Video Transcript

Okay, so welcome to our workshop. I hope you had a good break and that you’re really fresh for the fourth workshop. I know it’s hard, so we’re going to try to make it as friendly as possible. So, today, we’re going to discuss one-time customers. And this topic is actually really interesting because, as marketers, and I’m sure you’re all familiar with it, we invest a lot, and we put a lot of effort into acquisition.

And we try really hard to create this beautiful content to be unique, to be creative, to engage with communities, and to give the customers the best experience we can give them. And then we get customers who buy once, which is great, but never come back. And this is very sad. And that’s what we’re going to discuss today.

Before we start, I’m going to present ourselves. I’m Daphna. I’m the Marketing Manager at Optimove in New York of the North American market. And here with me is Agathe, and she’s our Director of New Business straight from New York as well, or Frnace, depends. So, let’s start. So, 59% of first-time customers never come back.

And, as a spoiler, this number can even get higher, and it does get higher. In some industries, for example, in e-com and fashion, it can get even up to 70% or 80%, which is kind of crazy if you think about, not only the amount of time and effort you spend into acquisition but also about the money you spent into acquisition.

And acquisition costs are rising all the time, and they’re like three or four times higher than the cost of returning an old customer. So, we have a real incentive to be more mindful about the type of customer I’m acquiring, who am I bringing in? And most importantly, what’s next? What do I do once they’re already there?

What do I do after they purchased for the first time or made the first deposit if we’re talking about gaming customers? So,s what do I do next? How do I make them buy for the second time or make a second deposit? And why am I talking about that second deposit or second purchase and not about generally repeat rates?

The reason is because second rate, second purchase rate or second deposit rate is by far the biggest hurdles for marketers to overcome. It is much harder to get a customer to buy for the second time or make a second deposit. It is very hard. But once they’re already there and bought for the second time or made the second transaction, it becomes easier and easier.

So, as you can see in the graph behind me, only 41% of the customers make it to the second transaction. But after that, after they’re already in and bought twice or had two deposits, it becomes easier and easier over time. So, how should we approach one-timers? So, Agathe and I build this session kind of to be a DIY session, so it’s meant to be very practical.

And this is something you can definitely implement on your own with your team once you’re back at your offices. So, the first thing to do is look at the data and understand how much time, on average, does it take to your customers to make that second transaction. And I’m saying “your customers,” because this number is different across the board.

And for one company, it can be a week, and for the other, it can be three months. And for a different company, it can be a year or nine months. So, this number varies, and it’s different, but it’s fundamentally for business. It is really important to understand that timeline if you want to maximize the chances of your customers coming back and making that second transaction.

So once you understand that timeline, and you really understand what’s the average time does it take to your customer to come back and make that second transaction, you could define the lifecycle stage. And you can understand when is the customer a new customer, and this is a customer that is in this timeline and will hopefully make a second transaction, and when a customer becomes one-timer.

And a one-time customer is what we’re trying to avoid. And the customer becomes a one-time, a one-timer customer once he exits that timeline, that timeline that we discussed. Once he exits that timeline of the average time that usually takes to a customer in your company to make that second transaction, once he exits that timeline, he becomes a one-timer.

And we want to avoid it, and why? Because once he becomes a one-timer and exit that timeline, it will be much harder to get him to come back, or she, and it will be much more expensive. So, we talked about that timeline. And once we understand that timeline, let’s say it’s 30 days for your company, so it is important to understand that in this threshold, it’s exactly at this point.

There is this threshold. And it’s important to understand when they’re becoming at risk of becoming one-timers. And this is something you just need to decide. If your window is 30 days, it can be day 20, it can be day 25, it can be day 28, or 29, but it’s something that is very approximate and very close to that end of timeline.

And, in that window, when the customers are at window…sorry, when the customers are at risk of becoming one-timers, we try to capture them all and automate campaigns that will help us to make them buy for the second time or make a second deposit.

So, this timeline, we’re going to talk in this session about this timeline, and particularly about this timeline. And this is the time we went to capture them all and help them to make that second transaction and pass that hurdle that we talked about. So, here’s an example of something we did with one of our clients, and I know it kind of like looks intimidating.

I’ll explain. This is something we did with one of our retail clients, but it’s the same idea. And it’s very relevant to all of our clients, gaming and retail is one. So, what we did here is we to tried to understand, where do most customers fall in terms of making that second transaction? So, all the numbers that you see here for this specific company is customers who already bought for the second time.

So when we look at this graph, and we see this orange line, which is accumulative, so we see that 15% of the customers who bought for the second time made that second purchase after one week. And then 23% of the customers who bought for the second time made this purchase, the second purchase, after two weeks, and so on.

And then we see that after three months, 70% of the customers bought for the second time. And what it means is that for this specific company, in particular, 70% of the customers, most customers made their second purchase within 90 days.

So, for this company, the timeline is 90 days, three months. And once we understand the timeline, and when a customer becomes at risk of becoming a one-timer, we, as I said, want to capture them all and target them and automate campaigns.

But how do we do it? So, the first thing, before I hand the microphone over to Agathe, is to note that not all one-timers are the same. And what do I mean? I mean that because they’re all at risk of becoming one-timers, it is not a good enough reason to batch and blast them and send them all the same message just because they’re all at risk of becoming one-timers.

It is really important to understand who is standing in front of you, who is the customer and approach them differently according to what makes each one of them tick. And it can be different reasons and different things. And I’m going to hand over the microphone to Agathe that right now, and she will get more practical and discuss with you about her tips and best practices.

– All right. So, once we have the window of risk, right, we’ll have a good understanding of when do we need to interact in order to minimize the chances of the risk actually of customers to not come back, right? So, we’ll have that understanding, whether that window of risk is day 75 out of 90 or day 15 out of 20. We’ll have to find that out for your own company.

Now, what we’re going to go into now is more of a DIY, do it yourself, DIY, do it yourself session. And so what we’ve done here is just find ways of segmenting your one-time or risk of being one-time customers in a way that makes sense for your business. Now, because I’m guessing that you all come from very different businesses, or even if you come from within the same industry, different types of customer base, we’re going to look at ways of segmenting that across the board or meaningful.

What do I mean by meaningful? Segmentation is a buzzword, right? So there’s a really kind of bad habit in the marketing space, which is to just take segmentation as the idea that, you know, trump’s them all into segments for the sake of it. There’s a billion ways of segmenting in a way that’s completely meaningless, right? Recency of transaction, for some business, has absolutely no impact on customers’ future behavior.

So the ways that we want to group customers are ways that make sense for your business that we found through your data will, you know, move your customers and elicit similar reactions. So, we’ve picked four, day of the week, time of day, order amount, and product categories, and we’ll go through them. I’ll switch this. Okay, so the first one that we’ll look at is day of the week. So, this is all coming from Optimove customer data.

Can you guys maybe show me by your show of hands, how many of you guys are working with or in the gaming industry? Okay. And anybody else in e-commerce, retail, or anything else? Okay, that’s Boris. That’s it. Cool. So, we’ll adjust.

So, this is all Optimove data, and the first thing I will look at is day of the week. Thankfully, all of this really applies across the board, and it’s coming mostly from our Strategic Services Branch. So, we have a graph here. And what we’ve done is trying to find a correlation between when people first transact and then have the second transaction, right? So vertically, you have the day of the week of the first transaction, horizontal, you have the day of the week or the second transaction.

And what you can see is that there is a very, very, very strong correlation, right? People are creatures of habit. If I tend to play on a Sunday night, that’s probably going to be my behavior and my pattern with variances. If I tend to shop on a Tuesday evening when I come back from work or whatever reason, I’m probably going to do this on the first and second order. So, what does it mean?

It means that we need to be consistent within that window of risk that we’ve already talked about. We need to be consistent about how to address customers and when to do so, right? The thing that we shouldn’t be doing is probably set up seven different campaigns per day of the week, right? That’s probably not the most efficient, emotionally intelligent way of doing that. It’s probably going to take a lot of time for your marketing team. So, typically, what a lot of people do with that is that they have one large automation that is picking up on multiple of seven, starting from the day of first purchase, in which case you will have that second email or communication that will fall on the correct day of the week, regardless of what that was for that same person, right?

So that’s one of the tricks for implementing this. With that correlation, it’s true across the board. The second one is time of day, right? So, in that same way, this is not rocket science. And, you know, you’re not going to drop off your chair in surprise that there’s a correlation there, but here we have a similar…you know, actually it’s a different customer, and we’ve done a similar analysis, time ranges of transactions for the first transactions, time ranges of the second transactions.

Here, you had that same correlation. But here you have an additional level of difficulty, which is, unlike the day of the week, for the time of day, you actually need to time it a little bit earlier when they would typically come back and play or come back and shop, right? We actually…I don’t know how many of you, if any, were here last year, but we had the CMO of Dollar Shave Club who came to speak on stage. And, Dollar Shave Club is a subscription income business in the U.S.

And she said something like if a customer is giving me their email address, it’s like a friend is giving you a key to their house for you to come visit. If you come to the house at 3 a.m., and you wake up the kids, and the dog start barking, nobody wants you, right? Marketing is a little bit the same thing, and the timing of the email is one very good way of applying that principle. Third one is order amount.

So, this is interesting because it looks the same, but actually, it’s less intuitive, okay? So, this is a third brand, and this is an e-commerce brand. And we’ve actually grouped, you know, first order and first transaction by the order, like the average order value baskets, right? And same thing for the second order, second transaction horizontally. Here, you see that same correlation, right?

What am I saying? I’m saying that people who spent $100 on their first transaction will most likely spend within that range on their second. It’s counterintuitive because a lot of the time you have marketers thinking, “Okay, I put in a lot of money to acquire those customers, I have my foot in the door. Second-order is the time for me to try to upsell you and cross-sell you into a higher spend so that, you know, I can make my money back,” right?

It doesn’t work, not for a second order. It’s to high a risk to take. A lot of the time, there is an expectation that you will target customers or, you know, whoever, your clients with something within a similar range. Which means that also, typically, so in the retail space, for example, this habit sometimes of, you know, “I bought a blazer and the marketer will have recommendations within that same categories, but sometimes much higher price range,” also doesn’t work, right?

So, there’s a lot of reason to be mindful about this. And another fun fact, which is…don’t know if it’s a fun fact, it’s at least interesting, is that, of course, there’s a very high correlation between the volume and the spend of the first order and the future value of those customers within the next 12 months, right? So, you want obviously to get that ticket price up, but you want to be mindful as to what you’re making of it afterwards, okay?

Fourth one is product category, okay? So, this is a little bit meatier, so I’ll walk through it with you. So, this is an ecom customer of Optimove who…that light is not getting better. It’s an ecom customer of Optimove who sells luxury children’s clothes, right? So, we’ve run an analysis and, again, surprise, but it works the same in gaming. We’ve broken them by categories of first purchase.

And horizontally here, you have the spend, so how much they spent on their first order. Vertically, you have the probability of them to not come back, right? So the higher you are up on here, the less likely you are to come back for a second purchase. So, what is it that we’re talking about here? We’re going to be talking about, how can I engineer my acquisition efforts to ensure to the maximum probability that I’m generating customers who will be more likely to come back, right?

A lot of the work is done before. You need to actually email people to make them come back, okay? So, if you’re looking at the bottom large circle, you have one category of first purchase, which is girls, right? So, if purchase from girls is apparel, and what I’m seeing here is that it’s a category that has, like, pretty high spend and also very low chances of people not coming back or high chances of returns, right?

So, if you run an analysis with your own categories and your own products, it will tell you some really, really good information. For example, everything that’s at the bottom, right, in the bottom two quadrants are excellent categories to recommend for acquisition, right? Because I will be generating people who come back, right?

So, of course, the higher the better, but anything that would be here would still actually be a better product category to recommend for your acquisition effort than anything that is on the higher quadrants. It doesn’t mean that we’re dismissing everything that’s up there, it only means not now, right? Third purchase, fourth-order, fourth transactions, whatever, but just don’t play around with recommending categories that generate one-timers, right?

So that’s one example about, how can I control it before, okay? And then second, product category, number two, I’d say. So, this is the same brand. And we’ve split up the categories of first order and second order, so similarly to what we’ve seen so far, and what we’ve tried to try to find here is very simply correlations, right?

We’re trying to identify, what are commonly walked paths by customers from one category. You know, what are the categories in between which they travel from first to second. Why? Because I want to know what to recommend, right? We have the window of time for intervention. I have the day of the week. I have the time of the day.

I have a very strong idea as to how much, what’s the ballpark average spend that I’m gaming for, right? And now we’re saying, “Okay, what do I recommend?” Okay. And so here, you can see highlighted, you have some high correlations between, I don’t know, like baby accessories and baby apparel, or women’s clothing.

So here you have this path that obviously you don’t want to take a risk, so these are probably the best ones to recommend. On the other hand, everything here that’s nothing are paths that nobody ever walked before. So, it’s up to you if you want to take the risk, but maybe not at that crucial time, right? And so I think what we were talking about earlier is that segmentation is really, really idiosyncratic, right?

It’s unique to your brands. We’ve just looked at four that, across the board, by a very large probability, there will be a correlation for all of your business between the four things that we’ve just looked at and the likelihood of people to come back. But it could be really anything else, right? So, I have different examples. But beyond that, it could be, you know, have they reviewed? Did they call customer service? What was the level of, you know…?

Are they part of the loyalty program? Were they recommended into your brands? Like, did they click on the site so many times before their first transaction? I don’t know, right? We just need to find out. Its unique per data set. I’ll just go very quickly through those three ones. Returns is more relevant to the e-commerce world, but I think that a lot of retailers are trying to avoid actually returns, right?

They see it as something that could be potentially negative, and, you know, it’s a lot of margin, and you have to pay for shipping, etc. We have found out, for a lot of our customers, that people making their first transaction with a partial return, so they return part of their order, actually was often credited to VIP customers, you know, in the future months, right?

So, actually, it doesn’t mean that there’s a causation, it doesn’t mean that you should be trying to make your customers return. It just means that if you have a subset of customers who have a partial return on first order, those guys you should probably set aside as a sub-segment, you know, time allowing. The second one is basket variety. So, if you forget about the basket idea, this is applicable very much so to the whole gaming world and reach pretty much any business.

As a rule, the more varied and the more cross-sectional the initial basket and purchase is, the most likely your customers will be to spend more and to come back, right? So, there’s a lot of things that, you know, we can talk about maybe later as to, kind of, how can I encourage that cross transaction, right, which is recommending things at checkout and things like these.

But this is something to keep in mind. The third one is really important, and this is applicable for gaming, for e-commerce, for financial, for anything, first purchase discounts. The way that you’re utilizing promotion should be very, very different for your first orders and your second orders and for everything after that. A good example of that is high discounts, right?

High discounts are probably really good for your loyal customers sometimes. It creates a very different type of behavior for acquisition, right? Acquiring customers or new games on high discounts creates a lot of cherry pickers, right? You might attract people who are there only for the discounts, right? It’s encouraging a behavior. It’s not going to look or be the same thing from the third transaction onwards.

But for those early stages, it can be dangerous. On the other hand, if you think about anyone who purchase first with no discount at all, that’s something else, right? It’s either an impulsive buy or something that’s really need-based. And so those people could also be treated differently. In the middle, we have you know everything else, and I’ll give you an example from an Optimove client who wanted to look at first purchase discounts.

And so we segmented, and we looked at their customers who made their first order with no discount at all, 5% discount, 15% plus, and 20% plus, right? It turned out that people who made their first order with a 5% discount, so the lower level of discount, spent 20% more over the next year than anybody else, including the customers who had first purchased at no discount.

Yeah, that’s the surprising fact, right? It’s not that intuitive, right? So, there’s a lot of value in looking at those other ways of segmenting. That’s my point. I’ll leave you with that, and I’ll hand it back.

– Thank you, Agathe. So, to summarize, we talked about the timeline, and how important it is to understand when a customer is a new customer, when a customer is a one-timer, and when he or she are at risk of becoming one-timers. And then as Agathe explained, if you really want to declare war in your one-timers, it’s very important to understand and be mindful and to pay attention to that correlation between the hours and days of when a customer is making a first purchase or first transaction to when they’re making that second transaction.

And same goes for order value or deposit value. And it’s really important also to remember that product category is something to be mindful when you’re trying to acquire customers, and not once they’re in to be very mindful about that. And the last thing, the last tip we want to give you today is test. And when I say test, I mean, test the behavior of customer who already made that second transaction.

And why is that is because when you understand their behavior and analyze it, you can really kind of reverse engineer it and see what worked and what didn’t work. And once you fully understand that and you implemented on your new customers, you’d be able to maximize the chances of them coming back. So, that’s it for today. We hope you enjoyed it and that it was helpful and productive for you.

And we’re happy to answer any question you might have.