Noibu: Why the AI Copilot Is the Wrong Ecommerce Strategy

[00:00:00] Kailin Noivo: In Claude, we have a connector. Takes two seconds to set up. Not only do we connect all your behavioral data captured by Noibu into Claude, we also connect your third party, your Shopify, all your ads data. And then all you need to do is run a prompt automatically, and you could pick your schedule. And it will not only actually find the performance improvement, it will build the business case as to why it should be done, assess the ROI.

Alright, folks. Nice to meet you all. Quick show of hands. Who's been asked or is asking their teams to implement AI into their businesses? Perfect. From there, it's something that we've seen across the board on our customers from small to large. But on the other side of it, who's getting a little bit of AI fatigue? Because it seems like every product you use is, like, sending you an email with some stars and saying, hey. Great news. We've just launched an AI copilot. Cool. We actually think that the AI copilot's the wrong strategy, and I'll explain to you why. So we actually think the winning value is going to be connecting all of your data into a single source of truth to actually connect everything and harmonize across multiple channels.

We built a product called Noibu. And for those in ecom, which is, I'm assuming, everyone here, we effectively collect heat maps, your UTM parameters, some of your marketing data, your session replays, your errors, and all that really, really valuable data that we typically had locked up at a dashboard. But we've been really starting to lean in AI first and trying to understand what that means for our customers. And, really, what that means is actually not only connecting our data to your system of record for AI, like your Claude or your OpenAI, but it's also connecting your other ones. And what that enables you to do is unlock some pretty crazy value that I'll talk about today. So just going back super quickly, what would you do if you had a thousand interns tomorrow? Well, you'd probably watch every session replay. You'd probably look at every heat map. We've taken that approach by enabling folks to connect their data to an LLM, but then also providing them specialized skills to be able to extract the value.

Not gonna ask you to raise your hands again, but I bet you it resonates with pretty much anybody that you're being asked to do more with less, integrate AI. Everyone I talk to feels very, very constrained when it comes to resources, and no one's looking for more projects. So what we realized is people wanted less data and less AI, queryable data, and really what they wanted was outcomes. So what we built was effectively prepackaged skills that run automatically on your store to do jobs to be done. Great example of that is how often do you rebalance your ROAS between channels? How often do you look at the heat map data from the people landing on the actual website and then correlate it back to inform decisions? I bet you if you had a thousand interns, you'd probably do it every day. How often do you take a look at every URL and every campaign to understand why it's slow and open up and take it with your agency? Well, we've automated that too, and the results are pretty crazy.

So we've served mid market and enterprise customers and Shopify and every platform for a long time, almost a decade, but we wanted to test this with some of the smaller brands to see how resilient it is. And some of the early outcomes for this brand, Cars and Cards, on Shopify is they're able to automate some of their ROAS rebalancing. And, no, it's not the Terminator scenario where it's making changes without your permission. It's more serving work to you. Like, hey. I noticed that when this happens, your ROAS decreases. Do you wanna make this change? And the results are pretty crazy. In just about thirty days, their conversion went from point eight to 1.1 on the website. Here are some quotes. This one's really interesting. A nontechnical user of ours was able to improve their CLS directly without their agency, which is a Core Web Vital metric, which is really, really, really cool.

But really more importantly, we're talking about the Noibu MCP, which enables you to connect all of your behavioral data into an LLM, and then we give you the skills, which are effectively the prepackaged prompts, to be able to deliver outcomes pretty automatically. I'm gonna provide a quick little demo here. So in Claude, we have a connector, takes two seconds to set up. In the connector, not only do we connect all your behavioral data captured by Noibu into Claude, we also connect your third party, your Shopify, all your ads data. And then all you need to do is run a prompt automatically, and you could pick your schedule. And it will not only actually find the performance improvement, it will build the business case as to why it should be done, assess the ROI. And from there, it'll actually do a merge request automatically and send that to your development team so that they can review it and merge it or whoever does that in your business. All of this is automated because in ecom, everyone's teams are lean. Everyone's being asked to do more with less and implement AI, but their targets go up every year and the budgets stay flat, and in some cases, they decrease. So you need leverage somewhere, and that's why we went down the path of building this product.

Lastly, before I wrap up and take questions, what is the data we bring to the table? We bring all of your acquisition data, your behavioral data, all your heat maps, your sessions, your user events, everything related to technical site performance errors, all your session replay. This has always been valuable data, but it's been pretty much locked behind resource capacity issues. No one has the time to look at every session replay. No one has the time to look at every product review and then look at the user journey. We've automated all of that to the tune of a thousand interns, kinda like how we like to position it.

As part of a wrap up in thanking you for taking the time to chat with us today or take a look at my chat at least, scan this QR code. If you take a demo, we're willing to give you a month free of Claude, first 10 people. We're also willing to run this free assessment for you folks. And, yeah, I'm gonna pause to see if there's any questions, but we're really, really excited about this, and we think it's a game changer.

[00:06:14] Midroll: If you're listening to the ecommerce toolbox, you're entitled to a podcast exclusive website audit. Go to noibu.com/podcast-audit for a free scan that uncovers the hidden friction blocking your conversions and shows you where you're leaking revenue.

[00:06:29] Audience: Great presentation. You know, as a merchant, when I walk around here, what's different to a couple of years ago is every point solution is now trying to be an everything solution. What I mean is earlier, someone was just an email provider. Now that email provider is saying, thanks to AI, I can also be your reporting provider. I can be your ROI improvement provider. And so I think for me as a merchant, it's confusing. So, like, should I move from my 10 solutions to this one solution? While I have you, I'll just ask the second part as you answer that. I noticed that this is amazing because, basically, what you're doing is connecting all my data to Claude. What I have found is despite the models getting really good, they are not the best statistical models. You know? So Claude is like, I'm not an engineer, but it's a language model. It might make mistakes. So even though you might be giving it really good context data, Claude could be giving me poor hallucinations back. So those are my two questions.

[00:07:24] Kailin Noivo: Great questions. First off, coming from an x point solution, it's you guys that brought us here to be the platform because that was the number one feedback we'd get. So everyone is building point solutions. Now everyone's building consolidated platforms candidly because that's what people are asking for. They don't want to pay for a monitoring system that has this and then another system that has that, but it kinda has part of this. And then you replicate that over 12 systems, then you're paying multiple times for the same data capture and storage. So first off, we think that's the dominant strategy. On the second front, a 100%. That's why we think the future is verticalized tools because not only is it the context window, but it's also the data structure in the MCP that's really, really valuable. And, I mean, look. Candidly, our sales process is evolving to just be connected for you and give it to you, and then seven days later, you're begging me not to turn it off is the experience we've seen so far. So we kinda wanna let it speak for itself. And our answer to that is we're a vertical software. We only sell to merchants. We have platform integrations, so we have a high level of context. And then from there, we have specialized skills that have been developed over the nine years of our experience. And when you merge all those things together, it's really, really powerful, and we're so common in it. We're letting people kinda take it for a spin themselves with just take it for a drive. Thank you all for the time, and I really appreciate the chat.

[00:08:49] Midroll: The ecommerce toolbox expert perspectives is brought to you by Noibu. To find out more about Noibu and how we can help you debug your ecommerce site and rocket your revenue, visit www.noibu.com. That's noibu.com. And then make sure to search for the ecommerce toolbox expert perspectives on Apple Podcasts, Spotify, or anywhere else podcasts are found, and click subscribe so you don't miss out on any future episodes. On behalf of the team here at Noibu, thanks for listening.