The Customer Segment You’re Probably Forgetting
Getting to know the customers in the Reactivated lifecycle stage can decrease your churn rate by 20% and generate more loyal, long-term customers.
– [Motti] Good evening, everybody. I think you can see me okay on the webcam.
– [Recording] The broadcast is now starting. All attendees are in listen-only mode.
– Good evening, everybody. I hope you are well. My name is Motti Colman. I hope you can all see me okay. I’ve got a white wall behind me and a bit of light there, so I’m just…yeah, I think that’s probably best. Pleasure to meet you all. Thank you for joining us today on what’s hopefully going to become an extended series or webinars that Optimove will be running. So again, I really appreciate your time.
Just to give you a bit of sort of, you know, sort of the house rules, this webinar is going to be 45 minutes long today, of which about 30 of it I’ll be presenting. I’ll be talking, obviously, about this concept here, “The Customer Segment You’re Probably Forgetting.” At the end, there’ll be time for a Q and A, so we’ll do about 15 minutes’ worth of Q and A at the end. And littered throughout, we’ll also try and make it as interactive as possible. So I’ve got a couple of questions that are going to pop up on the screen for you, and it would just be kind of on a survey basis. It would be interesting to get that information back pretty quickly. Hopefully I’ll be able to use it and tailor some of the answers in the way that I actually present on this based on those questions.
So please do interact. Please write questions into the chat box. And yeah, let’s hopefully make this interesting. I’m going to keep my webcam on for the entire time so that you can see my face. It’s probably a little bit more interesting. I hope so, at least anyway.
How Do You Approach Marketing Strategies for the Reactivated Customer?
Okay. So we’re going to be talking today about, as you can see here, the customer segment you’re probably forgetting. But that doesn’t answer the question now what that customer segment is. We’re going to be talking about the specific parts of the customer journey or we talk about the idea of a reactivated customer. So once you start bringing customers back from churn, it tends to be a customer that people don’t really approach, we think, in the right kind of way. So it’s going to be something that we’re going to be trying to focus, trying to understand various different elements of that customer’s behavior, and seeing how we can actually approach that customer to make sure that they become a customer that gets fully reengaged with the brand, hopefully becomes active, spends a lot of money, and increases their loyalty.
So to kick that off then, what I’d like to do is actually present the first poll right now. And that’s going to be a poll talking about you understanding your unique marketing strategies for your actual reactivated customer, so it would be interesting to understand how you guys approach it.
So the poll should be up on the screen now. If you can send us here your answers, that would be a big help. I’ll just wait a few moments or just wait for some of those answers to come in. And when we do, we should have the results out shortly. I can see already they’re coming in live. This is extremely exciting. It’s like the Eurovision Song Contest. This is great. Okay. Thanks. Thanks very much, guys. This is really helpful.
So the vast majority of you–and this is not a massive surprise–the vast majority of you said that you’re reactivating customers from within part of your general marketing strategy. There’s a small part of you that do so kind of sporadically. And also very good to see that a few of you have a full strategy in place, only a handful of you guys on the webinar that actually do that strategy properly. But even for you guys, I’m hoping today to get to actually harness that and show you further the other, I suppose, 90% of the customers that is…of you guys that are sitting here on this call. I think hopefully, there should be kind of a real-world information ahead that we can go into.
Observing Customer Behavior Across Different Types of Business
So enough about that. Let’s first kind of understand a little bit about who we are. As you know, we are obviously Optimove. Business, we’ve been around since 2009. We’re working with about 250, 250-plus different brands, operators, clients across a wider verticals, so, you know, all the way from e-commerce through to insurance and banking, gaming, amongst many, many other things. And for each and every single one of these clients, what we are able to do is to provide kind of full customer marketing cloud that leans very heavily on data science. So we’re trying to bring the world of data science, understanding everything we can about the underlying customer, together with marketing automation, and hopefully trying to help our clients be much more intelligent with the way that they personalize their customer interactions.
We have offices in New York. We have offices in London. I’m actually from London. You may well be able to tell from my accent. So I’m initially from London. And the main office is in Tel Aviv, which is where I am now, which is why I have a bit more of a tan then you’d expect somebody from London maybe to have. The company now has over 120 employees, so it’s a big company and constantly growing; lucky to be a part of.
And these are some of the main brands that we work with, again, going across various different verticals. Now, the opportunity to work with some of these brands amongst all of the others that we work with has given us kind of a wealth of data, a real wealth of data so that we are seeing very, very interesting insights into the way the customers behave across all different types of businesses, all different types of verticals. And what we’re able to actually do as a result of sitting on so much data is we’re able to kind of look across industries, across markets, and see if we can pull out any kind of interesting conclusions. This, obviously…particular webinar is one that we feel is a very interesting conclusion of what we’re able to see. So we really have that opportunity by scanning so many different types of customer interactions with businesses to actually see how customers really operate and what we can pull out of that.
A little bit about me… I know I’ve got the webcam on, but there’s a picture of me at my wedding in happier moments. It’s probably been a little bit more downhill since that point. But I am Motti Colman. I am one of the directors of New Business, which is our fancy way of saying “Sales.” So I’m the Director of one of the sales teams here in Tel Aviv. More recently, I have spent the last couple of years working for William Hill. Some of you guys may know that. And historically, when I was back in London, I was really kind of a city boy. I completed my ACA in forensic accounting doing sort of very big corporate breakdowns and searching for lots of money before then moving to portfolio management. So I’ve kind of got a bit more of an analytical background and financial background to coming into this sales team.
Approach to Customer Segmentation: Lifecycle, Segmentation Layer and Micro Segments
Okay. So moving on, and hopefully getting us started in the right direction, whenever Optimove approaches any kind of data project, working with any particular client, we do build, for each one of our clients, a completely bespoke, a completely unique, predictive model for each client that we work with. But we will always follow the same framework. So the framework that you can see in front of you now is how we approach customer segmentation. We first start looking at the lifecycle of the customer, we then move on to the segmentation layer, and finally we move into the micro-segments.
Now, we, today, will be talking specifically about the lifecycle, and then we will go a little bit deeper into the segmentation layer. We’re not going to touch micro-segments. Micro-segments is what helps us understand predictive analytics. And I’ll mention a few kind of theories and interesting points from it, but we’re not going to go into too much detail about it. We’re going to really focus on the first two. What that should do though is also be now an opportunity for you guys to come back and tell us. This will be the second poll that we’re looking at. And what I’m going to be looking to see is what type of segmentation that you guys do, so kind of basing on the way that we talked about the micro-segmentation, medium segments, large segments, etc. etc. So again, if you could answer that back, it would be very helpful. We’re just going to see the results once again coming in live.
Okay. Interesting. Interesting. Okay. So we’ve got our…actually quite a lot of guys here, quite a lot of the people sitting on the webinar, are doing quite advanced, which is really nice to see, so some kind of a level of micro segmentation. The biggest majority of you guys are doing an intermediate, what we describe there as an intermediate, so looking at various different segments. It may be not as well organized as maybe the way that I set it out previously. You’re still definitely looking at segmentation. And they’re still quite a large chunk of users, so a fairly even split here. So quite a large chunk of you that are just kind of talking about large segments. That would also go back to that initial poll where some of you mentioned that you weren’t really targeting reactivated in any specific kind of way. So it probably lends itself quite nicely to that initial poll as well.
Lifecycle Stages: New, Active and Churn Customers
Okay. So what is that first stage, that lifecycle stage? So lifecycles tend to be things that most people are working with already in some which way, shape, or form. But ultimately, the lifecycle is simply trying to find where the customer is on their journey with you, so from that point where they register or become what you would distinguish as a customer all the way through to the lifetime of the last, you know, interaction they ever have with you as a business. So where on the journey is that customer with you?
Now, again, most people do take lifecycling relatively seriously. It’s not something new. It’s not something that we invented. And the way that most people approach those first three that you can see up in front of you now, you’ve got your new lifecycle, your active one, and your churn lifecycle. They’re fairly self-explanatory. So your new lifecycle deals with all of your customers that have just come in over a particular period of time. And then what happens next is that those customers either continue to reengage with the brand, and they continue to purchase things, and they become always an active customer, or they drop away and they become a churned customer. They could also drop away after becoming an active one as well. But these tend to be the three that most people focus on.
Now, when we’re… I’ll also be talking about lifecycles. And again, very specifically for this particular webinar, we feel that there’s a very interesting lifecycle. This is right here between active and churn, which we call the “reactivated,” where they’re back from churn lifecycle. So that tends to be the one that people miss, which is, of course, what we’re going to focus on today.
A little bit of a tangent, but just to kind of give you a flavor for why it is that we’re talking about this. So this is a little bit of research that I did myself. And I was looking at UK crime rates. I was quite astonished to find that for every 1000 people in the UK, there are 109, .96 crimes committed. I’d like to think of myself in the 881, but that is still… I mean, that, to me, was interesting. What was more interesting to me was that 26.8% of them reoffend. So as soon as the first crime is committed, the chance of actually committing a second one, you can see there, has gone up. And then if we jump to that final third statistic there, repeat offenders, so individuals who have offended, who have committed crimes more than 11 times, have continuous reoffending rates of up to 47.2%. So we can start to see that once people start to display a particular type of behavior, it may well be that that kind of snowballs on an ongoing basis.
Why is that interesting? Well, let’s look back here at the idea of churn, and the idea of activity, and the idea of these various different lifecycles. Now, again, this is from our ability to look across various different verticals and various different markets. But what we see, what we tend to see, is that when you have your new customers, the movement over a defined period of time, usually about a month… So after that month is up, and the customer being new, 60% to 70% of those customers naturally churn. Okay? So 60% to 70% are still…obviously, that’s pretty high, but that’s what we’re seeing is kind of a natural churn rate with your new customers.
If we then look at your active customers, we can see that actually only 5% to 10% of those customers naturally churn. Immediately that’s telling us, and it’s no… Again, this is no surprise, no secret to anyone, that your active customers, customers that are very engaged with the brand–the rate’s showing a lot of loyalty–will continue to retain that kind of level of loyalty on an ongoing basis. They will continue to come back. They will continue to spend money with you.
The Forgotten Segment: Looking at The Reactivated Customer as a New Customer
Now, what is a reactivated customer? I mean, we’ve spoken about it as conceptually. Well, what is a reactivated customer? So a customer goes to churn when they stop spending over a defined period of time. Let’s just say they haven’t spent with us for three months. We are now defining that that customer as churn. At some point, that customer then makes a purchase. So they appear within our database as a customer that’s active. Now, if we were to just simply look at that one interaction with that customer as a purchase, and we know that that’s a very old customer that’s been with us for about a year, we would, if we only looked at it on its own, we would think that that customer was the same as any of our other regular active customers. A regular active customer could have also made a purchase on that day.
If you take your mind back to the previous slide where we were talking about crime and reoffender’s rate, you will now see where kind of the connection comes. Because if you understand the history of this reactivated customer, i.e., that a week ago, two weeks ago, they were churned, and they hadn’t been creating any kind of interaction, no purchases for quite an extended period of time, you will now see that these customers, the reactivated customers, actually have a 50% to 60% natural rate of churn. That is obviously a lot more in common with our new customers rather than with our active. So even though in that one particular moment they’re displaying exactly the same type of behavior as our active customers, actually you will see from the way that they churn, i.e., the way that they reoffend, that these particular type of customers shows a lot more in common with our new customers.
If we then look at future value…so this was the bit that I said we’re going to touch upon some of our predictive elements. I won’t go into detail about how we do this. Please take it from me that our predictive analytics is exceptionally accurate, and probably this kind of centerpiece of why we do…how we do what we do. Looking at that kind of information, we have seen, again, if you look across the various ways that the future value or the expected future value will be of your customers, you will see that a new customer shows a lot more in common with a from-churn or a reactivated customer, which, again, is obviously much lower than active, once again building up this idea that our reactivated customer has a lot more in common with a new customer.
Understanding The Value of the Reactivated Customer Through Segmentation Layers
However, that only is as things stand today. The whole idea of this webinar and the reason that I think it’s such an undervalued segment is because the potential to do something with that former churned customer is significantly higher than the new customer. You’ve got two things at play here. Number one, acquisition costs for customers is significantly higher than Reactive entry costs. So immediately we can understand that the ROI or potential ROI from bringing a reactivated customer back to active will be higher because you should probably have to spend less to get that customer. Now, that’s the first thing.
But the second thing and probably more important thing is that there is a massive potential to do so, to reactivate those customers, because there’s been a very large history of those reactivating customers. There’s a lot of things that you know about that reactivating customer based on what they actually did with you historically, Sso what they spent, what they liked, how often they came, all of these various different pieces of information, the data that you have been collecting on those customers for that entire period of their lifetime is obviously much deeper, much wealthier to you, as a business, than a new customer who maybe has just made their first purchase yesterday.
So what we then need to do for that particular lifecycle, for that reactivated customer, is to look within the segmentation layer. So like I said to you before, the segmentation layer, that becomes our second layer of segmentation. And what we will be doing in the segmentation layer is looking to try and study each one of these individual customers through various different lenses, so looking at different types of behavior, like I mentioned before, given the wealth of knowledge that we have on that customer based on their historical behavior and trying to use it as a way of understanding how we could personalize their particular marketing and the engagement with them.
Okay. So some of the key segmentation layers that we’re going to focus on for this particular group of customers, the reactivated customer, the first thing we’re going to look at is the number of times that they’ve churned. We’re also going to look at reactivation sources, so actually what came about as a result of their activation, then purchase frequency, their past and present activity, and the highest achieved segment that they have had with us before. So let’s look at them all in a little bit more detail.
Reactivation Source: Organic vs Reactive-Based Customers
The reactivation source. Okay. So when a customer is coming back, of course, there’s only one real way that they could reactivate. They could reactivate by making a purchase. So we’re seeing them come back to the site, they’re making a purchase, and they are reactivated. However, what caused the purchase to take place in the first place? Was it by an offer? So did we send these customers some kind of discount, some kind of offer that made them think it’s a good idea to come back, or did that happen completely organically? Did they just, for whatever reason…they wanted to buy a shirt or some apples or whatever it might be, and remembered us and came back.
Now, we could see from, again, from all the various different customers that we’re looking at is that when a customer comes back to us organically, i.e. , they just, it suddenly switched on in their mind, they remembered that they like us, they have a much higher propensity, so twice as likely, to then move from reactivated into the active lifecycle naturally. Therefore, the sorts of engagement you need to have for this customer doesn’t need to particularly revolve around offers and, you know, various different things like this. Much more likely that you should be focusing on experience. So it may well be a newsletter that says, “Welcome back. Lovely to see you. Just, by the way, these are all of the other products that you might be interested in buying.” So they don’t really need offers anywhere near as much as our next group of customers, which we’re going to see…so the offer-based customers or the reactive-based customer.
That customer received some kind of offer. They were churned. And we’re targeting our churned customers. We sent them out a whole bunch of offers, and they looked at one of those offers and thought, “Well, that was very interesting. I’m going to come back.” Now, we’ve just seen from before, these particular customers have a much lower propensity to move to active naturally. So therefore, the way that we need to treat these customers is to continue giving them ongoing promotions, so things like discounts, buy one get one free, free shipping, whatever it might be. But we have to kind of continue on this journey of giving them offers.
Purchase Frequency: Guiding the Reactivated Customer to Multiple Purchases
Okay. The next thing that we can look at, purchase frequency. So when we’re talking about reactivated customers, reactivated lifecycle is, again, one of these lifecycles we look at that has kind of a window within which they’re being defined as that. So that window is somewhere between two weeks to a month. Now, the question is within that period–let’s call it a month–how many purchases have the customer actually made? So we can see again. We can look here, and we’ve got our line in purple looks at customers that made multiple purchases, and we can look at our customer that makes a single purchase, so a multiple purchase is two-plus.
Now, all of you guys won’t be surprised to learn, and I’m sure this is something that you think about probably quite often, that the idea of a one-time purchaser is probably the vast majority of your customer base, customers that have just some kind of one-time purchase and they never come back again. The same is going to be true here when we’re talking about the reactivated lifecycle, because it’s in that kind of incubation period over that course of the month. You will see here though that there is a dramatic, dramatic increase in survival rate from customers that make more than one purchase over that period of time. So, again, this should be coming into our mind, but what we need to try and get the customers to do within that incubation period, within that reactivated lifecycle, is to move from being a single purchaser to a multiple purchaser. And we’ll see it here.
So just, you know, kind of a few key stats for you. This is where the similarity, again, with the new lifecycle comes in, so i.e, like those one-timers, those first-timer purchasers. And we’ll see that 30% of our customers that make a first purchase don’t move… Sorry. Thirty percent don’t move from a first to a second purchase. Okay? So there’s a very, very large drop-off of customers post that first one-time purchase. However, once they have made that second purchase, 70% of them then move from second to third. So we can actually see, again, straightaway, that as soon as we get past that hurdle of a first-time purchase, it moves us into kind of a different category of engagement with that customer, the loyalty kind of…it expands almost exponentially. We have actually written blogs and other pieces just on that kind of topic alone, so it’s a very, very interesting way of having to look at it. Of course, what we’re trying to get into our mind is the idea that whatever we do during that period of time, if that clock is ticking towards the end of the incubation period and the customer’s only made one purchase, even if it means to give them some kind of very, very heavy discount, we should really be trying to generate that second and beyond the purchase.
Focus on First Time Churners for Better Results
Okay. The next thing we’re going to look at is the number of times the customers have churned. Now, customers have these kind of cycles with us where they become active, they may churn, they may reactivate, they may go active again, and they, you know, over an extended period of time, a year or two years, they have done in this process where they’ve left us and come back. We call that the kind of zigzag effect. Well, we don’t call that the zigzag effect. It’s called the “zigzag effect.” And I imagine that some of you guys have probably spoken about it at some degree before. Well, what should be interesting, too, is to actually understand the implications of this zigzag effect.
So you can see that customers who have historically only churned once and have now reactivated, only 18% of those customers will naturally move from that reactivated into active. However, customers that have churned more than once, so have actually been on this kind of zigzag, this cycle of bouncing backwards and forwards over an extended period of time, naturally have a 51% chance of moving from reactivated to active.
So immediately that should be telling you…and you won’t be surprised to understand if you look at on this next stage… Here’s our zigzag phenomenon. Here’s our 18% versus 51%. What you should be trying to do is to find a way of incentivizing your first-time churners much more than you would do to your zigzag churners, okay? So your customers that have come back on multiple occasions need less of an impetus from you in order to get them to become active. Why? Well, simply because this is part of their natural cycle with you. So they naturally churn, they naturally come back, they naturally become active, and they naturally churn again. And they’re very used to doing it, so they’ve done so multiple times. It doesn’t mean that they should be left alone. But if you already have one particular type of offer and you only want to give it to one particular segment when comparing these two, then of course, what you should be focusing on here is your first-time churners.
Engaging the Reactivated Customer Based on Past and Present Activity
The next thing we’re going to look is past versus present activity. So, of course you’ve got this whole wealth of time that we’re looking back on in the past, and we’re now, in this reactivated lifecycle, looking at the present. What that will be able to do is some kind of comparison between their entire history with us as a customer versus where they are today. So we’re going to effectively be able to understand the potential that they have based on past activity. If a customer’s only ever spent a total of £50, $50, €50, whatever it might be, then the fact that currently they only spent £30 with us during that reactivated period shouldn’t be really of any surprise. Vice versa if a customer spent hundreds and hundreds of pounds in the past while right now they have only £10, well, we’re seeing that there is the potential that this could be a customer that we could do certain things with. We’re also going to look at past and present activity, not just in terms of spend, but may be in terms of campaign responsiveness, etc. etc., so actually able to have a very kind of deep understanding of the customer based on what they did versus what they’re currently doing.
And taking that, what we’ve just described there was almost like a video, so looking at that entire lifetime. When we’re talking about this idea here of the highest achieved segment, we’re just taking a snapshot of the top that that customer received at any particular point in time. So we could be looking at historical seniority in terms of loyalty schemes. So you may have some kind of loyalty scheme where customers can be bronze, silver, gold, diamond, platinum, whatever it might be, so we can actually have a look back and see when they were active all that time ago in years, six months, or whatever it was, what kind of level of seniority they were. We could also look at their longevity. So they’ve been a customer of ours for two years, three years, four years, all of these kind of ideas of longevity, seniority, giving us something that we could hook into. Because what we’re able to do is actually remind the customer of that, “Hey, you used to be a diamond customer.” Or just refer to them again as a diamond customer. “Good to have you back.” “Hey, you’ve been a customer of ours for three years. So great to see you again.” All the time we’re trying to inspire this kind of historical memory of engagement that they had with us, of the loyalty that they had with us to kind of reengage again today.
Historical High-Tier Customers: Improving Reactivation Rates Among VIP Customers
The final one we’re looking there is historical high-tier customers. So it may well be you have some kind of VIP-type way of looking at a customer, of understanding your big spenders. And if you do have customers that fell into that category historically, then of course it’s, you know, very exciting if one of those customers comes back, you will understandably be pushing much more your marketing efforts into that particular customer, be it from a phone call to the customer, very heavy offers, or, you know, really overloading the way that you want to engage that customer, because, of course, they’re worth a huge amount of money to you potentially.
Okay. So, what we next do is take all of our information, and we’re going to try and turn it into an actual marketing plan, so use it as a way of actually targeting customers, engaging them in those very individual ways. Of course the objective here is to increase the percentage rate of customers who migrate to the active lifecycle. So each and every single one of those examples that I gave you before, the kind of…the way that we’re going to take that, the way that we’re going to turn this into something useful, is to try and build a marketing plan that’s going to improve their reactivation rates, so the principles, of course, that we’re going to look at, we’re going address reactivated customers on a weekly basis. We’re going to look at past activity and current activity. We’re going to differentiate between customers who reactivated organically versus offers. We’re going to take number of times for them to churn into consideration. The only one that I haven’t mentioned here is the idea of a…the purchase frequency, because that’s pretty self-explanatory. If they haven’t made a purchase once, then we need to send them additional campaigns.
Optimove’s Approach to Reactivated Customers
Now, this here is kind of a very nice graphical way that Optimove looks at our marketing plan. You guys will all have your own kind of scheduling capabilities or the way that you write things. The way that Optimove approaches things is this…it’s kind of on a calendar approach. So we built out a marketing calendar that lets you target very specific segments of the customer base. We have different campaigns that you can schedule for each and every single one of those particular target groups, as we like to call them. When we’re talking about target groups, we’re talking about behavioral types, so using things like segmentation layers that we just spoke about to filter out different types of customer. What we’ll start to see is customers move very naturally from one particular target group into other target groups over that course of their lifetime. But again, you guys will be approaching this in kind of the way that you do. Maybe you’re building out some kind of journey etc. etc., which is, again, another good place to stop.
I’ve been rambling on for quite some time now, so we’re going to lead to our final poll where we’re going to be talking about your marketing plan and whether it’s fully automated, how it is that you actually approach that. So, again, any points here would be very, very helpful.
The results are starting to come in. Okay. So that’s weighed very, very heavily towards basic and none. Okay. So most of you are running this on a kind of a fully manual process or a couple of basic automation tools that you were using to help. Nice to see that there’s a few of you that are doing this very, very advanced. And that’s great, because it will obviously help you, once you’re doing things on an automated basis. Of course everything we’re talking about is going to personalization. The more we personalize, the more groups we’re going to be working with. Without automation, that’s going to become very, very heavy and quite difficult to handle. So you guys that are looking, that responded to that “none” or “basic,” should really be thinking about the ways that you can start automating flows a little bit stronger.
Create a Marketing Plan Based on Segmentation Layers
Now, I’m going to breeze through because we’re getting towards the end of our time. So I’m here, but I’m going to breeze through this, which is quite a breezy part. Anyway, this is just kind of the way that a marketing plan could, or should, or might look. So again, it’s just taking into account all of those different types of segmentation layers. So you might decide on Day 2, that we’re going to look and use the idea of whether you reactivate it by a campaign or not, so the particular source. And you will just see it here, the way the flowchart works. So if, you know, if you wanted to reactivate it by a campaign, i.e., organic, then like I said to you earlier, we might give these customers a newsletter or a soft message, “Great seeing you again.” We’re showing them a hundred different products that they might be interested in. Alternatively, if they have been reactivated by a campaign historically, then we might need to continue to give them those particular type of offers again and again and again to turn them from reactivated to active. So it could be something like 10% off your next purchase.
Day 10. Okay. So we’re now a few days down the line. We’ve got a bit more information about the customer. Now, we’re going to look at the number of times churned–excuse that typo there–where it says reactivated by a campaign. It should say number of times churned. So we can see here that, again, if that number has been more than three, i.e., kind of that zigzag phenomenon that we spoke about before, so customers that are doing this naturally, our offer is going to be “Buy three, get one free.” But for the customers that have only churned one or two times, so not very used to actually churning, it may well be “Buy two, get one free.” Okay? So again, just taper that off slightly so that we’re pushing our more heavy offers on the customers that need it rather than the customers that don’t, not wasting good money on customers that don’t really need that anyway. Okay? So try to really maximize the ROI for many of these campaigns.
Seven days later, we might now have a look at campaign response. So this is something kind of entirely different. And when you guys are building out campaigns, I would imagine this actually becomes kind of the key behavioral element. Did they respond to a campaign, or did they not respond to a campaign? Optimove would be very happy to include those in as well. We just kind of layer it over the top of behavioral elements as well. So if it’s a no, you might want to send a reminder. If it’s a yes, we may be sending out a newsletter. So again, some kind of way of reengaging the customer based on how well they responded to the campaign up to that point.
Seven days later, we might now look at cost versus present activity. So historically, they were very strong with us, but currently they have kind of low present activity. So we can see that there’s high potential, so we might give the high potential customers that aren’t doing very much at the moment 15% off their next purchase. If there’s high present activity, so they’re doing very nicely but they were doing very nicely historically, we’re probably not that concerned with the customer. But we are quite excited at the prospect of what they might bring, so we’d say maybe they’re some kind of intermediate offer, “Ten percent off the next purchase on orders over X.” And then if they had low past activity and low present activity, then of course what we’ll be saying is we’re not really going to be expecting them to have very much activity going forward. So it may well only be a 10% off the next purchase. So again, tapering the particular offer to the particular type of customer, trying to use our best offers that we’ve got on the customers that could potentially maximize the amount spent.
And then we move all the way down to Day 45, if we go that far, and we’re going to have a look at additional activity in the last month. So we might look at how active the customer has been over that period of time, how many purchases that they have made, yes, no, again, etc., etc. So really trying to use the various different layers of segmentation. So it is probably a good point, kind of to recap and do a reminder, the way that most people tend to target customers and when they build out these kind of marketing journeys is that they will look at something almost as simplistic as campaign responsiveness only. I sent you a campaign on Monday. Come Friday, I want to check, did you or did you not respond to it, yes or no, and build out our next option beyond that. And that’s fine.
But as we’ve just kind of seen with all the various different elements that we’ve brought in this, many more kind of emotional reactions that are taking place, many more behavioral insights that we can actually understand, many different ways of targeting the customer based on how they have performed, how they have purchased, how they have purchased historically, numbers that they spent, etc.etc. So what we want to try and do is enhance the data, enhance the marketing actions that we’re taking, not just to look at campaign responsiveness, but to look at real behavioral elements.
So that kind of is the bit where I’m done talking. I’m more than happy, and I will be very happy, to come back again and answer any questions. So feel free to start typing them into the chat box if you can. I’m just going to have a sip of water. The questions will pop up in a moment.
Preventing Churn Before it Happens
Okay. So I’ve got my first question here, and the first question is as follows. Okay. So from Mike… Thanks, Mike. “How can we prevent churn in the first place?”
Okay. So that’s a little bit out…obviously outside of the scope of what we’re talking about because we’re talking about what we do with customers once they have churned. But it’s still a very, very relevant question. It’s certainly in the scope of what we do. So how can you prevent churn from happening in the first place? So it’s a question that will be answered very simplistically now, but very broadly, if you can think about it in the same terms of the way that we just approached this entire webinar, which is that we would focus on our active or on our new lifecycle and start to break it down to focus on very specific segments that show high risks of churn.
Now, again, we didn’t specifically look at our predictive modeling. But within the predictive modeling, that is, once we go down to the level of the micro-segment, we’re able to see all of the historical customers that did churn and all of their customer journeys that took place prior to their churn, and use that as a way of really anticipating when other customers are heading down the same path that are going to start churning and being able to pick very specific parts of their journey. So we may monitor things like activity. How inactive have they been over a particular time? Did they return any goods? Did they write in any complaints? So the customers, all of these kinds of red flags that we start to see lead to churn.
So one of the indicators that we’re able to present back to our clients is something which is called “risk of churn” or “propensity of churn.” And actually we’re able to target customers based on, dare I say, “I want to target only my new customers that have a risk of churn greater than 80%.” Or alternatively, I mean, that’s kind of a general way of looking at it, of using a very specific kind of miracle way of doing it. But you can also look at some of those segmentation layers. In the same way that we just broke down reactivated customers by a different segmentation base, we would do the same for our active and new, and try and determine which of those particular segmentation layers and underlying segments are actually showing very high risk of churn.
So that would be in answer to your question, we would be doing the same kind of systematic approach, sort of a breaking down of the customer, looking for behavioral trend. And then ultimately… It’s actually a very good question, because I…when I talk to my prospects and my clients about the way that Optimove works, or kind of the way that any key marketing or CRM strategy should work, rather than being reactive… And actually the behavior that we talked about today has been kind of reactive, but preventing churn is very proactive. It’s understanding the customers might become a type of customer that’s undesirable to you and actually be able to target them in a very specific way based on their historical behavior. So maybe that’s a little bit more long-winded than I expected, but yeah, that’s the answer.
Measuring Churn with E-commerce Platforms
Okay. Let me just see now. We have another question from Colin Thomas [SP]. Thanks, Colin. “How do you best measure churn with e-commerce?” So that’s also a very good question. So how do you best measure churn with e-commerce? The answer is that there’s no one particular way that we measure churn. It’s certainly not uniform across all of our different clients. When we’re talking about the measurement of churn, the first indicator clearly is the time since they’ve last spent. Okay? So that tends to be the kind of the key thing that we’re looking at. And we would build the churn lifecycle that would have some kind of general 45, 90 days, 120 days, whatever it might be, based on all the ways that the customers are actually operating. But when we think about churn, and we like to do so rather than on a kind of generalized basis, we like to look at churn factors for individual customers. So if we use that example of activity and activity-based, if I’m the sort of customer that is active, you know, once a week… Every week I come back in on Thursday and make a purchase. I pick up all of my shopping, all of my groceries every single Thursday and I make a purchase. If I wanted to come back in a week from now, it would be an indicator, given my frequency and my returns, that I was actually at a very high risk of churning. I might not have churned yet, but I may well be a very high risk of churn. Conversely, if there’s a customer that comes in once every month, if they didn’t come back in for a week period of time, we won’t necessarily be concerned. So we like to look at churn almost on an individual customer basis. We would be looking at things like activity, but it may well be other parts of the customer journey as well, things like the terms and complaints about some…that you collect that may well be kind of an indicator towards churn. But the actual churn itself will more likely than not be the number of days gone active. But like I said, it may well differ on a per customer basis.
Recommended Frequency of Campaign Launches
Okay. Do we have any other questions? Okay, Noam. Noam Zitkus, I think, is how we pronounce your surname. “How often do you recommend to launch a campaign?”
Again, another very good question, too. If we go back to the marketing plan that we looked at before, you saw there that we were also talking on a kind of weekly basis. That’s just for that particular principle. Ultimately, it depends on what kind of industry you’re looking at. If you’re talking in a kind of an online gaming environment or an online games environment, we may well be talking about campaigns every three days, every five days, something in that kind of range. When we talk about e-commerce, we’re looking at it much more in a kind of a 7 to 10 day… It tends to be more of the natural cycle.
It will also depend on which particular lifecycle we’re targeting. Active customers, which we have already understood as showing a very, very low propensity to churn, so they have…only 5% to 10% of them are actually churning from period to period. We might only be targeting our active customers once every two weeks, once every three weeks depending on their level of activity. Our churned customers, again, we’re not going to bombard our churned customers. We might run some kind of very heavy reactivation type campaign for our churned customers. That might be, say, once a month. But the reactivated customer or the new customer, those were the two lifecycles that we said have something called an incubation period. So that might be two weeks or a month period of time in which after that, they’re either going to become active or they’re going to become churned. So those are the two lifecycles on new and on reactivated lifecycles that we would probably target those customers more often than any others, so it would be something in the region about of a week.
More Customer Segments to Look at in Addition to the Reactivated Customer
Okay. Next question coming in from Chris… Thanks, Chris. “Besides reactivating, what additional segments should we be looking at?”
So from the first question, we might…we already spoke about preventing churns. So of course, those are the types of segments that you want to look at. So churn prevention campaigns become very good. One also is another kind of segment that you should for sure focus on, and I would imagine that you’re doing so already but maybe could be done in a more systematical approach, is the VIP whale or however it is that you refer to that customer. Again, this is something that the world of online games, online gaming, does very, very well because they understand that their VIP, their whales, are the big customers. They’re going to be receiving, in some cases, 90% of the entire revenue of the business, so they’re a very significant segment of the rest of the base.
When it comes to e-commerce, I don’t think we do that probably as well. So I would be looking at a very systematic way of looking at maybe early VIP customers or VIP customers that already exist, maybe tiering them, maybe having them look at their historical activity, just trying to find ways of constantly making sure that these customers are spending. Because they can outspend in a big way, you know, kind of big portions of the customer base, a very small proportion can outspend that, so that would probably be one that we would focus on.
Marketing Tools: Best of Breed vs Integrated Solutions
I am going to take the last question now because we’re almost out of time, and it’s from Penelope. “There are so many different tools in the market,” Penelope says. “What’s the trade-off between best-of-breed and integrated solutions?”
Again, a very good question and important question. It’s my personal view, and obviously, I’m going be a little bit biased here given where I’m sitting and where I’m working, but it’s my personal view that one should always be looking to take best-of-breed where commercially viable. Certainly the way that Optimove approaches this is that we are or we hope to be the best-of-breed from a data intelligence, customer modeling campaign automation orchestration kind of a space. That’s our superpower.
Where we’ve never really necessarily, I would say, pretended, or what we don’t claim to have is any kind of superpower our self is on the execution side. You will have many different marketing clouds that are exceptionally good kind of from an execution side, so very good at sending emails, very good with their SMS, with their approach to various different channels that they’re offering all at different times. So their marketing automation may be very good. Again, I would be always looking…for myself, I would always be looking to take best-of-breed where commercially viable. So in some cases, it’s not commercially viable, or people just like to have something maybe a little bit more basic, the groups lumped together. But where possible, I will be looking to be as smart, as intelligent as I can be and trying to take something that focuses, so it is very focused on the level of intelligence that they can provide. And conversely, I would also be looking to take different execution channels that could also provide best-of-breed for each and every single one of those.
So I’m being told that I have no more time. I can probably sit here and talk about this for hours. But thank you all very, very much. Just again, a bit more housekeeping, after this is finished–I’m going to sign off in a moment–but there will be a…just a little…another poll, I suppose, or just a questionnaire. It would be very, very helpful to me to ask if you could answer that. It’s going to ask you how well I performed. If you give me anything below a four, I will know. So please do whatever you can that feels natural and feels right.
Thank you, again, for joining us, and hopefully we’ll see you on another webinar in the future.