Improving conversion and reducing revenue leakage at Mejuri

Speaker A: Foreign. Welcome to Tech Transform, the podcast that explores how technology is reshaping the enterprise landscape. I am joined by Kaelyn, president and co founder of Noibu, an E commerce analytics and monitoring platform that unifies site monitoring with experience analytics to surface both revenue risk and conversion opportunities in one console. And we're also joined by Rohit Cpto at Majuri. It's a leading direct to consumer jewelry brand and he's going to share how they've leveraged Noibu's platform to drive measurable results. So Kaelyn and Rohit, welcome to the show. Very excited to have you both on. Before we dive in, could you each introduce yourselves to our audience, tell us a little bit about what you do. We can start it off with Kaelyn.

Speaker B: Thanks so much for having us on the show, Christina. My name is Kaelin, co Founder president Noibu. Like you mentioned, Noibu is an E commerce analytics and monitoring platform. We'll get into what that means just in a little bit. But, but super excited to be on the show.

Speaker A: Fantastic. Well, great to have you on. Rohit, what about yourself?

Speaker C: Hey guys, uh, thanks for having me here. I'm um, Rohit. I support multiple functions at Majuri, all marketing, product management and engineering. As Krishna mentioned, we are a global direct consumer jewelry company with 59 worldwide stores. Really excited to be here and talk about E commerce and E commerce monitoring.

Speaker A: Fantastic. Well, we're lucky to have you both. They're both here to talk about from two different perspectives, I think. So to kick things off, I want to start by, by looking at the day to day reality E commerce leaders are operating in today. It's far more complex now than it was even just a few years ago. Things like, you know, more tools, rising costs, even the fact that consumers have much higher expectations, it's all contributing to this. So from your perspective, what are the biggest challenges when it comes to driving revenue and conversion growth? What do you think about this, Rohit?

Speaker C: Yeah, great question, Christina. Uh, I would say from the brand perspective, from retailer's perspective, we start with the macros. What happened last year with the tariffs, that was one of the biggest headwinds, that the industry is still struggling and we all had to figure out how to adjust to this new reality. On top of that, the second macro factor is what's happening with AI is really rapidly changing the consumer behavior. And that's one thing that every single brand leader or someone in my shoes is constantly trying to stay on top of. But having said that, like what you mentioned are the list of ongoing challenges and within the given context, I would assume that personalization and ability to connect the dots, those are the two challenges that doesn't seem m to go away. So when you think about personalization, it comes down to one to one. How do you crack that one to one personalization. What do you know about your customers so you can offer them the best features, best product selection and best discovery experience. And then when it comes to connecting the dots, it's all about format. Um, click to conversion. And no one seems to have really cracked it. We have taken some strides over there and we have some improvements along the way. There's so much more to here.

Speaker A: Yeah. So there's just a lot going on. I think for all of us. There's so much going on. But you've highlighted a few of these important things that e commerce leaders are having to confront. Right. Kaelyn, is there anything else that you would add on something that you're seeing? Maybe probably even you would confirm exactly what Rohit said as well?

Speaker B: Yeah, I think it's kind of a bit of a perfect storm. People definitely built out large infrastructure during the COVID era, uh, largely on the back of horizontal platforms that kind of bring you 80% there, but you have to bridge the last 20% yourself. And I think what we are kind of seeing with like the specialization of commerce engines in like the Shopify and those types of businesses is that really kind of owning your niche and doubling down compounds over time. So just to really add on to what Rohit mentioned, we're seeing a lot of cost pressure. People are not willing to have multiple tools that kind of do the same thing because then you're paying for data capture and storage for kind of the same data in like three to four different places.

Speaker A: Okay, so the perfect storm pretty much is how you summarized it all together. And we'll go back, we'll explore a couple of these points more today. I want to explore how these challenges surface in a real concrete example. So I'm going to throw it back over to Rohit. Rohit, thinking back to when Majuri first started using NOIBU for front end error monitoring, what problems were you trying to solve for? Exactly. And how did those initial challenges influence your broader approach to conversion optimization across teams?

Speaker C: So the context here is that we had just completed our re platforming to Shopify and it's similar to rebuilding the plane while flying it. We really had to keep the business running. We were really hoping that once we had this new tech stack, most of our follow up problems would go away. But some of these error monitoring problems were still unresolved. And essentially what's going on was that there's the core problem was the signal versus noise. Right. Like we would get this fire hose of all the alerts, but no real data behind the customer context or the commercial impact or how many users are impacted. And this is such a common problem that engineering or tech teams faces. And that's where NOIBU comes into the picture. And we did a two week poc, I still remember that. And it immediately identified one spinning card issue for us. And it was going on for several days or maybe a couple of weeks since the migration. And we were just unaware of that revenue leakage. As soon as we saw that, we were able to investigate really quickly and were able to resolve it. But you can imagine how many of those revenue leakage is happening for any brand that's out there.

Speaker A: Fantastic. Well, so we're seeing about, learning about how these real world issues, we're learning about what happens as they're happening. The next question then is how can brands stay ahead of these issues like what you mentioned Rohit. So throwing it back over to Kaylin and we already talked a little bit about this, but AI. So AI is becoming increasingly central in E commerce as it is across most industries today. How can we make sure that it's genuinely adding value and we're not just following some hype trends? And also how can brands, how can they leverage it to do more than just detect errors?

Speaker B: Yeah. So like Rohit mentioned and the cost pressures and customers asking us for features that were adjacent that we didn't have two, three years ago is part of the reason we started going down this journey. In commerce, we don't care about 10,000 bugs, you care about your top eight. So like getting from 10,000 to eight was a really hard problem that took us years to solve. And we actually solved the final mile of it by implementing AI. And it wasn't until that like we had it working but like it really wasn't like quote unquote perfect as defined by like. Right. Nine out of 10 times until we implemented AI. And that was probably in like 2022. So at that point we're like, okay, we have to broaden out. And what does that mean? Well, now that you've stabilized the website for bugs, you want to make sure that it's fast. Right. So what do you do to make it faster? Well, you need to know your core web vitals. You need to understand as you're making releases if it gets faster or slower, and you need to do automatic release updates with AI where basically you don't have to pay someone to go look after every code release, what KPIs got better and worst, you can just automate that. So we went from like errors to errors and performance. And then we said, okay, well once your website has no bugs and it's fast, what do you want to do? Everyone wants to do AB tests and they want to basically innovate. So that's why we launched page analysis to help you understand your customer journeys. Where's your marketing spend going? How are they converting? What's the behavior on the website? So that more behavioral layer, and that's really where we helped consolidate a few vendors for Rohit's team as well, is bringing all of that into a single platform. But to answer your initial question, like where does all of this play in with AI? Well, how I think about AI is that everyone's doing the same thing today. I use Whoop and Whoop's a great product, but everyone's taking the data they have and connecting it to a foundational model and enabling you to like, interact with the data you have is basically everyone's AI strategy right now. And that's our first AI product. But how I'm really, really thinking about AI is how can we help automate micro workflows? And what I mean by that is a good example of a micro workflow is when someone does a release in most organizations. And Rohit, let me ask you, when you guys do a code release, what do you do?

Speaker C: After we of course run all the testing that we have to, we go back, look at the metric and ideally it's an A B test. So we can go and compare performance.

Speaker B: Exactly. So you look at your performance. Did we introduce any new bugs? There's like a very common checklist. Right. So with our new release monitoring agent, the agent just does those checks. It'll tell you in 1 hour, 6 hours, 12 hours, 2 days. Did your website get slower? Did you introduce bugs? Is there a large change in conversion rate? So we believe that the value is in amalgamating hundreds or thousands of those micro workflow automations.

Speaker A: That makes sense. So the first step you said when it comes to the AI part is that people are plugging into the foundation models like you said, and they're using it to chat and ask questions about, uh, the data and so on. But one thing I'm thinking about here is how manageable is that overall setup for teams on the ground that's the next topic actually that I wanted to discuss, which is around tool consolidation. So when brands consolidate tools, you've already mentioned quite a few already in our conversation today. What should they be consolidating around? I mean, is it data types, is it teams, is it outcomes? What does a good consolidation look like in practice so you don't lose visibility while you're trying to simplify everything? Rohit, I don't know if you all have experience with this, what are your thoughts on the tool consolidation?

Speaker C: I would say ideally it has to be around the outcome, but you have to be really careful with how you are defining that outcome. Right. And if you're one of those organization which has a very clearly defined KPI tree, it's easier to get there. You can have a point solution for a specific node on that KPI tree which is what you really need, or you can keep going up on the KPI tree as you consolidate. So what I mean by that, like if you take Noiboot as an example, you think about errors as an engineering metric, it can be part of the KPI tree. But I think of this in terms of sales per visit. That's our core digital funnel. So when I think from the sales per visit perspective, error monitoring is just an element of it. I think about the funnel analytics as well. I think about the session replace which is typically used by our product managers or more by product designers. So all of these tools can now be consolidated against a single outcome which is sales per visit. So the starting ideally you want to go through the outcomes, but you need to have a strong KPI tree where you know how this all KPIs ladder up. But you can also do it along data types and teams as well. Like if you call workflows, that's where teams come into the picture. Let's say linear. So linear is a good project management tool. I have rolled it out across marketing, product management and engineering. Typically it's used for engineering, so it's a workflow solution. So it applies to other use cases as well. Similarly, like you can go on and on and look at like other examples.

Speaker A: Okay. I think of course simplifying tools can make a really big difference, but I think the real test is how all of this holds up in moments of uncertainty. Right. And we've already mentioned it with the tariffs that started last year, there's so many different things that we can't predict right now. So throwing it back to you Rohit, what operating habits help you protect revenue and still keep momentum, especially in these Times of uncertainty on conversion improvements, as we've been discussing, without slowing delivery or without creating any war rooms.

Speaker C: I would love to say there's a solution out there, but for me personally, it comes down to culture. It comes down to the mechanisms that you have established in your company. So one big bucket, I would say, is continuous improvement. Do you have proactive monitoring? Do you have a, uh, really robust culture around experimentation, A B testing? Do you have mechanisms to review those outcomes and create feedback loops so you can adjust your roadmap in an agile fashion? This sounds theoretical and execution is not easy. It requires a certain form of discipline. And that's where the culture and mental model comes in, where you allow for calculated risk, where you create, allow the freedom to learn from mistakes. Processes like postmortem and all those things comes into the picture as well. So, um, in my mind it's a lot about process and culture. And the second element is I keep bringing up connect the dots. I think it's really important, especially in the period of uncertainty. Right. For us at Majiri, we sometimes forget how important it is to understand the new marketing campaigns which can happen inside out. To give you an example, I remember we were scratching our head about the conversion and drop off and we saw that it's all coming from meta and it led to this big deep dive. But what they did not take into account that we had completely changed how we were doing campaigns on meta. We went from conversion driving to more clickbaits, which worked really perfectly in our context. But it had a negative impact on one of the KPIs. But overall it worked for major. Right. So I think that connecting the dots is often missed in companies. So when you focus on those culture of processes and mechanisms, right, like you get better over time.

Speaker A: Okay, so connecting the dots, the culture piece, culture, of course I think is incredibly important. So Kalyn, I want to throw it back over to you and I want to talk about one of my favorite things, which is data, the data itself. So many teams, they have lots of data, but. But they're struggling to act on it. Do you have some sort of framework for turning data into decisions? Like how do you connect all this data from customer experience, all the signals that you're receiving into business impact so teams can prioritize?

Speaker B: Good question. A common theme we hear is data rich, insight poor. Rohit would know this. The amount of data capture from tools is really, really high. And a lot of the data captures overlap with each other, which actually ironically, like slows down. The website, creates conflict of like third party code. So you really actually don't want like 30 different things pulling data from your website. And then the worst part is these providers are actually charging you to store it because that's one of the things that they basically use to come up with their pricing. So how we look at this is if you want to turn data into insight and you're a horizontal company, meaning you sell retailers, banks, this in the absence of retail being 90% of your business, you're optimizing for a different variable which is like portability of data. Like basically you're bringing somebody, uh, build your own adventure versus bringing them an opinionated product that they can like fine tune the last 5% in an easy way. So like I think we realized that like pretty early on where when we were just an error monitoring company people were like, but I have new relic. And we're like sounds good. But like how many bugs have you solved in the last month? They're like well not that many. And we're like why? Well there's 20,000. We don't know which ones actually we should work on. And like okay, well you have two options. You either buy our product or you hire a person to study the errors and try and understand which ones are hurting conversion. And, and I think that initial insight could be applied across multiple different categories of tools. So I mean I'm just going to be super candid. Like I don't think you can solve this problem if you're serving multiple masters. And the only people that are going to be able to solve this at a data and an intelligence layer is doing it vertically. And unless there's some super quantum leap in AGI, even with our AI products, it doesn't have the context to just like solve that problem for you. And there's also another thing which is kind of separate but adjacent to this. Rohit would know this. First off, collecting, storing and transforming that data so that it's actually usable is really, really challenging. Multiple different formats, different things and you have to solve that problem before you can get to the end state of automating some of those micro workflows or to be able to interact with your agent in a way that you get very valuable insights. So to answer your question directly, I think not only the reason why NOIBU is well positioned, but I genuinely don't think this problem's solvable horizontally and I think that's how most people are approaching it. So yeah, that's my bit on it. Obviously I'm a little biased but yeah,

Speaker C: yeah, I'll just jump in with one comment over here. Like, I mean, completely agree with what you said. And the way I tried to bring this up in m my team meetings is like, know your business, know your customer, know your goal, and they're all related. And when you do that, you know what you're seeking, right? Like you know the questions you want to ask or answers you're seeking. And that's when data becomes easier to action on.

Speaker A: Um, and I do want to quote back to sum up everything that we've just said. I really like one of the first comments that Kaelyn made, which was that we're very data rich, but we're insight poor. Data is incredibly valuable, but you got to work with it. So back to you, Rohit. New technologies, new features are moving very fast in this space. How do you all at Majuri innovate quickly but also without introducing unnecessary risks to the customer experience, especially around releases, for example, what guardrails, what signals, what shared metrics help you move fast with confidence?

Speaker C: I would say like it again goes back to the culture. I think you need to have strong culture. You need to have strong set of operating rhythm and process. It all starts there. It's the culture of risk taking. It's the culture of looking back and understanding what went wrong and really truly getting into the deep of like, what were you thinking, why those decisions were made. And that's how you take corrective actions and assume that those hygiene factors are in place. Then you're looking at the basic table stakes, like whether it's uh, NOIBU or any other analytics tool around monitoring to make sure after the release you have a way to monitor progress and make sure everything went right. We ideally like to a b test everything to reduce the risk and also to be able to quantify what's working, what's not. But I think most of the tech teams or leaders that I talk with, these are table stakes at this point. The part where I've still seen folks struggle is what I call silent failures or the revenue leakage. And these add up over time. And that's where I, uh, really love what Kelly and you guys are doing because that's really help us cut through the noise, find stuff that's actually hurting customer experience or creating some friction and find those opportunities and go ahead and fix them and prioritize and fix them. And it can save hundreds of thousands of dollars or millions of dollars.

Speaker A: Yeah. And can I ask, do you have any like, tips, tricks when it comes to establishing the culture, the mindset that You've mentioned like anything that, any behaviors, anything you've done at Majuri to be able to prepare within the company.

Speaker C: Yeah, I would say, like I'll give a simple example. When I joined we were not experimenting, we were not doing a B test and I had a very simple goal. I want my team to run one a B test per sprint, no matter how they do it. Just a very simple. I was like, I'm not going to put any constraint, I'm not going to care about being data driven and all that. Just run one a B test. It sounds simple. In order to run an a B test every sprint, every two weeks or every month, it takes a lot to get to that point. And once you get that discipline in it organically helps you clean up a lot of mess. It helps you figure out like, do you have the right team, do you have the right processes, do you have the right data, do you have right the tools? So for me personally, sticking to the basics really helps. And then of course you have to be clear about your expectation. You have to support your team because when you think about experimentation, things going to go sideways, you're going to make mistakes as you just mentioned. Uh, so being there to support your team and also create mechanisms to not make the mistake again, I always say that like don't repeat the mistake and it goes back to the learning. So it's like go back, look at where the company is, where your teams are and go focus on the basics.

Speaker A: Yeah. So learn from your mistakes. Make a culture where people are not afraid to fail because you're going to learn from it and move on from that. Caitlin, do you have anything else to add to that around like the culture, the mindset? Because it sounds like you, uh, all@ uh, Noibu as well, you're innovating fast. You've got to have the right culture to be able to do that.

Speaker B: Yeah, it's so interesting. I've seen, not in our organization, but with customers and prospects when you start to have silos, and Rohit knows this, I'm a huge advocate of the ctpo. So the consolidation of technology, product and marketing in retail, I think that is like the smartest consolidation. And I'll explain to you why. If you look back, you have three distinct, sometimes four distinct departments that roll up to two strong function heads. You have product and technology and then you have like E Comm and marketing. And when you look at this, they each have their own tools. They'll have an APM performance tool, they'll have like a Heat, uh, mapping ui, UX digital experience tool. And then they'll sometimes have like a analytics uh, tool that kind of somewhat sits in the middle and that's kind of like the Switzerland, it's like the neutral ground. But then you have your own tools. And often I've seen in very poor cultures, obviously fortunately not with Rohit and Majuri, but those tools start to get weaponized. People fine tune them. They don't give full access to different teams. And I've seen that create a lot of challenges. Right. And even in some buyer, more old school customers. Right. The concept of putting revenue signs beside technical issues.

Speaker C: Whoa.

Speaker B: I don't want this team to have access, I don't want this person to have access. Or you create a lot of conflict internally if you're not growing in the same direction. So from a cultural standpoint, I'm a huge fan of the everyone rolls up to the same leader. That's the first thing that I would say. Where marketing's using a tool, engineering spends 90% of their time questioning the validity of the data. And then it becomes about disproving the validity of the data. And that's where all the effort's going versus like solving the problem. And even if the data's not accurate, it's directionally accurate. So from a cultural standpoint, I think it starts at the top and unifying teams. And then as much as the tool could support culture, I think that is going to drive the most amount of efficiency because sometimes you're looking at data and you don't like what it's telling you and you have two options. You either work on the solution or you try and disprove that there's a problem in the first place. And I find when there's a unified leader and a unified tool, it's a lot easier to actually roll in the same direction on that topic.

Speaker A: That makes a lot of sense. Unified leader, unified platform. Right, Unified tool. So I want to shift gears a bit. There's this one question that I wanted to come back to. Kaylin, you've described noivoo as an E commerce analytics and monitoring platform. So when you say platform, what does that mean in plain terms? Like what workflows become possible that aren't possible with point tools? This centralized view that you were just discussing right now as well.

Speaker B: Yeah, it makes sense. And I want to preface this where you almost need to build a series of point solutions that are interoperably connected. And what I mean by that is like you don't want to build like a platform that kind of does everything but does nothing. And how we approach this is we went really deep for a good amount of time on certain things, and then you expand it out. But to answer your question, what can you do? Well, first off, connecting these systems and managing integrations, especially as SaaS products are shipping code every week. Disaster. You don't want to pay someone to do that right from there. For me, a platform, it needs to be out of the box, but have configurability. Kind of like very similar to what Rohit's team did with Shopify, right? Using Shopify out of the box, but you're using a custom app on the front end through hydrogen, where you can control the customer experience. So for me, a platform, it needs to be good enough to serve the teams that it says it's going to serve. It needs to be opinionated enough where you don't need to hire big teams to manage it and there's very little overhead and. And people need to actually work out of it in unison. And for us, it means that we need everybody from customer service to marketing to product to engineering, collaborating out of the tool and trusting their part of the tool that they care most about. And that part of the tool needs to be powerful enough for them to be able to do their job good.

Speaker A: Well, we're coming towards the end of the episode, and I do have one last question for both of you. A very important one. So to wrap things up, uh, this E Commerce leaders, they have a lot on their plate right now. It's clear they're trying to keep up with this rapidly evolving space. Lots of different moving pieces. We've only gotten a taste of it through our chat today. So if you could give e commerce leaders three practical strategies for 2026. So one for visibility, one for speed of action, and one for team alignment, what would they be? So, Kaylin, I don't know if you want to start this one off visibility.

Speaker B: I actually, maybe this is a contrarian opinion. I don't know if people have a visibility issue. And what I mean by that is like if you have a data access problem, that's very kind of a T0 problem. What I'm hearing in the market's actually the opposite. People have too much data. So I would say definitely you want to solve visibility problem. You shouldn't have tools in the org that other people don't have access to unless there's sensitive information that they shouldn't see. Everyone across these teams should have access to all the different tools regardless. They use them daily. So I would say from a visibility standpoint, you obviously have to solve that first. If you're either missing data or missing access to data, that for me is like a T0 problem. From their speed, this is by the way the biggest organizational drain that I see. It's what I just kind of went on a micro rant about. You have different tools that are pointing you in slightly different directions. It's like going on a boating journey and like you're one degree off. Not a big deal at the beginning, but like that's the difference between ending up in Panama or Cuba if you're leaving from Miami. Like that's a big, big, big difference. And I think teams are dragged down in speed because what ends up happening is you end up spending weeks litigating the source of truth and then by the time you do that, you kind of forget why you embarked on this conversation anyways. And like you stop working on the problem or it becomes less of a priority. And then culturally I think ideally you unify everyone under a single team. I think in the absence of that it is pretty challenging. But with that said, if you have a good trust between the teams and an incentive alignment, I have seen that work as well because obviously in some organizations at a certain scale you can't unify that. But yeah, it's not too different from what I just mentioned, but that's my take.

Speaker A: Okay, cool. Well, what about you Rohit, what would you say? So it was one for visibility, one for speed of action and then one for team alignment slash culture.

Speaker C: Uh, I agree with Kalyn on the visibility piece. It's not about getting more data, it's about figuring out how to efficiently and effectively action on the data. So actually I'll just talk about Nuibu and personally it has never been easy try to even coach people to use different tools. Even simple thing as session replace, it's not easy to watch those session replace. So it's not about whether you are able to do track errors or check uh, your funnel performance or session replace. It's how you connect the dots is the key. Wherever seen, slowness in execution is down to slowness in decision making. I think that's a really important piece where it's all about the leadership, what culture they have created culture around risk taking, sifting it around, risk taking mechanisms, process tools, support, like they all come in together to really help teams move faster. And then that concept of one way door, two way door where uh, it's a bigger decision, like you take more time. The way I like to talk about this. It's a one way dough. Try to break it out into multiple two way decisions and still be able to figure out a way to move faster.

Speaker A: Perfect. Well, thanks to you both. Kaylin Rohit thank you for joining us. Thank you for sharing all of your thoughts, your experiences. We've covered a lot of ground today. I think. So thank you both again to our audience. We hope you took a lot away from today's episode. For further information on what we've discussed, please head over to www.noivu.com. it's spelled N O I B U dot com. We'll be back next week with another episode in our podcast series. Until then, make sure you subscribe to this podcast on all major platforms and follow the conversation on our socials at EM M360Tech, on X and LinkedIn. And for more great daily content, head on over to EM360 Tech.

Speaker D: The e commerce toolbox, AI and retail is brought to you by Noibu. To find out more about Noibu and how we unify error monitoring, site performance and experience analytics to uncover growth opportunities and skyrocket your revenue, visit www.noibu.com. that's N-O-I-B u.com and then make sure to search for the E commerce, toolbox, AI and retail 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.