picture possible. So you understand exactly what you're configuring. And then again, I'll skip the right hand side, but we're more than happy to talk to you guys about that and would love to in the future. Then we've got monitoring alerting. So this is the future bit. So we've already got the onboarding help and the monitoring alerting on activities. Next up, we've got monitoring and alerting on your personalization strategies. So within the actual front end stuff as well. And then following that, we will have it on monitoring alerting on your items feed as well. So if you send a full feed or a partial feed throughout the day, whatever you're sending, and there's an issue with that, we can flag that with you guys straight away. You'll get an alert and you can take the corrective action. You can go in and you'll see exactly what's wrong. It'll say what you need to do to remedy it. And we'll go from there. And in addition to that, we would be rolling this out over the coming year onto Fred Hopper customers as well, so that everyone gets this value. And then a quick plug for our roadmap portal. The idea here is that this is to help with our global product team with transparency and feedback avenues from you guys as well. So the link is in our support portal or CSMs can provide this to you. This is all based on product boards, which is basically a powerful tool to manage the product lifecycle. And we're going to use it to present our plans, gather feedback from you guys, and enable you to suggest brand new feature requests straight to us. So it's hugely helpful for transparency, but please take a look and please keep engaging with that. So far it's been great. Now I will hand over to Ernst, who heads up our support and operations team for an update on what we've been up to with regards to peak. Thank you very much, Tom. So as we're heading into the final quarter of 2024, peak is once upon us. And what we wanted to take this opportunity with is just talk a little bit about all the things that we've been doing and will continue to do going into this period to really make sure that we are as resilient as possible for this very critical period for our customers. So the first slide we have here is all about the key dates. And we want to share a couple of the things we're doing slightly differently this year. So the main date that we're working towards is that final period in November where Black Friday is situated. We do understand that for most of our customers, peak starts when the Halloween sales kick off. And so what I'm going to share today carries over across all of that. The first thing we've done slightly differently this year with new running operations is we've implemented what we're referring to as a soft freeze that started on October the 1st and will carry on until about January the 6th. This does not impact our customers' ability to make changes to the site. This is a self-imposed freeze on major infrastructure adjustments or rollouts and things like that. So some of you might know that during the summer we did a lot of work with regards to modernization of our infrastructure, rolling out some changes based off the OpsWorks end of life from AWS and things like that. And what we wanted to avoid going into this period is throwing up any changes too close to the Black Friday core peak period that could result in some instability. So this is a self-imposed period where we will not be doing any big infrastructure related changes and we are being very cautious about what we allow to go through. We'll be working with the various teams internally to make sure that we risk assess anything that we do need to push out that's maybe at a lower level. So this could even involve some work from the professional services team on data and things like that. Really trying to make sure that we ensure stability going into this period. Then on the 18th of November, so a couple weeks before Black Friday, our hard freeze kicks off. The hard freeze is where we will not be making any changes and we will encourage our customers to not do the same. Hard freeze changes that we will not do is for example fast version upgrades, go lives or enablement unless pre-arranged with the professional services team and risk assessed accordingly. We'll also not be going in unplanned re-indexes. So this is where we might do a re-index for a customer. We won't be doing things like that or deploying new e-commerce foundations. Our customers are obviously still free to continue doing their work with the merchandising studio behind the scenes, making changes to their cathodes and that sort of stuff as per usual. We are not blocking that. This is mostly from the trade-off problem and exo science. We then have obviously Black Friday week that 28th to the 1st of December is the core weekend and that entire week we'll have some additional capacity around the various core teams on our side. We have a short break between the two freeze periods. So after the 9th of December to the 16th, we will also allow those hard freeze changes to take place. Again, risk assessed before we allow them to pass through and then we'll go into our second freeze of the period of the year, which is the December 16th to January 6th where those same changes will apply. Just to make sure that going over the Christmas period when many organizations will be in reduced capacity and there'll be many holidays with Christmas and those sort of things coming up that we don't end up doing something or pushing something like that could cause frustrations for our customers. Our customer support team, as I'm sure you all know, provide 24 by 7 emergency support. So any trade-affected instances and things like that, you can contact the support team. That is not going to change. Over the course of this year, we have been upskilling some of our colleagues in the U.S. to provide assistance in the evening hours UK and EU time. While they are more junior, they will act as sort of a triaging point and address more lower priority issues in as close to real time as we can. Should something urgent come in, getting the right people in very quickly to make sure we can address that just like we've always done. I've left the core office hours on there for anybody that's curious. So anything that's affecting a test site or is low priority, those will be the hours that we focus on those. But your emergency and critical things during this period and throughout the year, contact the support team through the normal channel and they will jump back as quickly as possible. So what I also wanted to take an opportunity is just to re-tap with our customers some of the key processes that's really important going into this period. And at the top that list is capacity planning. So this is where you will let us know what you're planning for peak, how much uplifts in your normal traffic you're expecting, any sort of changes to the aim that we should be aware of. And the reason we do this is very simple. One, while we do have dynamic scaling in place on our infrastructure that will try and maintain operational stability throughout, when you hit a successive sale period where there's a rapid increase, so pushing something out through social media or something like that that's very quickly going to draw a lot of traffic onto the site, that rapid increase the system will struggle to deal with. And so we ask to prepare for that as best as possible by scaling the environment around those periods so that it's ready and waiting for that extra traffic to come in. We will also obviously be scaling in shared infrastructure and overall architecture on our side, but knowing when the customers are going into different sales and the level of that sale allows us to sort of plan around that and making sure that for those critical we have all the right people available should something go wrong. Load testing is another one that is really important going into this period. Now this is not a required process, but more something that we recommend to customers, especially on the bigger side or who have some confidence or lack some confidence that the infrastructure will be able to handle the requests for peak. Get in touch with your customer success manager, talk to them about setting up a load test, which we can arrange for you where you can send through the traffic that you're expecting from the peak period and we can scale the way we would scale normally and see how that handles it. If we do detect any issues we can prepare for that before we hit the main event as it were. The next one here that's really important as many of you know when we do encounter an issue we will try and alert you as a customer practically notifying you that we've detected something and we're taking action or in some cases detecting something and we actually need you to take action. For this we require an emergency contact and we encourage you to engage with your CSM on your regular meetings with them and make sure that they have the most up-to-date list. We know that sometimes there is a start turnaround within businesses and sometimes people leave and they are the ones that would have gotten these alerts. So yeah please get in touch with your CSM, make sure that they have the most up-to-date list of your emergency contacts so that when we log a ticket at three o'clock in the morning to notify you that we have to take some action the right people on your site are going to get notified and are aware of that and it doesn't create undue panic within your organization. The last thing that we didn't have so much last year so for those of you that are aware or hopefully most of you are already subscribers we do have a threadhopper and an EXO status page which you can access without needing to log into anything and that will show any up-to-date maintenance and will also show any incidents that we have. This is a human control page which means we will put something up on there it's not an automated action and it means that all the information that's on there will be real-time provided by an incident manager like myself or some of my other senior staff as the issue develops. We really encourage you to subscribe to those pages either for an email or even RSS feed things like that if you prefer and to get those real-time notifications as we detect incidents. While you can always log us a ticket in many cases we will be telling our customers about this and then you can get the information faster than waiting for the ticket to get updated because this will go up en masse. So I did talk in the beginning a little bit about some of the stuff that we've been doing behind the scenes to make the product and the infrastructure more prepared and reliable. A lot of what we've done with the cloud infrastructure is to focus on monitoring and alerting and I know that many of our customers have shared criticism in the past that our monitoring alerting does not feel up scratch and that was one of my big focuses coming to the operations team this year is to do a full audit and modernization of the monitoring solution. We have now completed that we have a brand new tool in place and that is much more modern than the old infrastructure that we use to monitor our site and the operations team now have full control over this meaning that we can add and remove alerts as we need and we're able to be a lot more reactive without needing to wait for it to be developed and having longer turnaround times. Provisioning so actually scaling up environments making sure that we have the right capacity in place in different places is also much easier now and can be done in some cases through the support team so when you log into the support team for some capacity planning they will be able to provision that themselves without needing to engage with the operations team and from the operations team we're able to to monitor those and to adjust in real time as we detect issues. There's a lot of automation in place and I've already mentioned this sort of dynamic capacity that's in place in a lot of areas where it will try and maintain operation efficiency by adding or removing instances as the traffic ebb and flows system access for us behind the scenes to make it so that we don't need to manage so many passwords there's a lot of SSO involved now and just storing things like that. We've also decreased the complexity meaning that the operations team are able to do a lot more independently without needing to go to the more senior engineers that might not always be available at the in the wee hours of the morning. It also means that we can turn that around since that team is geared around 24 by 7 availability just like our support team. We've made improvements to security and extended our disaster recovery and automation processes as well which hopefully we won't need going into this period but it's always good to know that we have them just in that worst case scenario. And I think that was the end of the the peak bit at least my bit but if you do have any questions or you have any concerns or you'd like to have a larger conversation about peak reach out to your customer success manager and they can set one up most of these I joined with customer success manager just to help them talk through some additional topics so feel free to reach out to your CSM and we can set some of those up for you. Cool thanks very much Ernst and yeah just to reiterate please do sign up for that status page just if worst comes to worst we're covered then and so that is one action for you guys. Right let's take a look at what you can be doing. So I am merely introducing this topic we have already provided this if you don't have it you can follow this link here or you can reach out to your CSM and they can provide it to you but some of these may be a little obvious to you it depends depends who you are but we wanted to make sure there's a full list here so that you can make sure that you've ticked everything off and go into peak with confidence even if some of this is just hygiene and some of it you do you do regularly. So as I said I'm just introing it please have a look through this even take a screenshot if that's easier. We've got one of our top SMEs subject matter experts on the call today to run through in detail some of the higher value points on this list so over to you Andreas who's the head of our experience consulting department. Thanks Tom yeah I just wanted to use the opportunity to just introduce a few of the things we've seen that that worked for other retailers in the past in preparation for peak. So I just wanted to pick yeah really three topics so one is around the visual merchandising and the rankings so how you can prepare that related to that perhaps how you can play with the filters on your side to again make the navigation a bit more well intuitive or possibly even exciting for you if you go with creative on the naming and then later on we'll cover a small bit around surge that isn't all that needs to be done there usually and also a little bit around AV testing but ultimately the main bit and you can probably tell that from your from your own experience is around the visual merchandising so in incredible terms that would be the rankings, ranking cocktails and the result modifications and I suppose the this is quite an obvious and I have to admit that but the probably the biggest help in any kind of peak preparation is scheduling so I think that's that's something if you haven't done this in the past I'd really recommend preparing rules up front and just then configuring them so that they fire at a certain time especially if you've got some some campaigns some promotions going live at well early morning you don't want to be awake at night so rather just schedule your rule set it up so that it fires let's say Sunday morning at six in the morning six a.m. and then you don't have to worry about having to do this across all the different categories that you might want to influence the next bit and yeah I guess I've learned it from from experience is to just double check some of the the rules that you might have in place for regular trading periods so if there are some restrictions for example on markdown products that you might have to lift when you when you go into a promotional period just double check that they're all set up correctly and that they're all being lifted in the right time because what you don't want to have happen is that let's say if if you do a blanket markdown of most of your products in a certain category but then in your regular trading period perhaps you exclude or you you demote anything that's marked down you don't want to have those two interfere with each other so again just double check that they're all set up right and then when it comes to to really configuring your ranking strategy for a promotional or just the peak period in general there's a few tricks I tend to use a lot and they're mainly around what what basically what data points to look at to decide what products should come first there's a lot more to be said around this that there's enough time for entire workshops around this but just the I would say the two key points here are in either the first day or the first few days of a campaign promotion you probably won't have the the right level of sales performance historic sales performance to look down or look back to and that's simply because if let's say you've had these products in the past but they they were full price and now they're 50% off it might not be that interesting to to know what they have sold that before just because now the the prices and with that the demand will be completely different so if you need something that's that's a lot more short term just having a look at product detail page views in the ranking cocktail tends to be a good idea so that still allows you to look at much shorter time frames and be much more reactive to trends then the the second bit and admittedly that's again quite obvious but still nevertheless often overlooked when you go into the later phases of a promotion our chances are that your your stock will be fragmented so perhaps there aren't all that many sizes left that you had initially so at that point shifting your focus to to promoting those products that are actually still available tends to be a really good idea and especially in that period now after black friday when you're when you're basically moving from the the black friday madness onto the the christmas gifting season what you don't want to have is that somebody's especially browsing for gifts only then to find out that well all of the sizes they were interested in had already sold out during black friday so that's a good idea too to just put extra emphasis on the the availability of core sizes and then another trick that especially during the like the heavy markdown periods you might still be able to to get a little bit more extra margin out of your products if you use result modifications to do not just as you show a breadth of categories but also the the breadth of different price points so that's why you could enforce a mix of let's say some marked down products some that are perhaps not as heavily marked down or perhaps in full price so just to help you there with with the management of margins and i suppose that's that's the the key things around the the rankings and ranking cocktails and result modifications if we now go to the the other part of the product list of pages essentially the the filters of facets there is a few things you you might not have done in the past that that are worth trying so when you do go into especially from a trading period or regular trading period sorry to a the peak period in the sense of markdowns or gifting the the needs from customers for for filters might change quite drastically so for for black friday there there might be a demand from customers to to see especially heavily marked down products now if you want to do to do that and promote that filter is up to you of course i know that there's certainly branding and again margin management guidelines around it but if it was only up to the customers then perhaps having a filter for discounts somewhere high up in the list might be a good idea or just a like a toggle only show me markdown things or when it then comes a little bit later to do to the gifting season you might want to you can have the for example the price filter with some brackets that are typical gifting thresholds so let's say something less than 10 or 100 pounds or whatever type of product you're selling so perhaps adjusting the the brackets that you use there to to match what what people typically spend on a present can really help there or if you maybe if your filters are not that exciting from the the naming side of things perhaps you could just rename the values so instead of just saying only show markdown products you could for example call it only show me the black friday offers so that's even if you can't change the background you might be able to simply change in your federal facet configuration near there's a lot more that can be done with the filters but i appreciate that's perhaps a little bit more specific so if you do have any questions around that again reach out to your customer success manager they they can probably advise or get you in touch with the right person so let's go to the next page so the um yeah the the remaining two bits that i tend to prepare upfront before black friday it's be just a bit of a check on search so not so much the like a change of configurations but really just a review that we're all set for the like the top trending terms in the peak season and the the best idea there is to really look back so don't just look at the the searches that are currently trending but rather look at the searches that were trending last year otherwise it's just too easy to miss all of the let's say the christmas themed searches and um yeah you want to look at let's say the top 100 top 200 whatever um your team can manage uh to just see what the the results look like uh perhaps already thinking a little bit ahead into what product data will you actually get in perhaps any any christmas products that haven't launched yet so for example if you if you know there will be a thousand more products that are um in let's say starting november they're going to be categorized into a christmas department or christmas category you can usually already predict what the search results for those terms then will be like or if there's any sort of side effects that you might have to be aware of so just to give you an example let's say if if um the let's say that the entire christmas department will be filled not just with um say christmas trees and christmas decorations but perhaps with all kinds of random stocking fillers you know like the candles and so on you might just want to double check that the the christmas trees and the decorations and so on they will really be promoted higher than um the let's say they did more presents and gifts so that's that's worth double checking um and for that you you will need to know what what people are looking for um so for that check last year's and how your product data will be structured so usually um if you don't know from last year perhaps it's worth reaching out to to your buying team or to your data specialists and if something doesn't quite look right the usual mechanism around it is to add synonyms if there's something missing so perhaps link the search for christmas to let's say baubles or or any of the other terms that might be related to your particular business or use result modifications to to block or demode certain categories if they're not all that relevant and similar for the rest of the site um if you prepare for for peak it's usually a good idea to just look back into the like what happened last year um again it's quite obvious that it's yeah i sometimes make the mistake and only look at the the current trends instead of looking at at last year's trends um similarly if you've got any ab tests that you've run during the year it might not be that applicable to the to the peak season so that's it's worth double checking that the the conclusions you've drawn from from tests are really valid for the for the peak season um which i guess also means in turn that it's really it's worthwhile setting up a few day b tests really just specific to the peak period um and technically i suppose that's two of them so there's the like the heavily promoted area around um like friday and then perhaps you can learn from that uh how shoppers react in in this highly competitive highly mocked down landscape and then um run some other tests how people who are perhaps more on the like the gift shopping side how they react for example to to merchandising strategies so try to run a few tests there um with a few caveats there of course that's uh yeah if you're unsure about how your system so how your shop and how the integration to fredber will be able to handle ab tests it's always worth double checking with your technical teams or perhaps also with our account managers to see that there isn't any impact on the the technical side of things but yeah that should give you the chance to really learn from the peak season and be even better prepared next year yeah um if you've got any more questions then again do reach out to your customer success managers or to myself thank you very much andres um so the final bit is is we're going to stop talking now um don't worry it's it's pretty much over but we just wanted to send out one poll for this so there are multiple topics that we are investigating right now obviously we know all of these are important for different reasons but if you have strong opinions or if you have um if you know that you want to drive in in any of these directions then please flag uh within the poll which ones you're the most interested in um and that's one second i will launch this poll now um so you guys should be seeing that now um the poll and you should be able to click on that um but the few main topics here are performance reporting and insights action ability making sure that you guys have all you need to optimize your website within insights and there is already some bits coming there so don't worry if this isn't a huge um one for you then then it's already advancing but if you do have some thoughts to share with us then please do reach out because we need to make sure um and i suggest as well um we know that this is a big one making sure that that suggest is as optimal as possible in terms of results but then also the control that you would need associated with that product tagging um what data you need uh to use the tool in the most optimal way personalization um improvements on fred hopper there's already personalization options but but what do you need to see next and how exactly do you need that delivered to you international insights i already went through that but um about how we show all of the different websites in in one panel and then conversational experience almost like a chat gpt interface for for your website so that can help with search or even when you're on a product page to make sure that things are compatible there's so many things that can be done there we want to make sure that we're ticking off the best things first i'll just leave that open for about 30 more seconds in case anyone else wants to click on everything else and then we will call it a day just like to thank you all very much for your time uh and thank all the presenters it was really interesting from my point of view hope that you the customers uh loved it too please do provide feedback um via surveys or just via your csm if that's easier or whatever you prefer all right i will end that poll now and thank you very much have a lovely day