The (Costly) Reason Why Zalando Bans Customers | Growing Ecommerce

Episode Transcript

Mike: 0:00:09
Mike, Chris, how are you doing? Good. Hold on, we have one step before that—the intro. Here we go. Welcome to Growing E-Commerce. I’m your host, Mike Ryan, and with me? Chris. Again. Hey Chris, this is episode 13 of our co-hosting. I’m proud of ourselves. Lucky number 13. Let’s see. Exactly. And before we get started, Chris, I have a question for you. Did you watch Dumb Money? Not yet. Okay.

Chris: 0:00:43
It’s on my watch list.

Mike: 0:00:48
Have you watched Margin Call? No? You should do that. It’s on my watch list too. I was telling Chris that I was actually born in the same city as Roaring Kitty and then went to the same college as Roaring Kitty. So, we knew each other briefly before the millions. The big question is: have you participated in any way, shape, or form? This is very personal, Chris, because actually, we were building our cellar and garage at exactly that time. We were out of money, probably because everything was at zero capital anyway. So, I completely missed out on this entirely, and now it’s one of the biggest new opportunities. One can only hope. But I will watch it, I promise. I really think it’s an inspiring story.

So, we’ve got a couple topics on the agenda today. Let me lead in to the first one here. Just last night at home, Steffi was saying that she’s waiting on some overdue return money from Zalando. Do you shop at Zalando? Are you a fan of that?

Chris: 0:02:03
No, but shout out to my wife. She is, and I love you. I will talk about the news you’re breaking now.

Mike: 0:02:03
Yeah, sure. So, these two things are unrelated, but I had to warn Steffi that there is some news from Zalando. They are tightening up their return policies. Basically, in a nutshell, they’re working to identify what they consider disproportionately high return volumes. I don’t know what those thresholds are—I don’t think they’re transparent about it. It could, of course, be relative to your overall order volume, or it might be personalized.

Chris: 0:02:33
So there are no specifics yet?

Mike: 0:02:36
Not that I’ve seen, no. But what they’ll do if they think you have too high return volume is, first, they’ll give you an email warning. So, check your Zalando emails; it might not just be a coupon. Then, they might temporarily shut down your ability to place new orders. And if they’re really not happy with you, they’ll ban you for a year. They’ll shut down your account for a year.

Chris: 0:03:02
Okay. We talked about that, right? Happy wife, happy life. Imagine your wife is not allowed to order from Zalando anymore.

Mike: 0:03:10
What is normal shopping behavior and what is not?

Chris: 0:03:16
That’s a big question, and that’s why I would love to get the specifics on it. Just yesterday, we had a big quarterly business review with one of our biggest clients, and return rates were a big topic. The return rates of this client were very low, significantly below 6%, so I think they’re doing a great job there. But it was again a massive question: how can we use and deal with this data in order to understand which products are actually very prone to be returned? And how can we use this data again to optimize our paid search campaigns? I think this is a very important topic for every retailer. As far as I know, especially when you talk about the big retailers, I haven’t heard anything remotely close to what Zalando is doing right now. Have you heard of any other retailers going down this route?

Mike: 0:04:06
I think Amazon for a while has had an abuse detection policy, but I can’t remember what the penalties are for that. This question about what is a good return rate varies a lot per category. Of course, it’s going to be higher in fashion, whether you touch on size variants, color variants, et cetera. I think there are actions you can take on the advertising side. If you have a product with a high average return rate, do you want to promote it as aggressively? Also, is there stuff that you could do on your landing page side? That’s definitely where a lot of investments are being made right now with AI, especially with virtual try-on technology.

Chris: 0:04:52
We’ll talk about that in a minute. One thing I find is that buying fashion online is so appealing to a lot of people because you can just order whatever you want and there is this possibility of return. You have your private wardrobe at home. For me, the big question is how restrictive will Zalando be? Because I think it’s a big part of the value proposition against offline shopping. I would love to get some specifics on it; I think it’s a very slippery slope.

Mike: 0:05:32
I agree. Steffi will buy stuff in multiple sizes and colors, and some of that stuff is going back 100%, but it’s done with the best of intentions. There are also Instagram influencers and TikTokers who are buying stuff, wearing it once for a video, and then sending it back. That’s clearly abuse. Some of it needs to be baked into pricing. In their press release, Zalando talks about how this is unfair to other customers because it’s tying up inventory. I think there’s truth to that, and it’s not environmentally sustainable either due to unnecessary freight movements. It’s disruptive to their processes, but let’s face it: it’s also about profitability. The word “profitability” was not mentioned there, and I don’t know why they need to talk about everything else besides that. Everyone knows that’s the name of the game.

Chris: 0:06:39
To give context, in the fashion industry, it’s quite normal to look at return rates between 30 and 70 percent, varying by category. Some products will be massive outliers, and this is obviously a massive driver for profitability. This is one action you can take as a retailer, and let’s see how restrictive Zalando is. It will probably have an impact on demand—likely the right impact because it’s demand you maybe don’t even want in the first place. But for me, the bigger question is something we discussed in a previous episode: return rates are first-party data that heavily impacts your profitability in Google Search or Microsoft Search. What is your take on activating first-party data like return rates in PMax or standard shopping campaigns?

Mike: 0:08:05
An average return rate is not the hardest data to calculate, and it’s super important. When people hear “first-party data,” they often think of remarketing audiences, but it’s your product data as well. A return rate is off-channel data that Google doesn’t know about. It’s decisive because they’re bidding on the order value. You can send an adjusted order value later, but how soon can you send that over? Years ago, back in 2018, I wrote an article about a retailer who was segmenting products based on margin brackets and return rate brackets. They were creating buckets at the intersection of these two. Even back then, it was possible to integrate this data. The question is, will people do it?

Chris: 0:09:14
When will people do it? If you want to go down this route, I think a smarter way would be to anticipate it and push products where you know, on average, you can deal with the return rates. That’s a more proactive way to manage it.

Mike: 0:09:44
Exactly. I’ll dive into two different options here. In the Google universe, you can use “conversions with cart data.” It’s a small snippet of code that unlocks your clicked-versus-bought data—the item that received the click versus what was actually in the order. This brings these effects into your campaign data so you can look at this at the item, brand, or category level. You can even intersect that with custom labels for margins.

Another thing is a package in the Google Marketing Solutions GitHub repository called the “Product Return Rate Predictor.” I’ve said this before, but some of the coolest things Google builds are not in Google Ads but on their GitHub. They took their Lifetime Value (LTV) predictor model and modified it to predict the conversion value after returns for every transaction. If Customer A orders $150 but returns 80%, and Customer B orders $120 with no returns, things look very different. The algorithm doesn’t always have the chance to learn that if the return window is long. This tool predicts that value so you can import a modified value.

Chris: 0:11:58
That’s super smart and definitely worth testing. Return rates are a huge dark horse. Everyone knows the impact on profit, but I just don’t see enough retailers building strategies around this first-party data. My shout-out to every online retailer listening: look at your first-party data and activate it. Mike, maybe we can test this predictor with some clients.

Mike: 0:12:40
We’ve talked about the advertising side, but on the landing page side, there’s the dimension of virtual try-ons. I’ve used some AR features in the past that were pretty underwhelming, but generative AI is a much better way of solving this. In the US, a company called Stitch Fix has a huge data science team working on virtual try-ons. You upload an image of yourself, and they use generative AI to show you what the clothes look like. They are a subscription company that sends personalized outfits, so it’s a perfect fit for them. The major platforms are tackling this too.

Chris: 0:14:09
It’s a great technology because it reduces return rates in a positive way. Instead of banning people, you help them understand if the clothes will fit or look good. It’s a better experience for the client. Who do you think is leading the pack right now? Is it Amazon?

Mike: 0:14:59
For me, it’s definitely Google. Amazon has virtual try-on, but it’s limited to shoes and eyewear. Google is leading on fashion with full-body try-on coming to Google Shopping and a standalone app. They also use AI to generate 3D spins from standard feed pictures so people can rotate items. They even have a virtual try-on API in Vertex AI using their Imagen model, which allows retailers to use Google’s technology on their own sites.

Chris: 0:16:00
Return rates are a massive topic with many ways to tackle them—technology on the website, activating first-party data, or using predictors. It’s a massive profitability driver.

Mike: 0:16:23
If it’s something hurting your company, there are options to address it.

Chris: 0:16:31
Let’s end on that positive note. There are solutions out there.

Mike: 0:16:38
We’re making headway, and some of it is thanks to AI. To the AI haters—of which I can never decide if I am one or not…

Chris: 0:16:46
Look, we had our debate about whether there’s an AI bubble. It’s working. AI helps you, for sure. Let’s move on. What about your shopping behavior going into the peak season? Black Friday and the holiday season—will you change your buying habits?

Mike: 0:17:16
Me? Or my wife? We are absolutely deal seekers. Price is one of our number one criteria. Not always, though. I just bought the I Spy book series by Walter Wick out of pure sentiment. They were massively overpriced because availability in Europe is limited, but I need my kids to have them this Christmas.

Chris: 0:17:55
I’ve also become a very price-sensitive person, but certainly not when it’s related to my kids.

Mike: 0:18:00
Exactly. The price elasticity is totally different with kids and pets.

Chris: 0:18:10
All jokes aside, there is some interesting perspective coming directly from Google on this peak season.

Mike: 0:18:10
A couple episodes ago, we talked about conversion lag and seasonality. October is a tough transition month—the “October blues”—because people start deferring purchases in anticipation of deals. On the other hand, Black Friday has become diluted. It’s gone from a day to a week, then a month. Maybe we’ll have a Black Friday year at some point.

Chris: 0:19:01
On a serious note, based on the data I presented at DMEXCO, is Black Friday still massively impacting Q4? While we don’t see the same pandemic-level peaks, it is still by far the biggest Friday of the year. But Google says there are other days just as big.

Mike: 0:19:39
They have a new thesis this year called “Triple Peak Week.”

Chris: 0:19:46
The marketers at Google are good. Can we call it “Peak Triple Max Week”?

Mike: 0:20:00
Their thesis is that Black Friday is just one of three peaks. The others are what they call “Surging Sunday” (the Sunday before) and “Cyber Sunday” (the Sunday immediately following). They’re saying it’s all about those three days. Sunday is a logical shopping date for consumers.

Chris: 0:20:33
Google is looking at this from a demand perspective. They say Surging Sunday has 107% of the demand of Black Friday. My question is: what does “demand” mean? I assume it’s search volume, not necessarily conversions. However, it’s still strategically important to be there when people start their research.

Mike: 0:21:31
I think it’s search volume too. That’s the peak of latent demand. In that case, it’s about moving up-funnel with Demand Gen campaigns or Meta activity to support search lift and conversion volume down the line.

Chris: 0:22:04
It feeds perfectly into our discussion about the lead-up to peak season. It’s wrong to look at isolated conversion data because a lot of conversions haven’t happened yet. If you look at “conversion by time,” you’ll see these days are vital. The worst thing a retailer can do is be defensive with ad spend in the weeks before Black Friday. The demand is happening, and you have to be there.

Mike: 0:22:58
I’ll be doing some research on past years in advance of Black Friday. We also have a free tool on the website called the SMEC Market Observer. You can see benchmark data like CPC and click-through rates. I’m working on an overhaul right now to add a competition monitor.

Chris: 0:23:36
It’s a hell of a tool, and it’s free.

Mike: 0:23:39
Well, nothing’s free. It costs an email address.

Chris: 0:23:39
We want your email!

Mike: 0:23:42
But really, this data is great.

Chris: 0:23:44
We have billions of data points. There are advantages to being a big player in the paid search universe, and one is having this overview. I really encourage everyone to look at the Market Observer; it gives you perspective on your own data. Time flies, mate.

Mike: 0:24:13
Shout-out accomplished.

Chris: 0:24:15
At least one shout-out per episode—that’s my new goal. But really, it’s a great tool.

Mike: 0:24:24
That’s going to do it for today. Thanks everyone for tuning in. Do you want to do the outro, Chris?

Chris: 0:24:31
I would love to, but there’s no way I can do it as charmingly as you. The floor is yours.

Mike: 0:24:40
I think I’m known for being awkward rather than charming, but let’s go. This has been another episode of Growing E-Commerce, brought to you by Smarter Ecommerce, also known as smec. To learn more, visit smarter-ecommerce.com. We have great free resources like our Market Observer, free scripts, and our blog. Thanks for listening, and we’ll catch you next time.

Chris: 0:25:10
All the best, guys.