Matt Zambrio: Thank you for joining us for our next topic, scaling e-commerce for Black Friday, from fragile to AI-ready websites. Today we have with us Meeky Hwang, CEO of Ndevr, leading enterprise-grade WordPress development and performance optimization projects. We also have Matt Dorman, President of Ndevr, overseeing strategic growth in technology innovation and WordPress and open source development.

Matt Dorman: Hello, thanks for having us.

Matt Zambrio: So I understand you two have a presentation you’d like to go through and share with us on this topic.

Meeky Hwang: Really excited. I am actually trying to, I’ll be sharing this.

Matt Dorman: We love technical difficulties.

Meeky Hwang: It happens, right? We’re all in a remote world, I think this is most of our daily lives at this point every day. All right, there we go, can you see the screen?

Matt Zambrio: Perfect, looks great.

Meeky Hwang: All right, so today we’re going to talk about scaling e-commerce for Black Friday. If you want to move to the next slide, since we got the great intro. The only thing really relevant here is if you want to find us on LinkedIn later, you can see both of our names, so go ahead and search our names, you’ll find us afterwards.

Matt Dorman: All right, so we want to talk about the stakes and get you prepped up for why this is important. We talked about Black Friday, obviously it’s a big set of days in the year, but why would you want to use AI for some of the implementations we’re talking about? And if you look at some of the things we’re going to talk about around audience and creator and developer experiences, the audience specifically are expecting very fast, very responsive, they’re expecting things that actually engage them more now than ever. So not just fast checkouts as in clicking fast, but fast in the sense that when they’re thinking about something, they want something.

Meeky Hwang: To give you a perspective on that, just to chime in, is that one-second delay is a 7% decrease in sales, which means that if you have a store that’s $1 million sales per day, that’s $70,000 lost due to the one-second delay. So that’s why having the fast site and fast checkout process and everything that’s seamless is very important, and that’s why we don’t want a fragile e-commerce store.

Matt Dorman: Yeah, and these are just standard typical numbers, but when you look at Black Friday specifically, in that hyper-focused time period when everyone’s giving deals, everyone’s trying to get them to their stores to check out, it’s not just about the seconds of delay making them want to think about things, but they’re going to start looking at other places and then they may or may not ever come back either.

Meeky Hwang: So OpenAI actually just announced this Monday, really early this week, that they’re rolling out instant checkout, which is in place, and now they’re planning to roll out to Shopify soon. That announcement alone just popped Etsy stock by 16%.

Matt Dorman: Yeah, so now we don’t need e-commerce stores anymore, right, Meeky? Well, let the LLMs sell all of our stuff now.

Meeky Hwang: Yeah, I guess you can kind of do it through there, but there are caveats right now, it’s not perfect. And I think, Matt, you actually tested yourself, but I haven’t actually tested it out, but from what I read, it’s just a single item that you can buy and there’s a lot of other caveats around it.

Matt Dorman: Yeah, I tested it out today, or yesterday, last night and today on my own paid account, and it was unfortunately not available. I don’t know if you want to switch that, I don’t know if you have tabs. So here’s my ChatGPT, I tried to get it to go to, specifically I was trying to be very specific, go to Etsy because I knew that they’re part of the program, get some stickers and try to customize it. Well, of course, you can’t customize it, so if you scroll down, it’s like, well, I’m a big fan of fat bear week, so I want some fat bear week stickers, if you don’t know what that is, you should Google it, it’s totally safe, see a bunch of fat bears.

So trying to get it to give me a fat bear sticker, end of the day, ChatGPT told me, no, it’s not available yet. It’s like, well, what do you mean, you just, like Meeky was showing, the stock went up 16% because this is available, it’s not even out. But I think at the end of the day, this first iteration anyways, you’re going to get to obviously surface your products, sounds like they’re going to utilize the Google product API or Google product feeds, so if you’re already doing that, and then also give you the ability, if you’re not on Etsy or Shopify potentially, I think those are the first two that are going to get rolled out, you could have that and then sign in and add your trade and you’ll be okay there.

But it sounds like it’s going to be a single product checkout, which if you’re running stores, you know that, well, that’s great, but I want to upsell and I want to cross-sell and I want to get my people to come back to buy other things, related things, and that’s what we’re going to talk about later. So hopefully they’ll improve the UI.

Meeky Hwang: Yeah, and also another thing is that there will be fees. So now Etsy shops and Shopify already have their fees, so for the store owner, you have to compensate the fees, but now OpenAI or other LLMs who are going to be rolling this out, they’re also taking the fees, and right now they didn’t announce how much the fees are going to be, but there will be fees they’re adding up. So for you as an e-commerce store owner, there will be a lot of other costs associated to it.

Meeky Hwang: Very nice. So moving on, when we were talking about audience experience and creator experience and developer experience, what we call is a 3D framework. So Matt, do you want to explain what this is?

Matt Dorman: Yeah, we like to use the analogy of a three-legged stool, three-legged chair where you have each of these pillars supporting a platform. So platform being your website, your commerce platform, potentially your social media platforms, your ad sales, again using the word platform. So all of those things resting on these three pillars where you have audience, creator, and developer. They’re pretty obvious, so any one of these things, if they get a little short or you’re not focused on one, then the whole platform starts to get wobbly, and if any of them break then the whole thing falls over, doesn’t matter how good the other two are.

Meeky Hwang: So to start with, to deeper dive into what these are, audience experience is pretty much what you see, the front end, what your customers see, feel, and interact with. So anything you see on the front end and storefront side on the e-commerce site or any platforms, that’s what we consider audience experience.

Matt Dorman: Yeah, and when you look at creator, that’s obviously the people that are creating the content for you, but also not just product details, but creating the social marketing feeds, the ad copy for you, if you’re cross-selling to other sites, you’re creating other copy there. So we like to have that in one platform that you’re delivering from one place.

Meeky Hwang: And developer experience is pretty much the engine room, technical backbone, technical foundation, or DevOps, anything that relates to your technology to push the code and make the features and everything, to make your site secure, stable, and efficient, that’s the developer experience.

Matt Dorman: Yeah, so first up is audience experience, right, Meeky?

Meeky Hwang: Yep. So there are things you can integrate with AI and some of the benefits, we just wanted to highlight a few that are more common and something you can look into to implement easier than other things. So the first one is AI search, which is AI-driven search, meaning that, to give you a perspective, if you don’t have an integration and have just regular search, sometimes you look for some products and if you have a typo or you don’t remember the exact name of it, those can actually give you zero results. So you might not be able to find the ones you’re looking for just because you don’t know the exact name or because you didn’t know how to spell it.

Matt Dorman: I was going to say, that’s kind of standard search today, or six months ago, right?

Meeky Hwang: Yeah, so as to when the AI search, like Algolia or Meilisearch or any of these other ones, Searchcraft or Chromia, I always pronounce that wrong. So they are pretty much what’s called a vector search, so it’s searching by the meaning, you don’t necessarily have to know the exact spelling or the product name, but with the context and all the vector search, it will search the right product or right context for you, which means you’re going to have a lower bounce rate because you don’t have to know the exact things to type in. So you can find it easier, which means they can convert better and leads to better sales, higher conversions and more sales.

Matt Dorman: Yeah, and if anyone’s looked at their search, what people are searching for, at least on a global basis, more and more people are searching websites the way they’re interacting with LLMs, where they’re asking it questions. So if you’re integrating with a basic keyword search and someone puts in a question like how do I check out, or what’s the shipping time, it’s going to give you some articles about shipping, but it’s not going to answer the question without an AI integration.

Meeky Hwang: Right. So the other one is chatbot, I think this is very common. So Matt, do you really use chatbots?

Matt Dorman: I use it to take out some frustration sometimes. I mean, some of them work really well and some of them just really don’t work and you end up in a deeper spot than you actually started with, and it’s typically from support-type ticketing questions for me.

Meeky Hwang: Yeah, I mean on the flip side, there were times that I used it that, if I was looking for basic questions that a lot of people would ask, I’ll get the answer right away instead of waiting for the next business day, 8:00 or whatever time zone people are working, to get the answer. So it actually does provide a happier customer experience with some of the simple questions.

Matt Dorman: Yeah, and I would say depending on the platform or the site hosting, whether you’re Shopify, WordPress, WooCommerce, or others, the integrations are all different levels, but this chatbot integration is pretty simple where you have to tell it where the information is, but the actual bot itself learns very quickly for the most part, and it’s going to be indexing based on your content. So that’s something good to have, and you probably want to make sure you implement it correctly as well.

Matt Dorman: Yeah, I would think you’re definitely going to want to monitor this and actually look at the answers it’s given, we’ve all seen the hallucinations, some of these are doing it as well, but those are more of the bigger ones where they have such a big mass of information.

Meeky Hwang: Yeah, and sometimes if you implement it incorrectly, you might come full circle back to exactly where you started the conversation from. But the reason we’ve separated this out with recommendation, that’s another set of AI tools where this is going above and beyond what we used to implement with recommendations, where you tag all your products and when you’re on a product page anything tagged relevant to it would get added, or maybe the top-selling, or maybe you’re just manually doing it. This recommendation is more attuned to the user’s interaction with your website and possibly even more if they use other tools on the site as well, so this is a big market increase where you’re going from 10 to 20% higher order value, they’re adding more things.

And then just to give an example from what you were looking for in the beginning, the fat bear stickers, without a smarter recommendation it might be just tagged as a sticker, so when you look for recommended items, it might be just other stickers not relevant to the topic. With a smarter recommendation, it will actually know the meaning of it, so it will provide you with a different product that’s fat bear sticker or item related, the better cups or mugs or tumblers, things like that, that’s more relevant to you because you’re specifically looking for that product. So that’s a better and smarter way to recommend it to upsell and cross-sell, that’s why it’s increasing 10 to 20% higher order values.

Matt Dorman: I think we’ll move along a little quicker because we’re going to run out of time if we don’t have the end Q&A and we have a longer Q&A setup. So behavior analytics, very similar to what we have today with Hotjar and FullStory, the tools have iterated on what exists today. And I heard on one of the earlier sessions where you’re talking about tracking how people are utilizing the site, and if you know what the expected user flow is, looking at Hotjar or FullStory and seeing how people are actually using it, it can give you recommendations on improving those things, but it’ll also just tell you where things are going wrong.

Meeky Hwang: All right, so moving on to creator experience, some of the better benefits, assistants, I think you can explain that better than I can.

Matt Dorman: Yeah, the assistants we’re talking about, and we’re trying to push forward that we’re not trying to generate content completely with AI tools, these assistants are meant to assist you in generating content. So a lot of input back and forth, ideally these, so Originality AI, the top ones are all integrated with your workflow. That’s the discrepancy right now, where a lot of these tools are outside of your current editorial process or creator process, where you have to go to the tool, do all the things, and then copy and paste, we’d like to see more of that being pulled into the actual content creation tools.

Meeky Hwang: So another thing is the content detection, this is actually interesting, I think I was talking to Matt about this. So one of the lines I put in here, sites that are perceived as human-written have 39% more value in site sales, which means that we’re making sure we’re not pumping out the low-quality chatbot AI-generated content, you want to make sure you have quality content that’s trustworthy. So what this content detection does is it’s actually detecting low-quality AI-written content, you want to make sure your content is authentic and it’s not penalizing your brand quality because of massive generated content that you’re putting out.

Matt Dorman: Yeah, this circles back to what I opened up with, we don’t want to prop up being able to generate content all through AI. So whether you’re doing it internally, you have a team, or a third-party team, it’d be nice to be able to check on them. So you could use this in two ways, either they can use this tool to make sure the content they’re working on with an AI tool is not too AI-like, or just validate the content coming in that you’re getting sent or being paid for, that they’re not just using AI tools, because I guarantee if you’re sending it out to freelancers, they’re saving their time by generating a lot of AI content. So something like Originality AI or GPTZero are the tools that can help you with detecting this.

Meeky Hwang: And this is more common in the educational industry, but this is also applicable across all industries, because it’s easy to tell sometimes if it’s GPT-written or AI-written, so you want to ensure that’s not the case for your content.

Matt Dorman: Yeah, and this next area is similar in that it’s going to be tools that will evaluate your content, but this is a little bit different where it’s not evaluating it based on quality, it’s based on how good it is for your identified audience. So we use ICP, the people you’ve identified as your persona, as your customers, the people actually coming to your site, you want them to purchase. So if you put all that data into these tools, ContentSeer is actually one that we built, we’ve been testing it out for a long time on content sites, works really well on brand sites as well, where it’ll give you grading on existing content or content you’re creating, to help you map keywords but also just general content in a way that those people are going to either want to buy something or want to come back to read more.

Meeky Hwang: And another application of this personalization or persona application is that you can actually interview your ideal persona. So you might not be able to have access to all your customers, then you can actually simulate the persona interview by just getting the pain points and understanding pain points, so you can apply that to your content or products or workflows so you can make it a better experience.

Matt Dorman: All right, so workflow automation. These types of tools are really around automating the actual process that you generate content. So we’ve talked about a bunch of tools, you can then somewhat automate those functions for going one step to the next. You can also automate the approval process, which is really nice in high-stress, over the next few months, you’re probably going to be generating a lot of content around your products, and maybe you have a workflow where a lot of approvals, maybe some of those can be removed by these types of tools. So you’re now accelerating the time that one person can get something from an idea to being potentially published, if you’re going all in for that.

Meeky Hwang: And then also blending into some of these things as a content detection, you want to bake in those checks and balances before you publish your content.

Matt Dorman: Yep. All right, moving on to our favorite experience.

Meeky Hwang: Yeah, we are primarily a tech agency and the two of us are very technical, so we use these tools daily for all kinds of things, so it might get a little bit boring, but we’re going to cut this one really short and go pretty high level. We’ve been using, I mean, Meeky, basically you could replace all these, like we were talking about before, you just use like Lovable, right, or Bolt, and you just tell it what you want and you don’t need a code environment, and it just produces a site for you.

So side note, we actually try to test it out because we want to know what vibe coding is and how far the AI coding can go, but unfortunately it’s not there yet, so we had a lot of fights with the vibe coding. And I think there was an article I was reading about this this morning, that Microsoft is actually trying to roll out something called Copilot, or they want to do the vibe workload, I forgot the exact term for it. So instead of vibe coding they want to make the AI do some of the work that people are doing, and the result is that it’s actually negatively impacting it, because people actually spend more time fixing what’s not written correctly, or your coworker is fixing whatever you started with the vibe workload, they’re spending way too much time fixing it than just starting from scratch.

Matt Dorman: I’m so sorry, I see what you mean, I will try better next time. How many times have you seen stuff like that around vibe coding or even ChatGPT and other LLMs?

Meeky Hwang: Yeah, this is exactly the response I was getting. I think I was fighting with vibe coding at one point where I was like, you didn’t do anything, how can you not fix this, like, you’re absolutely right, I completely didn’t understand your question, and then same results. So that happens, and then imagine that happening in other work, not just some of the specific work you want AI to do, so that’s giving a lot of headaches. But on the flip side, it does do the basic things, and more tedious, time-consuming things faster and easier.

Matt Dorman: Yeah, so we like to look at this as a flow of the way you’re doing development. So you’ve got code assistants helping you inline, your developers doing code updates, whether they’re on templates or in actual code. You’ve got code quality check, that’s not replacing but augmenting team leads and managers. And then you’ve got the next step of deployment, being able to put in actual AI testing or determining where to deploy things, that automation is important. And then finally, because that’s usually integrated with deployment, the testing, so you test during deployments but you also test after deployments and you’re monitoring.

So we talked about some of the Hotjar and things like that in the audience, a lot of those tools can be utilized in part of the developer experience where they’re testing out making sure all those are integrated, and every time there’s a deployment these all get run, because the last thing you want to do is have an outage that you don’t know about.

Meeky Hwang: Right, and also you want to catch potential issues before it happens, especially around holiday sales like Black Friday or even Cyber Monday, these days it’s not just two days of sale, it’s like cyber week almost, so you want to be prepared for all this traffic spike and all the potential issues prematurely before it actually happens, so that’s why testing and monitoring is very important.

Anyway, so moving on, we talked about three different areas of what you can apply AI integration. So just to give the recommendation on recap, it’s more, did you mute yourself? So audience, we have prioritizing the AI search, this is leveling out what you want to implement, so we feel like AI search and recommendations is something you’re going to get a lot of impact on on your website. So chat is important, but very easy to implement, so it’s almost a no-brainer, obviously it’s going to get revenue growth. Now if you implement it, you don’t want to implement it a week before your sale, you want to implement it almost now so that you have time to test.

Matt Dorman: Meeky, you’re muted maybe, I don’t hear you, or is it just me? Oh yeah, looks like we lost your sound, Meeky. I’ll do the creator while she’s getting it, interrupting me when you’re in. So we got creator, we talked about those tools around really the process of creating your content, and again not just in your e-commerce system or CMS but in all the systems. So if you can get the automated campaigns, the automated processes down so that you’re augmenting your thing but also integrating the detect originality, that’s going to allow you to have better brand recognition, and of course the end goal is SEO or AEO, to get integrated into those tools.

And then finally the developer, are you back yet, Meeky? So developer tools, and again we’re mainly developers but this helps support the other two really well, where these are always testing the things you’ve already put in place, these are integrating the things you’ve already put in place. So if you’re not doing this kind of workflow or methodology, something is going to get missed and something’s going to get broken along the way if you’re not monitoring, and you’re going to have downtime during that.

Meeky Hwang: All right, can you hear me? All right, so to wrap up, if you don’t adapt, actually no, if you adapt, it drives more sales, less loss, long-term resilience. On the flip side, if you don’t, Matt, you…

Matt Dorman: Well, if you don’t, and I think we talked about this pretty specifically, you’re going to have more downtime, and specifically around Black Friday. And we all know this, if you have downtime during Black Friday, the 7% or whatever percentages we’ve thrown out there are magnified by maybe a hundredfold depending on your industry and how much you’re pushing during that sale, so it’s a massive risk, and to do nothing would be the opposite of doing all these implementations. And by doing that, you’re going to be hurt by other people advancing fast, so you may not have downtime but you’re still going to lose.

Meeky Hwang: All right, so we do have an AI readiness quiz, we’re going to show a little bit but then we can move on to the questions. So if you are curious about how ready you are in terms of AI, you can go to this URL or QR code and just test it out. I’ll have a separate video kind of overexplaining what the results are saying, because it also gives you some perspectives based on your industry, so it compares with the average what industry is doing versus how your score is so you can figure out where you are in the AI readiness side.

Matt Zambrio: All right, some Q&A for you both, we have a couple questions from the audience. So first one, if you could only instrument five metrics for the war room, which are they and what alert thresholds do you set?

Meeky Hwang: Five, I’ll let, she said she was going to throw me under the bus. So five, yeah, I was going to throw you under the bus. So the metrics, I’ll start with developer experience. Some of the metrics you want to put in, I’m techy, so we want to make sure that your deployment, there’s a DORA metrics that you can actually decide to figure out how well your development workload is doing, so it’s like how fast you’re deploying and what are the failure rates, I might have messed this up. So there are four things, and also how quickly can you revert if there’s a failure, and I forgot the fifth one, so I’ll probably get back to it. So you can put those metrics in to make sure that your workflow for feature development and deployment is fast enough so you can always be on top of the most relevant features you want to push out.

Matt Dorman: Yeah, and from the audience side, I would probably go with your bounce rate as the first one, especially around Black Friday where you’re going to be sending a lot of traffic. And this is not, I would not, given that you’re sending from ads, obviously the bounce rate is going to be different than when you’re just getting organic traffic, so it’s hard to say thresholds on this one unless you have past data around these types of ads, so you know what type of bounce rate to expect because it’s going to be high, higher than organic.

And I would even throw in the obvious cart abandonment. This one’s hard too, because you might have really good ways of pulling those people back in that actually end up upselling more things depending on your technique, getting them to buy more things after they’ve abandoned the first time. So I’m not going to have good threshold numbers for any of these because they’re very variable based on the site and what you’ve done in the past.

Meeky Hwang: Yeah, also depending on the business as well, and also the size of your business, it’s really hard to just say one size fits all. There are recommended thresholds you want to get close to, but sometimes you might be doing way better than that, and you don’t want it to just be comparing to the average of the industry, so it really depends on your business how well you did compared to your previous data. Some of the basic things you want to track like sales, your sales number might be completely different from another company or even another competitor, so those things have to be set by your own term is probably a better threshold.

Matt Dorman: Meeky, do you want to throw in like time from ideation to product live? Like we do measure that in creator, that’s using all these tools, you have an idea of a product or a way of promoting a product, so you have to generate all the content, so that’s a thing you can track, especially when you’re in the weeds of sending out promotions really quickly, knowing how quick you can go from a sale idea or promotion idea all the way through executing it across your platforms. Knowing how fast you can get that up is something you should measure, so when the CEO says I need more sales, well, I can do this now and it’ll be up by X time today, it’ll have this measure, or it’s too late, we’re already done.

Meeky Hwang: Yeah, it’s similar to in the beginning when I was talking about how fast is your deployment, you also want to measure how quickly you can push out your content, because there might be some random bottlenecks that are not technical, but you want to eliminate that by reviewing it frequently.

Matt Zambrio: So it very much sounds like there are some things that are subject-based on the company and others that are going to be objectives that are pretty standard across the board.

Meeky Hwang: Yeah, I think so.

Matt Zambrio: Next question, are there any white papers or similar things to understand better how we architect our CX?

Meeky Hwang: So we have a checklist that we actually, there’s a bunch of metrics and how hard it is to implement each thing, so we can share that after the fact.

Matt Dorman: Yeah, that’s a good one. I was going to say the bigger brands like Forrester have some good things around individual products, it’s where a lot of those lay flat, around the integration between the two. When they’re really good at evaluating a single product and telling you how good this one product is, there’s not a lot out there that says, all right, these are the ones I need to use and then I’m going to throw this one in and how is that going to work out. I mean blogs will do that sometimes, but it’s going to be pretty lightweight without having someone actually come in and consult with you a little bit, or just internal, you get all the product information from the various places and then you can look at them and see how well they should integrate with each other. But the multiplier, it’s not going to multiply the 7% and then the 20% and then the 30%, they’re all going to overlap a bit.

Matt Zambrio: Next question, which bot classes do you rate limit or challenge during promos, and how do you avoid collateral damage to real buyers?

Matt Dorman: This is, I have an answer, but Meeky, if you want to, well, you can have your answer at all. So the hard thing is, you kind of right now have to spend the money and allow the bandwidth bills to go up a bit. Now that depends, if you’re on a platform that’s super expensive and you’re not on a more custom platform where you have control over that, it’s really hard to differentiate right now, especially at one point a lot of companies were going to block OpenAI’s tools. Well, now as of this week, that’s how you might get users to buy your product without even coming to the website, and they’re actually finally getting your content to the place where they’re going to make a purchase.

I don’t think the first iteration is going to be good, so it might be okay to still block them, but there’s not enough data around who’s too small and who’s not going to be too big in two weeks.

Meeky Hwang: Yeah, so it’s not one-fit-everything question unfortunately.

Matt Dorman: Yeah, unless they’re coming from Russia or China, right? I guess I could have started with that. I mean, sorry for international, but if I’m in the US and I got stuff coming from China or Russia, they’re just trying to scrape my site to then replicate it somewhere else.

Meeky Hwang: Yeah, I mean, like we did explain to one of our clients too, that you can’t just completely block everything, because there might be something, but if you know the patterns, because this was US-based, if you know that you’re not going to be targeting outside of US, just like Matt said, you can block out all the international IPs. So you’ll learn the patterns to block it, but if you block everything you might lose your customers or potential clients, so it just needs to be, and sometimes there might be some people actually putting in the weird, it’s not just a bot, so you also need to distinguish that, some things really look like the actual human interaction, those things unfortunately can’t be 100% blocked but you need to be a little open to it just in case you might be getting some customers in that pattern.

Matt Zambrio: No, I think those are great insights. And then last question that we’ll have time for, do you have a process mapping tool recommendation to see the end-to-end CX journey that you could share?

Matt Dorman: I do not, I’m going to say that I’m the least process person on the panel of the two of us, I implement what Meeky tells me to implement on the process side, but I don’t know if we have real tools around the mapping part of it. I mean, we’re big process-flow people, and those are very low-tech, but it’s almost the Excel of process mapping.

Meeky Hwang: Yeah, I’ll have to think about it and then I could share that with you after the fact, because I always remember things after the fact, so when I do that, I’ll share that with you guys so that it’s easy to access.

Matt Zambrio: No, I think the audience would certainly appreciate that. Just want to thank you for the time. Just one more thing, any final thoughts that you want to share, and your thoughts about Prepathon so far?

Matt Dorman: I’ll let Meeky go first again. No real final thoughts, I mean, this is an interesting world, this whole year has been an interesting year, as you’ve seen from a lot of the other presentations. Some of this stuff is not really new, so do the things that you’re doing really well and execute them the way you thought you’re going to be doing them would be one way of putting it in a nutshell. Yeah, there’s a lot of great content on these sessions, so continue watching the rest and watch the ones from yesterday that you missed.

Meeky Hwang: Yeah, and another thing, I think the whole thing of AI integration, I’ve mentioned it multiple times, is that you probably have to have checks and balances in place, because some of the things that are way premature, like the OpenAI instant cart checkout and all these other things, it’s a little early stage, you want to make sure you take advantage of it but at the same time it’s not perfect, because there has to be some checks and balances in order to be successful, because you don’t want to be trusting too much for any area. So I saw some of the other presentations around how to effectively use it, that was very helpful, but one of the things we want to make sure is that you use it wisely and also make sure you check your own AI content creation.

Matt Zambrio: No, that’s great. Matt, Meeky, thank you again, really enjoyed the presentation today, thank you for joining us. Up next, we’re going to have Moeez back for a Black Friday survival challenge and then we’ll end the day with a final presentation by Christian Helms. Thank you so much.