Personalization in 2025: The 5-Level Ladder from “Hi" to Predictive Service - Commerce Today Episode 126
===

Joshua: [00:00:00] Hi everybody. It's Joshua Warren again, CEO of the E-Commerce Agency, Creatuity and Host Commerce Today. Excited to chat with you some about e-commerce and today's episode's a good one. We're gonna talk about personalization, which we haven't actually talked about in a while. It's funny that there's so much talk of AI and so many people looking at that, that I feel like the basics of personalization are getting forgotten. And good example of that. And the way I wanna frame this is are you successfully personalizing your site or are you just creeping out your users? And if you think about it, think about, Netflix, Disney Plus, really any of the streaming services.

But Netflix was first and foremost at this. And now even YouTube. Think about how great their recommendations are, how they learn, what you like to watch, what you don't like to watch. Netflix, it used to be the. Thumbs up, thumbs down. Now that isn't even as big of a factor in their algorithm. It's literally just, or do you watch the whole show?

Do you binge watch this show all the data they collect and they're collecting a lot of data [00:01:00] they use to personalize your experience. And it's a good experience. Like you never stop and think, oh wow, Netflix is like really watching me. You just enjoy watching your Netflix. Flip that around to, I'm sure all of you have had an experience like this.

When we were moving into the new Creatuity office, I was shopping for some floor lamps, and I finally found one that I thought would work well in a particular spot in my office, and so ordered a floor lamp. Then the next day I'm browsing around the internet and I keep getting retargeted and seeing ads for that same floor lamp.

And then two days later, three days later, a week later, a month later, literally everywhere I go on the internet, I'm seeing an ad from that company with that floor lamp. And I'm thinking, I already bought your lamp. This is frustrating. It's a little creepy. Why is your lamp following me everywhere I go, this is not helpful at all.

And I want you to ask yourself, all the e-commerce merchants out there, e-commerce directors, marketing directors, which side of the line do you fall on? Are you closer to that great Netflix experience or are you closer to that [00:02:00] creepy lamp experience? And how much money could you possibly be losing if you're on the wrong side of that line?

So personalization is still very important in 2025. Again, everybody's excited about. AI and about all sorts of new platform features and things like that. And I feel like we're not talking about personalization and personalized marketing quite as much as we used to. I know that a lot of us got tired of it a few years back when it first came onto the scene, but it's big.

And one of the reasons it's big is especially the mid-market brands that I'm working with they can't outspend Amazon. Like you can't, whether it's on your platform, whether it's on your marketing. Your tech stack, you really can't outspend Amazon, but you can out personalize them. So when you're competing against Amazon for those customers and that customer's attention, you can niche down and you can personalize using your brand, your niche, your data.

And so that's why it's so important to get personalization, right? Few things. [00:03:00] From some recent studies. So if you get personalization right, your A OV typically increases by at least 10%. You can get your repeat purchase rate up at least 15%. Even, and this is the stat that surprised me, support tickets drop eight.

Percent when solid personalization is live. And we'll talk about that later in the episode of how even your support ticket customer service burden can decrease when you get personalization, right? Also, if you're a B2B brand, don't tune out. Personalization is just important, if not more so at times.

For a B2B brand as it is a B2C, I've been saying this for years, but your buyers, especially as the buyers shift to a younger generation, your buyers expect just as nice of an experience shopping on your B2B site as they get on B2C sites, as they get when they're shopping on Amazon for their personal purchases.

So I believe that personalization is, I look at it as a five step ladder. And you [00:04:00] gotta start at the bottom and work your way up. These five steps. So level one, step one of this ladder is recognition. So this is the really basic stuff. This is the stuff that hopefully your platform has built in and you just need to turn it on if you haven't yet.

This is greeting people by name if they're signed into their account, ideally, even if they're not, if you have a cookie set. And then showing them, say the last. Products they viewed so that they can see they, they understand that you recognize them, you are treating them as a specific customer and not just one of millions that are visiting their site.

Some quick wins here if you haven't already. Add merge tags to your email and text message flows if you don't know what that is. Merge tags. It's basically how when you get that email that says. Hi Joshua, instead of, hi, that's a merge tag doing that, and you would be shocked the number of sites, the number of e-commerce businesses I go into where when I look at their confirmation emails, when I look at their marketing emails, the marketing text messages even some of the content on their website, they're just not [00:05:00] using those merge tags.

And you absolutely need to be some KPIs to watch on this first rung of the ladder of personalization. Your email open and click through rate. So if you're getting the basics of recognition right, it should increase your email open rate. This is that old, again, if the email starts with Hi Joshua, I'm probably more likely to read it than if it just says, hi customer.

So that is the first rung of the ladder, recognition.

Now onto the second rung of the personalization ladder, contextual merchandising. So these, again, a lot of platforms have this built in. A lot of people don't always turn it on. Contextual merchandising is. When you see a site that says people like you also bought, or if you have a complete the look bundle on a product that you're looking at.

So quick win here, turn this on. If you're on a solid e-commerce platform these days, it should have a native recommendation engine. I know Adobe Commerce ships with an amazing engine [00:06:00] that uses their machine learning and AI capabilities through Adobe Sensei to do exactly these sort of things.

Other platforms have that as well, so make sure you turn that on. And after you do the KPIs to watch, here are your upsell and cross sell rates. Those should be increasing if you're getting the second rung of personalization. Correct. Now the third rung of personalization. Now we start getting a little bit fancier.

Now we get into journey orchestration. So this is when your email marketing, your text messages, your onsite banners, your offsite ads. When they're all referencing the same last viewed skew or abandoned cart, chances are you have a journey orchestration process set up there. So quick win for this one is have customer segments.

Please tell me you have customer segments and make sure that those segments are wired all the way through your entire e-commerce stack and marketing stack. And what I mean by that is let's say you're using Adobe commerce [00:07:00] and digital. Make sure that those segments flow and all this data flows between Adobe commerce and.digital so that when you are sending out marketing, you're not just sending out an email blast of here's our top sellers, or Here's what's on sale this week.

You're not even sending out an email blast of, Hey, don't forget about this cart you purchased, or that you've abandoned and complete the purchase. You're actually sending out email blasts that take into account where the customer is in their buying journey. They bought one thing from you, two things from you, three things from you.

Are they still researching their first purchase and really bringing all that data together and personalizing the outreach around exactly where they are in their, the journey and relationship with your business. All right, so there probably the biggest KPI is time to repeat purchase if you. Or selling anything that can possibly lead to a repeat purchase, whether it's consumables or if there's just other products that they might buy from you.

You'll wanna watch your time to repeat purchase. If you get journey orchestration correct, then this should start [00:08:00] coming down Now, rungs four and five or where it gets really exciting and fun, in my opinion. The fourth rung of personalization is predictive AI bundling. So this is when a. Machine learning model will predict when a customer needs a refill or an add-on, and it will prompt them to buy it before they even think about it.

So quick win here is the bolt on AI features and functionalities. I mentioned Adobe Sensei also Shopify and BigCommerce both have. Either features like this built in, or in most cases with those guys, you gotta go to a third party app or extension. Get that installed and start with one SKU that people either reorder a lot or that people will often add on and start using that system to predict the best time to get that offer in front of a customer so that they also buy that additional sku.

Your KPI here is ideally signing up. For subscriptions. So if you do this [00:09:00] right with this predictive AI bundling and you also have a subscription or auto ship feature, you can actually offer it to them and say, Hey, we think it's time that you need to buy this again. And by the way, customers like you typically buy this every three months.

So why don't you just go ahead and set up a subscription and we'll just automatically ship it to you every three months? You don't even have to think about it. Now, the fifth rung, I've already covered ai, so you're probably wondering, you know what technology is even better than ai? What is even higher up?

The personalization ladder than ai. And this one might surprise you. If you listen to the podcast a lot, maybe not. 'cause you know that I love blending service and the human touch with technology and ai. So the fifth rung of personalization, the very top level of this personalization ladder is proactive service.

And so what this can look like is an automated alert that. Let you know, Hey, this order that was supposed to ship and be received by this customer by this date, we haven't shipped it yet. It's not physically possible for us to meet [00:10:00] that commitment. We need to do something. We need to either source it from another source, reach out to the customer with an apology, somehow reach out and make this right.

Quick win here. You actually don't have to spend a lot of money for this. You don't have to get super fancy. You could just set up a shipping status web hook that triggers an apology email or an overnight reship through something like Zapier or make, you don't even have to necessarily get developers and code involved.

Of course this goes all the way up then to super fancy support platforms that can do these sort of things built in. So your KPI here is going to be your reviews. What sort of customer satisfaction or negative reviews are you getting? You should be able through this proactive service and using this data, not just a market to.

People, not just to sell them more, but to make sure you're giving them an excellent customer service experience and an excellent buying and post-purchase experience. You're gonna improve your customer satisfaction. So most of y'all that I talked to out there, you're stuck [00:11:00] on the first rung. The second rung.

Some of y'all, we'll make it up to the third rung. I don't see a lot using the fourth or fifth rung very well, so we're gonna talk about that. And the reason that's so important is lots of times moving one rung up, the personalization ladder is gonna drive quite a bit more new traffic and quite a bit more revenue, most importantly, and going out there and buying traffic.

So before you spend more money on your ads to buy more traffic in. Look at your personalization ladder, look at where you are and improve that. And that's gonna give you a much better ROI than just buying more traffic. You don't have to just take my word for it though.

I've got three different case studies that we'll talk about briefly. So one is a company they're pure play, DTC B2C, however you wanna call it, brand. Azio Beauty. And they used a predictive replenishment flow that increased repeat orders 18%. So they plugged in a predictive reorder engine into their e-commerce platform.

It actually models when [00:12:00] each customer is likely to run out based on the SKU order, quantity, seasonality. It really, it goes beyond just, the average time between reorders is six months. It actually looks at for certain types of makeup. For instance, people use this more in the summer. So during the summer, the replenishment rate, the time between reorders actually needs to be shorter than it would be during the winter. It uses that seasonality and whenever it figures out, okay, based on this customer's order, history, what they've purchased, what other people have purchased, the seasonality, these products, it'll actually send them a replenishment email. And it's not just an email saying, Hey, you might wanna buy this.

It literally has a preloaded magic cart. So they can just click through and complete the purchase, especially if you're using something like Apple Pay. This becomes a two click process. You click on the email and then you double click or double tap the button on your phone to complete the Apple Pay. And I mentioned this led to an 18% increase in reorders.

Their ROI on this project was actually 30 [00:13:00] times so huge. ROI. And I see that a lot with these personalization projects where if you're not doing anything in the space and then you move one rung up the personalization ladder lots of times those moves don't cost a lot of money and they can deliver an amazing ROI.

Second case I wanna talk about is MSC Industrial Supply. You can probably guess they are a B2B brand. If you've listened to some of my other episodes you might remember, I will call brands like this, a sidecar. B2B brand. What that means is e-commerce is the sidecar to the main engine of their revenue sources.

So e-commerce is a small but growing percentage of their revenue. And they actually use an AI cross selling system that tied into their e-commerce platform and pulled their e-commerce data along with all the other data from their ERP and went to that fourth and almost fifth rung of personalization with it.

They are a $3 billion industrial supplier and distributor. They replaced a sales intelligence tool [00:14:00] with a new, like I mentioned, AI based tool that pulled their e-commerce data. And it surfaced. It either to the customer on the website, or more importantly, to the sales person or customer service person that was on the phone with the customer.

So this had deep integration in their support system would actually surface due to reorder items. Wallet share gaps co. Related add-ons, all those types of things in real time, while the customer service rep is either working on an email reply or is actually on the phone with the customer. So it is predicting what the customer might wanna buy, giving that information to the rep so that they can provide that higher level of service.

They saw a 20 times growth in upsell and cross sell revenue. They actually convert these recommendations at 13%. So basically, if your reps aren't asking. And make sure you have good personalization data. But if they're not asking when they're on the phone, things like, Hey, based on what you've ordered before, it seems like you might need this product, then you're giving up revenue.

Because [00:15:00] 13% of the time, if you get it right, people are gonna say, you know what? You're right. I need that product. Go ahead and put it in an order for me, or go ahead and add it to my cart. They actually increased their incremental revenue per. Customer service rep by $15,000 per year. So $15,000 for each rep that is working on customer service.

Finally the last one is actually one that I've talked about before on the podcast. So Rural King, we're talking about omnichannel here. If you aren't familiar with them, they have about a hundred stores. It's a farm and home retailer. We actually tweaked their in-store pickup texts and notifications to include usage guidelines.

So think of it, the simplest example I have is a chainsaw. So when someone's going to the store to pick up a chainsaw, you can drop into the notification something that says, Hey, do you have the right kind of oil for this chainsaw? Do you have the bar oil that you're gonna need? If not, it's in stock in the store and we can add it to your order or you can pick it up while you're in the store. And that actually was lifting their basket size [00:16:00] for these store pickup orders by about 12%.

we've talked about the five levels and five rungs of personalization and the personalization ladder. We've talked about how much money your business can make by going up some rungs in the ladder. We haven't talked about probably the most important part, the creepiness factor, your privacy and creepiness guardrails.

And just there's five rungs on the ladder I have a five question litmus test that I will ask before. We implement a new personalization feature for a customer. So very first, and these are simple questions you can ask yourself. You don't need to pay me or an agency to do this. Literally just sit down and you're looking at a new personalization initiative.

Ask these five questions. One, when I say this out loud to a customer face to face just think about saying whatever it is, whatever personalized line is gonna be delivered to the customer. If you were face to face with the customer, if you walked up to them in a store, would you say this?

Because some. Times. I find that there's fairly [00:17:00] creepy things that as marketers, whenever we're designing these emails, we'll throw it in there because we can, and we never stop to think, I wouldn't ask this person that. That's obviously creepy. Second question is, does it use only data the customer knowingly gave us?

And this is tricky. I know there's a lot of value in customer data platforms like Segment where you can do enrichment, you can pull in data from other sources. You might have marketing partners that you're sharing data with, but just be aware. The more data you're using that the customer didn't knowingly give you that you're getting through data brokers and other sources, the closer you're gonna get to raising that creepiness factor in their mind.

So using it strategically definitely makes sense, but if all you're using is third. Party data and you're starting to get really creepy with it, it's probably actually gonna turn customers off more than it is going to increase your sales. Third, will the message surprise or unsettle them? You've probably heard of the example that apparently may or may not be true, but.

Some time ago, there's a story [00:18:00] that circulates about how Target got their predictive analytics and predictive personalization. So good that they dropped a flyer in the mail to a teenage girl about baby products. And her dad gets it and gets really upset and is wondering, why are you marketing this to my kid?

Are you trying to encourage her to have a baby? She's still in high school. What's going on here? And it turns out she was pregnant and it was that, the data that Target had helped them predict that sooner than the father figured it out. And again, I've heard some stories that say that is true, that really happened.

I've definitely seen examples online of when Target sends marketing like that. I don't know if that exact story is true, but it's a really good example of. Don't send someone a personalized marketing message that might unsettle them or might surprise them if it's something that they themselves may not even know yet.

Just pause and reflect on if that's a good level of personalization or not. [00:19:00] Fourth of course, does it comply with your local privacy laws? I'm not a lawyer. And this is a free podcast, so I'm not gonna give you advice on. If it complies to the local privacy laws or not. So obviously take them into account with any personalization initiative.

And then five, most importantly, does it add real value to the customer's life? Are you doing something that actually contributes to their life or are you just doing something to try to squeeze an extra dollar out of them? Again, I get it. We're all in this business to sell more. That is how we keep the lights on, keep the paychecks coming in, but.

If you can find ways to do that adds value to the customer's life, it will be a much more sustainable personalization project, a much more sustainable relationship with that customer, and ultimately a much more sustainable e-commerce business. Now, I mentioned the sidecar B2B. I'm gonna briefly talk about how you can align your personalization efforts to the type of business that you have.

So if you are a sidecar B2B business, chances are. You don't have a lot of personalization, you're not very [00:20:00] high up that personalization ladder. I would say start at level two. Make sure you have level one covered, and then really focus your time on level two. Looking at what does your platform offer built in to cover that second rung.

Or is there an extension or third party service provider you can bring in to help you really completely implement that second round? If you're an omnichannel retailer, then level three customer orchestration, customer journey orchestration is absolutely important. Especially make sure you're tying in your store point of sale data with your e-commerce data.

And you're really orchestrating that as an omnichannel journey. And then finally, if you are a pure play digital, so basically you're B2C or B2B, but you're purely digital, all your revenue comes from e-commerce. You have to be at level four if you are not already on that fourth rung. You may not make it through the rest of the year.

This is the level of personalization that people expect out of pure play digital retailers. [00:21:00] Level five. As you move up to that fifth rung with that proactive predictive service, that actually becomes your competitive moat that you can build. So the more that you can do with that the more protected your business is gonna be.

Now, I always like to talk about metrics. We talked about a few metrics in the start of the show. I actually have a whole series of episodes and a whole chapter in the new e-commerce book that talks about metrics. So I want to talk a little bit about metrics for each rung in the personalization ladder.

So first rung of the personalization ladder recognition. Most important metric is email open rate. Click through rate can be a secondary metric. If customers aren't even opening your emails, then every other rung isn't gonna matter because they're not opening emails. You gotta get this right first.

So how to measure it. Use your email platforms reporting capabilities. Any email platform you're using must show you the open rate. If not [00:22:00] runaway, find a new one. Look at your 30 day rolling averages here. Good starting target. For open rate is above 30% for a house list, so your own data and then click through above 3%.

Second rung contextual merchandising. So your metric here should be your upsell and cross sell conversion rate. This will show up in your a OV average order value, but look first at upsell and cross sell conversion rates because the whole point of a complete, the look set up. Is a larger basket. So it's a higher a OV and it's that conversion rate increasing.

So measuring this, you can set this up in Google Analytics or whatever analytics platform you are using a good starting target for this. Is your upsell conversion rate being above 5%, or if all you can look at is a OV, then a 10% lift in a OV versus your baseline. Third level journey orchestration. Best metric to start with on this one is your [00:23:00] repeat purchases inside 60 days.

Because a coordinated marketing and buying journey should really shorten the time between orders. To measure this put your first time buyers into a cohort each month, and then calculate what percentage of those is buying again within 60 days. Good. Starting target as you move up that rung in the personalization ladder is to raise your repeat purchase rate, two to four points in the first quarter.

Fourth level predictive AI bundling. Best metric here that I usually track. Is your subscription or your auto reorder opt-in percentage to see how well targeted those offers are? Because your predictive flow should lock in future revenue. That shouldn't just be a one time order, one time lift in revenue.

So how to measure this, divide the number of customers who accepted a suggested subscription by the total number expose the offer. Super simple. Good starting target. If you're getting this right, you should see eight to 12% opt-in rate if you're selling replenishment or replenishable [00:24:00] SKUs.

Finally, the top rung proactive service. Two or three different metrics you can look at here. Just pick one. Based on what is easiest for you to measure. It could be negative review rate, could be customer satisfaction score, or a net promoter score. And that's because these proactive intervention should slash the where's my order questions before you get a negative review, you can track this as looking at one and two star reviews as a percentage of total reviews or monitor your customer satisfaction ratings. As a starting target, you really want to cut your negative review rate in half in the first six months or so, or raise your csat depending on how you calculate it.

Raise it by. Three points. Make sure that you anchor each personalization initiative to a single metric that matches where it is on the ladder. And don't use vanity metrics. I still see people use metrics like emails sent. Okay. You have a personalization system that can crank out a ton of emails. Cool.

Is anybody opening those emails? Is anybody ordering [00:25:00] off of those emails? That's what's important, not just the number of emails sent. So real quick, I would recommend that after you've learned about the personalization ladder here hold a 30 minute ladder on it with your e-commerce lead and with your e-commerce agency.

Choose one experiment, just a small experiment you can do about moving up the ladder at least one step. Assign a clear success metric and an owner, ideally, pick something super small, something you can implement in seven to 14 days. Start there and then review the uplift, see what you learn from it, and use that to design a larger personalization project.

So hope you find that helpful. As always, you can find me on LinkedIn as Joshua Warren with a Creatuity gold background behind my head. You can schedule a free 30 minute e-commerce problem solving session with me. Actually, I'm renaming. Those to e-commerce brainstorming sessions, I think. But stay tuned if you want to schedule one of those or please continue to like, subscribe, [00:26:00] comment, et cetera on the podcast episodes.

And I'll have more news in the next few weeks on the release date for that new e-commerce book I've been teasing y'all with. So hope you find this helpful and I hope you're having a great week.