Olly: Why fundamentals beat customization in ecommerce and AI

[00:00:00] Jennifer Peters: At the end of the day, the most important thing for a customer is to be able to find the thing that they want and to be able to check out. And I think sometimes in the push to customization and specialness, we block that by doing different things. You could have the most beautiful site in the whole world, but if a customer can't figure out how to add to cart, it does not matter how beautiful your site is.

[00:00:22] Kailin Noivo: Great episode coming up this week with Jennifer. Really cool to see the different opinions on do you customize, do you not customize, how much customization is too much, seeing it both from the merchant side and from the agency side. So, really good episode this week.

[00:00:36] Kailin Noivo: Welcome to another episode of The Ecommerce Toolbox Expert Perspectives. Joining us today, we have Jennifer Peters, who's the director of D2C, MarTech, and Digital Compliance at OLLY. Welcome, Jennifer.

[00:00:48] Jennifer Peters: And thank you for having me.

[00:00:49] Kailin Noivo: Awesome. Jennifer, we always like to start off with a bit about your history and your career journey. So, maybe talk us through your career journey so far.

[00:00:57] Jennifer Peters: Sure. Yeah. I got my start in real retail at Barnes & Noble back in the day, selling real things to real people in a real store. So, that's kind of exciting. I mean, it's changed. I think retail stores have changed a lot in some ways and not in others. But I made the transition over into digital, and it's really almost the same thing. You're just, the store looks a little bit different. So, I spent some time as a director of marketing and ecommerce at a church supply company in Nashville and then pivoted out to the Bay Area and have been here since. And I've done a couple of different things here. I've worked at an IoT company. I've worked at an agency. I am a Shopify developer, so I did work at an agency before I came and landed here at OLLY. And I've been at OLLY just a little bit over 5 years.

[00:01:42] Kailin Noivo: Very cool. Sounds like you've seen maybe some of the different kinds of trends in ecom, from "Hey, well, we have our own bespoke brand, so we need to customize everything" to "Hey, your product's not your actual website." So maybe talk us through your philosophy on that. Where did you, kind of, ultimately land from how you position your company online?

[00:02:01] Jennifer Peters: That's such a great question because I think every brand in their heart believes that their brand is so unique and so special and so different than other brands. But the reality is, when you're selling things in an ecommerce setting, everybody is pretty much the same. Unless you have some kind of wildly personalized product or just, like, over a million SKUs, we're all really doing the same thing. As a Shopify developer, I mean, obviously, Shopify is very templated, as are most platforms, and so there's so much that happens that's templated. But at the end of the day, the most important thing for a customer is to be able to find the thing that they want and to be able to check out. And I think sometimes in the push to customization and specialness, we block that by doing different things. You could have the most beautiful site in the whole world, but if a customer can't figure out how to add to cart, it does not matter how beautiful your site is. One of the brands that I worked with a couple of years ago was replatforming off of a, I would say, like, a legacy platform, and their top request was that it just needs to work all the time. And I really was just kind of blown away by that, but it's actually not that uncommon for brands to be in a situation where sometimes their cart fails, and sometimes it doesn't collect taxes. And, you know, the other things happen where there are bugs, and it's just the most basic element of a customer being able to check out is not happening. So, for me, web development, building an ecommerce store, building an ecommerce experience starts with the most basic: getting a customer to your site, getting them to the product, getting it in the cart, and getting to checkout. Everything else is gravy at that point. I think that the focus has got to be on the basics, and then you can add more on top of it. But I think the beauty of Shopify is, as well, like, you can customize anything, you can build anything. The question is: should you? And that's something I feel like I learned early on, like, you can do anything. Anything can be done. The question is, should you? And, you know, especially as you're allocating resources, you have to really think about, like, is this the best choice of my resources? This means I won't be doing this because I'm doing this. Which thing is the most valuable?

[00:04:11] Kailin Noivo: Okay. I love this thread. So, there was a big debate on build versus buy, right? I think Shopify and the other platforms kind of settled that. There seems to be now a kind of like customization versus anti-customization kind of argument, slash threat of thinking coming especially from the agency side. Is your argument out of curiosity that, like, there becomes a point of diminishing return on customization? Or how are you, like, balancing that? Like, what's kind of the threshold there, would you say?

[00:04:38] Jennifer Peters: Like, there are a couple of things. One, you're going to lose some of the fundamental benefits of being on Shopify if you over-customize. Like, you could plug in an app and think that you're going to do something, and it's going to be great, and then it just doesn't work because of customizations that you may have made. It's harder to maintain, especially if there's a handoff to another team or when you onboard a customer, I think, for the first time and you kind of look at their code and you go through their site, you can find all kinds of stuff that you're like, "What on earth has happened here?" And one of the things I see the most is brands that had someone who wasn't a Shopify developer working on their store. And so, you can tell this right away because everything is done in CSS. And that is one of, like, the gnarliest things to have to unravel from the beginning because, you know, you change one thing, you think it's going to do one thing, it actually does three things. So that, I think, is something I've seen a lot. And using CSS to, like, make a word red or, like, underline something, fine. That's fine. But when you're doing real development that should be done in a different way in CSS, you're kind of setting yourself up for future drama around making just, like, basic changes that you should be able to make. The other piece of that is really just like the cost of ownership. The amount of resources you have to maintain something that is incredibly custom if you're not with an agency, like, say, you're just a brand. That total cost of ownership is wild. I mean, you need to have developers on staff. You need to have people who are experts in compliance. You need to have people who are experts in tax collection. It's a lot to carry, and most brands don't really want to resource that. They might love the way a headless site sounds and how cool it sounds and all the things that they'll be able to do. But maintaining that is really complicated, and it really is quite a slog once you get into it. And then, really, you're trapped. You're trapped with it.

[00:06:30] Kailin Noivo: Do you think AI potentially flips any of this on its head?

[00:06:33] Jennifer Peters: Yes. I was going to say, I'm glad you brought that up. I do think there are some elements of AI now. Like, now, within the last couple of weeks, even, where you could build a store, you could run a store, you could run a whole business without ever even signing into the CMS. I mean, and that's incredible. I don't know what that's going to mean yet. I mean, the things I've seen and tried and tested so far are insane and just kind of, like, boggle the mind and, like, everything built around the logic of the way we do this and the way we think it works. It's very early days, though. So, I don't know. The other piece of that is that there are going to be brands who are going to be on that cutting edge, who are going to be, like, running everything with an AI agent, you know, and their own computer. But the majority of brands are probably not going to be there. So, I think it's going to be a slower transition than we think. I mean, you can look around right now and see brands, big brands that are still on, like, legacy platforms. So, I don't know what that's going to look like in 6 months. I feel like it's changed a lot in the last month. So, it's exciting. I think it's a great time to be in this business because you just don't really know what's going to come next. But I do think it's going to change. It's going to change a lot of things.

[00:07:47] 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.

[00:08:01] Kailin Noivo: It's interesting because, like, if you look at the foundational models, what are they best at doing? And there's, like, I think, a 70% penetration in coding. So, it's, like, the best in coding. And then when you look at, like, complex, noncomplex coding, it's really, really good at noncomplex coding. And then when you look at websites, they're largely noncomplex code. It's a lot of, like, front-end. So, like, I don't think people are going to build their own infra. Like, I don't know why you would try and solve that problem again, but the idea of, like, it evaporating the TCO of headless is actually kind of interesting.

[00:08:33] Jennifer Peters: It really is. It's very compelling, and I think that there are going to be certain types of developers who are going to love it and embrace it. And just like the advent of Shopify, there are going to be some developers who are completely fearful and will hide under a rock to avoid it. And that's always an interesting thing because I feel like a lot of front-end engineers that I've met over the years are not usually early adopters when it comes to new code or new ways of doing things or things that make some of the work they might have done in the past feel less relevant. But the ones who are kind of, like, leading with curiosity and excitement are going to be the ones at the end of the day that are successful.

[00:09:15] Kailin Noivo: Yeah. I think part of the attraction of Shopify is, like, you can run it to a certain scale without a developer. And I wonder if that now becomes true for things that are a bit more flexible, not template-based with AI. But it'll be interesting. I'm curious. I think you're right. I think it's, like, changing week over week. And by the time this podcast comes out, it might be actually clearer. But what I've observed is that we have customers that historically were on the edge of needing a developer. Like, they were just kind of at that 10 mil GMV, 5, 10 mil GMV range. And now they have one or two ecommerce people, and they're doing most of the dev work themselves, candidly. Like, just through agents.

[00:09:58] Jennifer Peters: They can run that.

[00:09:59] Kailin Noivo: So, I don't know, it's an interesting time. That's for sure. Alright. Let's hop into it. Curious. You've seen a lot of replatforms. What were some of the high-level? These are always kind of a hot topic, maybe less so now because I think a lot of people are stopping the replatform game, but curious, like, what are kind of some of the things that you've seen people underestimate when it comes to replatforming and the disruption of a business?

[00:10:20] Jennifer Peters: Gosh. I would say, you know, there's definitely the legacy platform piece, which I'm going to kind of, like, set that aside for right now because I think that's pretty obvious. I know a lot of brands that, 3, 4 years ago, all decided they wanted to be headless. And then if you look back at them now, they have all reversed that decision. So, I think that's a really interesting thing because a lot of times, agencies will sort of sell you this idea of how cool it's going to be, how beautiful it's going to be. It's going to be so different. It's going to be so interesting, but they're not really fully informing you on what it's going to take to maintain that day-to-day. So, I think, once that realization hits, there's suddenly a lot of backpedaling on how cool we think headless is. So, I feel like I have seen that a lot. The other piece is integrating with the rest of your systems. If you have an ecommerce platform that's, like, so complicated and customized and it's just sort of sitting in this space, integrating those is a lot harder. If you're integrating it into a CDP or your ERP or your CRM or, you know, any kind of dashboard, any kind of revenue results, KPI dashboard, suddenly, it requires, like, an engineer to do that and not just someone who can sort of plug and play and get things moving. And that's another complexity that you get when you're sort of, you know, off the beaten path of the accepted and common platforms.

[00:11:40] Kailin Noivo: Alright. Let's jump into some more D2C things. Talk to us a bit about, for those that don't know, what's OLLY, high level?

[00:11:48] Jennifer Peters: So, OLLY, we're a supplement company. If you don't know OLLY, if you have ever walked into Target and turned right, we are always on the first end cap. So, that's usually, the people are like when you hear that, you're like, "Oh, yeah. I know OLLY." Normally, I have a bottle on my desk, but I do not today. We are 11 years old. We were a San Francisco startup back in the day, and we were acquired by Unilever in 2019. So, we're part of the Unilever Wellbeing Collective along with some other really cool brands, like Liquid I.V., and Onnit, and Nutrafol and SmartyPants vitamins, and we have a great little group of wellness brands. We are known for our gummy vitamins, and the fact that they taste wonderful, and they really do. They're very, very good. But we are kind of branching out a little bit into the capsule space for probiotics and some different formats like fast dissolves and soft gels. Not everyone is a gummy person, but we have about 60 SKUs, and we're constantly looking at, you know, what people are interested in. And I think, like, our primary customer is millennial women. And as millennial women are aging, their needs and desires are changing a lot, and that's been a really fascinating kind of road to follow. You know, again, we're 11 years old. So, when we first started, we were making supplements for millennial women that were, like, maybe in their late 20s, early 30s. And now, you know, they have a family, and they're getting older. Some of them are in perimenopause, and so the needs are changing. And it is really cool to be in a business that is able to change and adapt to that so quickly.

[00:13:23] Kailin Noivo: Very cool. Actually, I didn't know this. It's funny. You just mentioned two of our customers, Unilever and Onnit. So, it's actually funny. They're actually Noibu customers. But you just recently, I believe, co-presented at Analytics Unite. This is something that's really resonating with me. A lot of brands are sitting on a gold mine of consumer feedback. Some of them are on the internet. Some of it is in owned-like things like reviews, support tickets, and VOC data. A huge problem that we consistently hear is that the data is scattered in multiple areas, and it's hard to, like, unify to pull signal. Maybe talk to us a bit about that. How have you been able to address that to be able to pull signal from various channels?

[00:14:02] Jennifer Peters: We have a great tool. It's called Yogi, and that's pretty much what we use to do this. That's actually who I was co-presenting with at Analytics Unite. And it does something, it kind of solves a problem that when we first signed with them, frankly, I didn't think was solvable. And that to me is kind of the best thing that can happen to me in this business, is when you see a solution to a problem that you couldn't even clearly articulate, or you thought was completely unsolvable and then suddenly someone solved it. And it's really amazing. And they were really the first to successfully scrape and dashboard Amazon reviews, and that was the big gap for us. It was like, we have so many Amazon reviews. We need to be able to put them together and gather insights from them. So, that was really the driver initially when we started working with them. But what we found was, yes, it's great to have Amazon reviews in a dashboard by product. We love it. It's awesome. But in addition to that, we're able to integrate our CX software and be able to pull those insights into that dashboard. We're able to integrate social listening and other platforms and other places where reviews live. So, we just, like, for example, we just added Costco reviews into our package because we just got picked up by Costco last year. So, buy us at Costco. So, it's really great, and their product has evolved really quickly, too. So, it's really exciting. But I think that at the end of the day, it's a tool that is probably one of the most widely used tools amongst different departments at OLLY. So, like, the brand team uses it, the marketing team uses it, insights, ecom, product development, innovation, they all use this tool because they can go and look at, like, "Let's see what people are saying about OLLY Sleep today." And, you know, it's a specific SKU. You go, and you look at the SKU. It kind of gives you, you know, the headlines. If you want to dig a little deeper into specific topics, you totally can. You want to see how it stacks up against our competitors in that space; you can. It's really great because we do compete in different segments. You know what I mean? Like, if you look at, like, dog food, dog food is kind of like dog food. We are in gut health, and we are in kids, and we are in multis, and we are in women's wellness. And within each of these categories, we have a different set of competitors. They're not, it's not the same. It's not just like, "Oh, here are the top 10 competitors." It varies depending on what the product is. So, being able to, like, really customize that and stack them up, like, apples to apples next to each other has been hugely valuable for us. And I think I will say there's not one thing that came out of this that we probably didn't already know. But I do feel like on the CX team, they know everything. They know everything literally. They know what people love. They know what people hate. They know what makes people mad. And, you know, when we share that internally, sometimes it feels and falls anecdotally. Like, it's just like, "Oh, it's a passing comment. A customer said this." But when it's in a dashboard, and you see that 30% of your customers hate this cap on a bottle and they can't get it open, suddenly, it's an actionable item. And that's probably been our biggest benefit and our biggest point of excitement around using these tools because suddenly, you can't hide from those things. Now, it's a piece of data. It's not an anecdotal story that somebody called in and complained about this, whatever. Now, it's there. You can see it. You can see how it sits in the context of everything else, and it really helps us identify those customer pain points, internal projects, packaging, all of these things. It's been hugely helpful to us, and everyone can use it. So, there's a level of transparency around it that we also really love, and literally anyone can use it. So, if you are curious about whatever, you can log in and use it and see what's in there. There isn't, like, a gatekeeping element to it or anything. Like, it's just, it's for everyone here. It's very cool. It's been a game-changer for us.

[00:17:48] Kailin Noivo: Yeah. It's funny you say that. I was recently on, and I received, it was an internal call, and it talked about turning qualitative to quantitative.

[00:17:56] Jennifer Peters: Mhmm.

[00:17:57] Kailin Noivo: And, like, AI really helped us turn qualitative to quantitative. Like, turn anecdotes into something that's quantitative. Do you know what I mean? Like, when X comes up, Y happens. On the AI thread, because I know that's at the tip of everyone's kind of tongue right now, talk to me a bit about how you guys are thinking about discoverability. So, like, historically, it sounds like keywords, SEO, SEO on Amazon. How are you guys thinking now about, like, I don't know, GEO, some people call it. It's got the other acronym, but basically, AI in the sense…

[00:18:28] Jennifer Peters: AEO. Yes.

[00:18:29] Kailin Noivo: Exactly. So, how are you guys thinking about that?

[00:18:32] Jennifer Peters: It's another one that changes on the daily. I think it's incredibly exciting. Our D2C business is just a really tiny portion of our overall business. And so, you know, the website's kind of always, and I would say that's true of any CPG brand that has D2C. Like, the majority of our sales and revenue are coming from in-store. They're coming from Walmart. They're coming from Target, grocery, and drug. All of a sudden, this website that we've had that, you know, maybe does just a tiny little bit of business, is now the point of entry to the brand, and that is super exciting. It kind of puts a whole new spin on what we're doing on my team around direct-to-consumer. Because now it's not about conversion. It's not about revenue. It's not about all these, like, super detailed KPIs that you think about when you think about what happens on an ecommerce store. Now, it's about, like, how are people discovering our brand? How are they trying to solve these problems? What words are they using? What questions are they using? What do the queries look like? Is our site readable? Is it indexable? Does our product catalog feed have enough digital attributes that are recorded in it to be useful? Most people don't. I mean, we've been making digital catalogs and feeds for Google for, like, forever, and they don't read every attribute. We've got, in the LLM world, unlimited attributes. But at the end of the day, a lot of what we do to do good AEO, and whatever you're doing to prepare for the agentic future, a lot of it is just good SEO, and it's just good business, and it's just good web development. So, like, the foundational and fundamental things, you have to execute those well before you even start thinking about, like, "What's our Reddit strategy?" You have to execute clean schemas, and good content, and relevancy, and just a good product feed. Like, a lot of these basics are the things that will get you almost all of the way there. Then you can start playing around with all the fun tools that do different things and measure impact and all of those things. But I really, really believe that if you're doing SEO well and you have a good website that is well built, you're more than halfway there. And there's a lot of temptation to overinvest in certain tools right now, too, because you just think like, "Oh, this is so cool. This is so cool. We should buy it. We should buy it. We should buy it." And then four weeks later, something else comes out, and it's way cooler. Or the way things work pivots quickly. Like, now there's a different way of doing this, and now it's reading it differently. So, I think one of the things that we've had to do as part of our strategy is, like, be cool. Like, don't invest in everything. Let's invest in a couple of things that we think are going to go long-term, but we don't know that. And let's stick it out and just see how it goes, and let's sort of try everything and use some tools that validate each other. And then also use tools that validate the human experience because that's not going to go away. So, we don't want to be creating all this content; it's very expensive and time-consuming to create content. We don't want to create all this content that no one's ever going to look at. We want it to check both boxes. We want it to be great for LLM indexing and understandability, and we want it to be great for human interaction. So, that's kind of been our intersection of content right there. It's got to do both things for us.

[00:21:41] Kailin Noivo: How is the traffic quality from these engines that you've observed? Like, is it converting at a higher rate, normal rate? How's the traffic relative to your overall traffic as well? I have an anecdote that I'll share after.

[00:21:56] Jennifer Peters: It's so small. It's really small. But I also, there was an event last fall. It was called the Retail Club AI Summit or retreat or something like that, and it was great. It was such a great show. And you would go, and you would listen to all these speakers, and they were all kind of talking about the same thing. And every single one of them had a completely different story about what they thought was happening. And again, I think it's like, there's nobody who knows for sure. And if they tell you they do, they're lying. And if they're an agency and they tell you they can fix this for you, they're probably also lying. Like, nobody knows how this is getting measured. Nobody knows what the data sources are. I mean, I know that one of the tools we use has multiple data sources, and they're all under NDA, and they can't say that they contribute to that. Like, there's just so much unknown, and you just can't spin your wheels on some of the things. And you know, sometimes, there's a pressure to show impact. You know, it was this percentage discoverable before the changes and this percentage after. Who knows if those are real? There's no way to know.

[00:22:58] Kailin Noivo: It's funny. I was just pulling it up. One of our customers that does about 10,000,000 sessions a month, I just pulled their last 30 days or 90 days data, and so they're doing, however, X million, and they had 600 ChatGPT sessions. So it's like…

[00:23:14] Jennifer Peters: So small. But, I mean, the tide is coming, but it's not, I think that's another thing that I'm seeing a lot of in sales is, like, "You're behind. Your competition is beating you. You have to hurry. You have to hurry." Don't have to hurry. Like, Amazon did not happen overnight. Like, we all saw that coming. It's going to be okay. I mean, there are things you can do to prepare yourself for what will come, and that goes back to that doing SEO well, building a good website, and making sure your product catalog is good. You can do those things, and that's good for your business, no matter what happens. And those are the things to me that you really need to be focusing on and not trying to game the system. Like, there were dark days of SEO where people were, like, putting in keywords in white print, you know, or white text so the computers could read it and people couldn't. And we do not need to go back to that. Like, we need to just do it right.

[00:24:08] Kailin Noivo: Keyword stuffing.

[00:24:10] Jennifer Peters: Exactly. It was bad.

[00:24:11] Kailin Noivo: No. Cool. Jennifer, as we look to wrap up, is there anything else you want to touch on?

[00:24:14] Jennifer Peters: I don't think so. You know, this is such an exciting time to be in this business. It's changing so fast. I would say don't be afraid of AI, and don't let other people scare you with AI. Be curious. Go out there and, you know, ask Claude questions. If you want to learn, that's a great place to start. You know, "What do I need to know? What do I need to do?" And face it with a sense of curiosity instead of a sense of fear would be my advice.

[00:24:40] Kailin Noivo: I love it. Very San Francisco of you. Thanks so much for the time, Jennifer.

[00:24:45] Jennifer Peters: Yeah. Of course.

[00:24:46] Kailin Noivo: It was a great episode.

[00:24:47] Outro: 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 n-o-i-b-u.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.