213. Your Customer Data Is Lying To You ft. Deneke O'Reilly

Speaker A: The question sounds simple until they try to actually answer it and they ask their marketing team, do we know this?

Speaker B: Hello listeners and viewers on all platform. Welcome to another episode of Commerce Untold. Um, today we have Denique O'Reilly, founder and principal of Relic Strategy. Hey Denique, welcome.

Speaker A: Hey, good morning Ethan. Nice to be here. Thanks for having me.

Speaker B: Sure, sure. I mean it's uh, we've been planning this for, for a while. Super excited you know, to talk about um, I mean your expertise and your contribution to the industry specifically around the retail space and omnichannel brand. Right. But I ah, think before we kick it off maybe just give us a quick intro of what do you do in Redix strategy. I mean who is your ideal customer and what type of services do you provide?

Speaker A: Sure, sure. Uh, yep, recently kicked off consultancy uh, here at Relic Strategy. Trying to apply my 20 years of customer um, data experience. And over that time frame I generally spent my time converting customer insights into uh, incremental revenue. Um, uh, at Relic we are focused on what I'm calling trying to help retailers build durable growth. And that's growth that kind of compounds over time because you've built a strong foundation based on kind of three areas of focus which I think are all connected and it's core to any business. Those areas are identification, activation and attribution. And I could talk more about that later.

Speaker B: Yeah, for sure. Uh, it's like the so fundamental. Right. And each one of them is like as difficult as it gets. Right. I mean it's totally complicated in our era where you have, I mean customers journeys are so complex. Right. I mean it's all over the place. Uh, nevertheless, I mean if you have also an omnichannel kind of experience. So yeah, I'm in. The attribution part is obviously one of the challenges that I mean anyone in the industry is obviously facing. Right. And you mentioned data, uh, like definitely as a cornerstone and as a source. But I think we're in an era where there is like too much data. Right. I mean each platform provide its own siloed kind of insights, uh, and analytics. And obviously when retailers with such a complex workflows, uh, architectures and, and system. How do you usually approach analyzing, you know, first phase one just what's really happening and how your process works. Trying to get some more clarity and probably provide some recommendations throughout this process.

Speaker A: Absolutely. And you're right, journeys are the most complex they've ever been. Uh, uh, it's always been a challenge to attribute what part of your marketing uh, is working Dating back to the famous quote of you know, you know, 50% of your marketing is working, you just don't know which 50% is driving it. And that got much harder with uh, the Internet and digital because now we have you know, multiple devices, multiple uh, points of pii or first party data from emails, phone numbers, cookies and so on. And after that you can disconnect from the Internet and go in person. So we still have our physical bodies as well and we like to shop. Shopping is still dominant in person. So you lose that continuity, you lose that identity. And if you are not strong in identifying who your customers are based on the information, data you know about them, it's going to be difficult to reach uh, them and engage them. And then it's going to be really difficult to understand um, who drove that conversion no matter what channel you're in. So um, kind of those three phases I mentioned are fundamental to uh, what you're trying to build and what, and if you're looking to build something that, that continues to grow and scales and you can optimize over time and I'm separating it from let's say just generating campaigns or testing new advertising channels, things of that nature. And they all kind of point back to the identification side of things. For example, if uh, my name's pretty unique but um, if you have multiple deniques in a platform you might, well you may not want multiple Denise. But which one is performing the behavior online, which ones purchase going into store? Um, how do you know it's all the same person? And ah, a more common name would, it would be more difficult. You may have multiple identities in that platform which makes it very hard to um, attribute performance, personalized campaigns and figure out which one is driving actual business income you want which is typically profitability.

Speaker B: Yes.

Speaker A: So when I, I, there are different ways of approaching it. Um, typically I start with an audit. But you're really trying to answer a few basic questions. Who your custom. Who are your, who are your, which, which one of your customers is driving the most profitability or margin and where are they coming from?

Speaker B: Mhm.

Speaker A: So you know where to get more of them.

Speaker B: Yes.

Speaker A: And if you have difficulty answering that question, uh, it's a good, it's a sign you might have uh, a data issue.

Speaker B: Yes. And what, what uh, from your experience, I mean customers know to answer this question or they are confused.

Speaker A: The question sounds simple until they try to actually answer it and they ask their marketing team, do we notice?

Speaker B: I'll get back to you.

Speaker A: Right, right. And then they'll Come back and say, all right, we may need some help. And it depends on your maturity levels, company size and platforms. I've been fortunate enough to work with many platforms over my career. Um, and I got started in email marketing, building targeted campaigns. So that's where I got started. So I was used to seeing moving data, handling data. And this is back in the days using Microsoft Access, managing, uh, data coding campaigns by hand. Obviously no AI limited, no journeys at that point, no journey builders at that, that point. And trying to match data up to what they purchased. So trying to personalize campaigns. But, um, the platforms have come a long way. So depending on your maturity, there are different paths you can take to try and resolve the issue. But it comes down to, uh, reducing the quantity of anonymous behavior occurring at your properties in person and online.

Speaker B: Mhm.

Speaker A: The more you do that, the more dependable your customer data and customer identities can be or will be.

Speaker B: Makes sense. Makes sense. So Danik, before we dive in further, what are the major challenges that you hear retailers are facing these days?

Speaker A: Um, you kind of alluded to it earlier, but they have more data than they know what to do with.

Speaker B: Mhm.

Speaker A: Um, it's whether that data can help them answer pressing questions like the one I posed earlier.

Speaker B: Mm.

Speaker A: And messy data or less accurate data sometimes will lead to sometimes lower engagement rates across your marketing channels.

Speaker B: Mhm.

Speaker A: Or not being able to answer which campaign, uh, drove this customer into the store? Basically attribution, which, which campaign should get this credit which tells you where should you invest your money to get more of those customers? Uh, so those are some of the challenges retailers are facing. And obviously as we become more privacy centric, people browse more anonymously hiding IPs. Nothing against privacy. It just makes it more difficult for you to, on the retailer side, to, um, giving the customers what they want.

Speaker B: Yes, yes.

Speaker A: Because you need to capture those signals.

Speaker B: Yeah. And this is true for offline journeys and also online journeys. And nevertheless, when they are combined, journeys, which we understand that most of journeys are today, are kind of a mixed online and offline, Right?

Speaker A: Yes. And it could go back and forth.

Speaker B: Yeah.

Speaker A: And it goes between platforms. Online, offline, it may start online.

Speaker B: Yeah.

Speaker A: Finish offline, or vice versa. And um, there are attribution. So you find some attribution techniques within, uh, Google Analytics, but typically most folks don't have all of their data in Google Analytics. So where the, where in the industry has been heading is to put all your data into one day, one platform. Right. Um, first it started off, you know, with like leveraging DMPs. Then you would have your internal marketing database with your marketing service provider. But then that was slow. Then came the customer data platform, the CDP M. I've worked with a few and I saw that industry mature. Um, but they're not all created equal. So there are many claims out there and um, some handled the duplication well or identity resolution to create what we, you know, call that ultimate golden record kind of one identity, all the behaviors and events and then others struggle to do that.

Speaker B: Yes.

Speaker A: But they might be good in uh, activation, activating on that data. So depending on your situation, you may need a two, uh, partner approach or uh, you can live with one customer data platform. And it depends if you're omnichannel or purely online.

Speaker B: Yeah, I mean that's, that's definitely, you know, a very important, uh, a very important issue. I think, you know, there is probably an overwhelming amount of tools and platform that is being used like all over the, all over the place. But both for the, you know, the backend side of things with logistics and inventory planning and obviously customer data on the marketing side and customer support and tickets. And I think today we have more tools probably to analyze the data right. Through AI and agents. I mean some of this data is just huge in terms of quantity. And those agents, I mean the risk is that they will not really go through the entire database. Right. Because they are also overwhelmed with the amount of data out there. But uh, uh, I'm sure this is something that helps you in the way you work and uh, probably implementing these tools uh, throughout your process. Correct?

Speaker A: Uh, they do. And there is more tools than ever. I'm sure you've seen that poster, uh, that used to show the marketing tech stack and this is probably 10 years old now. Used to be on the wall of uh, executives. I could imagine what that looks like now. Um, but at the end of the day you need to prioritize what's most important to you and found a platform that matches that because any implementation there will be pros and cons to the implementation.

Speaker B: Yes, I know we have like a SAS fatigue. Right. And I have a new term. It's uh, Sasmagadon. Right. So we are all like in a new era of uh, trying to figure out things and stop clicking buttons and just looking for solutions. Right. Like show me the bottleneck. Right. That's people are interested. Definitely. Uh, so a lot of things has changed. Right. Uh, and you mentioned previously Denique, about uh, incremental growth. I know you also use like durable. Durable growth. Right. As a term that you use quite often. Can you define what do you mean exactly by these two?

Speaker A: Sure. The, the way I, I view it is to build durable growth kind of that withstands the test of time. Right. That, that as you layer more, uh, understanding intelligence on it, it continues to grow. Um, BAS is with identification and that is, it's dependent on improving how customers are identified, improving your personalization and how you're activating on that data. You have mhm. And then your attribution will have its best shot in trying to quantify, um, what's driving the business, what worked from your marketing campaigns. And then you can use that and restart the cycle and invest that back into the marketing channels that work. But identification is not just, you know, I knew it was Danique that came to the website. It's obviously reducing, it includes that. But it's also what is Denique buying? Where did he come from? What are his interests? Uh, and what's his value rolls up under identification. A clear understanding of that will help you reduce wasted ad expense. So you can focus on kind of the, you know, the 20% that drives the 80% of your business. Kind of the 80, 20. Um, you can then invest in the, in the channels that are more, are most, uh, responsive to that customer segment that you care about the most. And this still depends on knowing that it was not ID 123ABC and that it was Denick O'Reilly that, that visited, uh, your site or purchased in your store. So the three areas are connected. But in my mind, all roads lead back to knowing who your customers are. And if we date back, you know, thousands of millennia, millennia ago or thousands of years ago, when people walked into the business, uh, or the store or the shop or the market, uh, they knew who their customer, who their customer was at that point.

Speaker B: Yes.

Speaker A: And it helped them cater the offering or the offer to that person. So it's peeling that glitch and going back to the basics. But it's more difficult, as we discussed, with the complexity with touch points, devices and anonymity.

Speaker B: Yeah. So assuming we've identified who are our, uh, customers or even, even who are our most lucrative customers. Right. The highest profitable customers, what's next in the process?

Speaker A: Let's uh, say you have that understanding, uh, you build out the right journeys that delight that person at each step in the journey. To put it simply, you break it down across the journey map and then you're going to want to build a campaign for, let's say, uh, that drives awareness for that user. And this is where Attribution is key too because then you want to look back like how did I get these folks? How did they hear about us? How uh, were they made aware of us? To kind of almost have uh, to look backwards before you can plan forwards. And once you map out the journey, build personalized campaigns and then obviously you're managing, you know, then that's where AI comes in with the right content, the right time. Um, and then you're managing frequency and building trust with that person through the journey and then you're moving into the stage of attribution. Yeah, but you have the ability at every click and every visit there's a chance you lose the connection of who actually was, uh, it the person I emailed that actually came in and bought. So building that structure behind the scenes is going to be important. And that's required tracking using different techniques, whether it's utms and coupons and uh, whatever your makeup is. But you want to develop strong trackability along the way.

Speaker B: Yes. And on online side of things, I think it's clear with utms and coupons that you mentioned, maybe just, you know, with today with choppable video, I mean you can attribute. So I mean that's fine. What are some of the techniques that you use? Let's say in an online to offline experience, someone just discovered a product online, but mainly purchased, uh, you know, at the end of the day. Right. You may have purchased, let's say in an offline store, like a big Morris store. What are some of the techniques that you can follow that, that unique user?

Speaker A: Um, there are a few. Uh, maybe the, the simplest one to explain is how good are you at uh, capturing that user's information at the point of sale? That's, that's key. And there are techniques to optimize that or amplify that. But you, you want to know who's, who's converting at the point of sale. Especially if not via, you know, beacons or along the way or apps and things like that also help. But the key is uh, how are you asking and how are you capturing it? Yeah, are you offering something at point of sale? And I don't mean and the rate at which you do that is what's important. Are you capturing 10%? It's not going to be good enough, is it?

Speaker B: Yes.

Speaker A: 30%. It's not going to be good enough because you've already, uh, because you have 30% and then if you're identification online is less than 20% then you're going to be down to a percentage Of a percent.

Speaker B: Yes.

Speaker A: And that's the optimization path you're going to be on. But it requires breaking it down into each journey, each step and plugging in the holes, fill in the gaps. And obviously taking care of the customer is a given.

Speaker B: Yeah, obviously. I mean just uh, probably exposing a coupon code online that you can take and buy, you know, offline with that coupon code then you can close the loop. Right. Obviously for, for this, for that transaction and then you understand this is the segment. So let's create some journeys to support that. So you mentioned a journey of existing customers. Let's say we've identified our top segment, top top customers, kind of profitability or sales and create variety of journeys to support them. What happens for like prospects or leads before they purchase? It's all about trying to capture some information as we possibly can around them and develop follow up journeys. Right. That, that, that serves you know, a variety of use cases to them.

Speaker A: Yeah. And there are many paths here. You can go down the path of choosing to do only um, acquisition campaigns where you capture that person's information. So let's say lead forms and limit. If the, if the marketing vehicle or activity does not allow you to understand who's signing up, let's say it's just a banner click, uh, then you avoid it or you put something along the path to help capture. It's called progressive profiling or some um, data capture along the journey to help you with the attribution. Yeah but it requires a certain amount of discipline to make sure that is set up. You almost, you have to have the measurement understood and locked in before you launch and before it's agreed to.

Speaker B: Yes, yes. Interesting. So the nick in relic, you also uh, implement those type of processes and workflows or you just consult. What's the business model for you guys?

Speaker A: Uh, we do both depending on the scope and the needs for that retailer. We have a team and we have partners that we've worked with um, over the years or I've worked with personally that understand the challenges deeply because they've lived it or I've lived it with them. Um, as I said, I started out with just Target emails but I ended up uh, at VP of marketing operations and managing very large, very large customer databases from first party, second party, third party, um, sources with partners and um, routine updates. So I've been in the trenches so I can, I'll be in the trenches with the partner in building the solution and helping with manage uh, the implementation. So we do both.

Speaker B: Yeah. And it's Always, uh, the best position, right? You bring your two decades of experience in that specific niche and you just service, you know, provide value to customers. That, that's a great, great, uh, uh, next step. So what drove you to, let's talk about your entrepreneurial, you know, side for, for a few minutes. I mean, what drove you to launch Relic? I mean what excites you about it? What are some of the challenges you're facing?

Speaker A: You know, I mean I, I felt the time was right in my career, um, and I, I started to, I, I see the same challenge over and over and I've personally worked on increasing identification rates, that specific problem for a very long time. And um, we've uh, achieved, I achieved success in that area and I felt other businesses have that challenge and if they are not, if they're not investing the time and energy in that space, uh, growth is going to be more challenging and it's going to be, they're going to ride with the seasonality wave and trends and political changes and things like that versus having a solid foundation to con, uh, bank on. Uh, personally I've always been entrepreneurial. I like to try and do different things. Um, and this is the first endeavor into leveraging something new. What I've been doing for the longest time, my, my strongest skill set and I hope to help retailers, um, even smaller retailers, you know, apply some of this methodology, uh, to their business. At the end of the day we're trying to help help them uh, succeed and help their customers, uh, be promoters. And um, maybe it's the working in email for so long and writing so many subject lines really I, I feel like I have a pulse on a customer. I still feel I have the edge on AI on subject lines. I know I, I need to do a nice little test here, but it's. You have a quick knack. You know when you build something and you feel like, oh, I can quickly tell which test is going to win and um, or which subject line is going to win. So maybe it's that background of uh, writing subject lines, seeing results thousands of times. Wow.

Speaker B: Yeah, that's quite unique. So I think the pain is clear. Right. We talked a lot about that from your customer's perspective and obviously you answer that pain and probably presenting some kind of a transformation or process for improvement. What are some of the, how do you define success? Right. What are some of the KPIs that are relevant to measure here?

Speaker A: Uh, the big one is um, potentially a uh, higher revenue per send or revenue per person reached. Mhm. Transaction or conversion rate from your campaigns. Reduction in duplication, but uh, the driving revenue, incremental revenue per person in the audience. If it's email per email sent for example.

Speaker B: Yeah. Well this is always like very easy to measure. Right. About, very hard I guess to achieve. Right. But sometime the gaps are quite visible that it's very fast, you know, to create some meaningful improvements. Right. When you work with customers and obviously uh, these are the things that you identify quite uh, you know, in your process and try to figure out, you know, to provide some recommendation how to improve things. And uh, the results are sometimes visible in the short term and the midterm. What is usually the time it takes to see some transformation here?

Speaker A: Yeah, um, so let's say we have the system implemented and we're monitoring identification rates, duplication rates, um, you'll start to see, and one to three months you should start seeing a change in engagement across your campaigns. Whether it's a reduction in opt outs, uh, or an increase in revenue uh, per individual. And if it amplifies, if you create more journeys or more, more touches or integrate different channels. So it's a email, sms and if they don't respond there, you send a targeted direct mail piece later on. Um, and after the attribution you can, you use that to reinvest in that channels and you're trying to make this as automated as possible along the way obviously. So what you built before is generally hands off outside of maintenance or improvements and they're already prioritized. But you're reinvesting into this engine over and over as you gather more data and the data you gather, the, the attribution to the sources that are working will continue to improve as you can identify more along the way.

Speaker B: Yes, yes.

Speaker A: So lists should grow, engagement should increase, disengagement should decrease and um, and then revenue uh, should follow.

Speaker B: Yeah. And all of these are really uh, like clear parameters. Easy, easy, easy to measure. Right. It's not something vague or it's going back to the data, uh, so the niche.

Speaker A: But it's less theoretical.

Speaker B: Yeah, exactly, exactly. Less theoretical. And you mentioned a lot of like uh, the way you work and the process and it's data driven, trying to find out, find you know, segments of customer types both on the retention side or the acquisition side. Try to implement variety of workflows and journeys. But there is also like the content piece which is very, very important.

Speaker A: Right.

Speaker B: I mean what do you write on those email scripts? I mean the SMSs, how do you craft those messages in a very intelligent way? Tell us more about that.

Speaker A: Sure. Uh, so the content should be based on what you know about the individual which comes back to obviously identification and how well you're capturing data. Is it real time data, is it accurate and does it actually belong to that person? So let's assume that's, that's resolved then the content you're, you're choosing, I like to base it on you know, growing the business thinking from a growth perspective. There you can come up with a list of automations or triggers. Let's say you prioritize them based on the impact which is typically a combination of you know, how strong the offer would be for that campaign and the scale, the scale of that campaign. Let's say you start there and depending on that campaign type, whether it's a, you know, as simple as an abandoned cart or a back in stock campaign, um, uh, you, you build the content to of course basics. At this point you want to have a strong hook, strong offer up to you. If you, you don't always need an offer if you're really relevant and just relevant. Relevance been the name. Relevancy has been the name of the game for I don't know, maybe 50 years. Personalization is another form of relevancy. You're just trying to increase the percentage of relevance or increase the relevance across your campaign as much as possible without being uh, obtrusive or creepy let's say. Yeah, but you're trying to solve that person's problem through relevance by showing them uh, by by wowing them or solving a problem for them in every campaign. So content you can uh, is easier than it used to be but you always want to monitor that AI and make sure it's, it's still driving the relevance. Um, understands the context which is obviously key in any generative um LLM. So make sure the context is key. Uh, it helps you write it and then you still need the human involved because I still think we have a good edge on AI in terms of um, understanding what the person really wants to solve.

Speaker B: Yeah, there is the strategic and the thinking stage. Right. Or layer and then there is the execution layer. I think we need to you know, train ourselves to be there on the strategic layer and making hard decisions and let you know AI to the grind like the repetitive task and the execution

Speaker A: and more on the execution side. Your data catalog, your product catalogs, um, all need to be well structured and available and real time. So for example if a price changes on the website at 1pm but you're only uh, or price changes two times throughout the day on a product, but you're only updating the product data clog once a day, you're going to miss a price change and then that impacts when you can send the campaign. If you send it before the price change or after the price change, you can very easily, uh, scale a problem and create customer complaints and customer service issues. So you want to stay connected to, um, the customer in terms of opt outs and what customer service is telling you. But it's not only what the customer sees. Um, it's also what's built behind the scenes and setting up the, from the titles, descriptions, um, colors, prices. Do you want to show prices in the campaign requires strong discipline, but it's riskier. Or do you want to say something? Click, uh, for best price or click to see today's price. You know, you make those decisions along the way for every element in the campaign.

Speaker B: Yeah, that, that makes total sense. Uh, great. Denique. Anything, uh, else? Uh, first tell us how you know people. Companies can reach out to you and to Relic.

Speaker A: Sure. Um, appreciate that. Uh, go to Relic R E l I q strategy.com and you can schedule time with me there or send me a message. Um, no, I think, I think I've, uh, said maybe too much. I talked a lot here. But, uh, I appreciate the time and the opportunity to kind of share my story and what I'm working on.

Speaker B: No, I mean, it's a very, very important topic. It's one of the most, uh, important topic for any marketeer out there or anyone who's just, uh, running this type of businesses. Uh, anything, uh, else you want to add?

Speaker A: No, uh, I, I think we're all set here. You know, the three things I leave you with is identification, activation and attribution.

Speaker B: 100%. Yeah. I will never forget that for sure.

Speaker A: All right, thank you.

Speaker B: All right, David, thank you so much. It was a pleasure hosting you here.

Speaker A: All right, it's been a pleasure. Take care.