The Three Critical Components of Real-Time Personalization
Reading Time: 16 minutes
Overview: The Three Critical Components of Real-Time Personalization, presented by Andrew LeClair– Sr. Product Marketing Manager, Pega at RevTech Summit 2021.
About the speaker: A results oriented, self-motivated, and inquisitive marketer with a proven record of accomplishment in thetechnology space. My strong organizational, critical thinking, and communication skills allow me to providecustomized deliverables summarizing complex concepts in a succinct format.
Hi everyone. Thanks again for joining me here today.
Now, over the next 20 minutes or so I’m going to take you guys through the three critical components of real time personalization. And in doing so we’re going to talk about why these are absolutely critical to delivering the customer experiences that our customers are demanding from us.
And where I wanted to start today is actually with the story, an example, if you will, of what can go wrong when we’re not able to personalize in real time. So this beautiful city right here is Minneapolis, Minnesota. And this was back before COVID. When we were able to travel, I think it was almost a year ago to the date. But anyways, I was I was out in Minnesota for a client meeting. And if you’re familiar with the Twin Cities in December, you’ll know that in reality, they look a lot more like this, right?
We get to the airport, these are these are my colleagues and I plus this one random guy who decided to jump into our photo, but based on our facial expressions, I think you kind of know where where this one’s going, right? So, we’re at the airport, we’re sitting there, and I get this email. Now, my legal department made me blur a bunch of this. So I can’t tell you the exact name of the company. But I assure you guys, this is a true story. I can’t make this stuff up. So we’re sitting there, and we get this first email that our flights delayed. So no real shock, right? It was dumping snow at the time.
We all travel for business, right? Or at least we used to. So we get it, there’s nothing they can do. So what do we do? Right? We head to the bar, grab a beer, have some dinner. And while we’re there, a few more emails come in some more delays. Inconvenient? Yeah, absolutely. But you know, okay, at least this company is keeping me informed. But then we get this last one, right, which actually canceled our flight, so I get to be stuck in Minnesota for the night. So not exactly an ideal situation, right? I’m not exactly thrilled at this moment. But then this happens.
The next email that they sent me, literally an hour after my flight had been canceled. They hit me with this. Don’t be tardy booking your summer travel? Are you kidding me? You just delayed my flight five times. And now you’re gonna tell me not to be tardy. You’ve got to be kidding me. So obviously, for me, not such a great customer experience.
But this is exactly the type of thing that happens all the time when organizations can’t personalize experiences in real time. And that’s why you see stats like this, right? So at Pega, we went out and surveyed over 5000 customers to figure out, you know, what are their thoughts about how us companies are engaging with them? And the major takeaway from that survey was this, right? So two, out of every three customers believe that as brands, we simply don’t care about their needs. And if you think back to the example that I just showed you, it’s pretty easy to see why.
That’s just me, right? I’m just one customer as one example on one channel. But think about the scale of your business, right? You’re trying to manage this, for literally billions of these interactions for millions of customers across hundreds and hundreds of programs on dozens upon dozens of channels. There’s a lot of complexity there. And unfortunately, that leaves us exposed. There’s not a lot of chances for us to get this right. But there is a ton of opportunities for things to go wrong.
And it only takes one right? One poor experience and that customers out the door. And if you think about why this is the case, right? Why are we struggling to give customers what they want and make that consistent and relevant across channels is because all the technology that we had today, it wasn’t built to actually support customers, right? It was built to help businesses like ours push products, to big segments and large batch campaigns. And it was all built around the concept of us picking something that we wanted to talk to the customer about way before that interaction ever happens. And one way or another the story go a little something like this, right?
So for years marketers, people like me, right? We get a brief from one of our stakeholders, and it was our job to go sell that thing, whatever it was. And we say, sir, yes, sir, right away, we’re on it. We run off and we gather some data, right, we would do some analysis. One way or another, we end up creating a segment, let’s say, you know, 100 people or so. Then what I would do is just blast this thing out to everyone all at once. Then I would, you know, map some complex thing using a bunch of sticky notes where, you know, if someone clicks, wait 37 minutes and 28 seconds, and then send them this or do X, Y, and Z on this other channel.
If they don’t respond, hey, I’m going to put them off in this other segment. Then I’m going to blast those guys at some other arbitrary point in time. I would do this literally hundreds of times every single month. But here’s the problem. The problem is that the average campaign in any market gets less than a 1% response rate. That means 99 times out of 100. They’re still ignoring everything that we’re saying. There’s literally no other profession in the world where we can fail and 99 times out of 100 and still be celebrated as a huge success.
This go–to–market is absolutely 100% broken. It just doesn’t work anymore. And in times like now, right with so many people in such dire straits and so much, you know, huge social issues dominating the headlines, it’s really, really a bad optic. And so if that’s the case, right, let’s talk a little bit about why this is broken. Why does this kind of go to market no longer work. And the reason this stuff fails is because most of what we want to push has no real relevance to the customer, right? We’re just out there trying to hunt for more money hunt for more revenue.
Honestly, guys, people know that right? No matter how creative we get, they’re gonna ignore it. Heck, they’re running from it right? Opt out, unsubscribes are at an all time high, because at the end of the day, it’s all based on what we want to sell. It’s not based on what people really need. And it’s pushed out when we feel like center something right, or when a overly simplistic business rule tells us to do it, right. It’s not about people, it’s not about their needs.
It’s sort of soul crushing, honestly, because this kind of approach has a complete lack of relevance, a lack of context and timing, and empathy, which if you think about it, are really key aspects to personalization. And that lack of personalization, you know, being able to give a customer what they need when they need it, that has a price, right?
When what you say looks and sounds and feels like whatever everyone else is saying, You’re basically putting a cap on your relationship, right? And you hit this wall, this ceiling that you just can’t get past, right, you can look up, and you can see all this wonderful potential, all these things that you could do with this customer journeys you wish they would go on, right? But it’s like they say, in Boston, where I’m from, you can’t get there from here, baby.
All of us, right, we get here, eventually we hit this wall, every single marketer has ever right has faced this. If you want to get past it, you’ve got to start to change the way that you think about customers and the way that we go to market. That’s exactly what we’re here to talk about today.
If that’s the problem, how are we going to fix it? Right? What do we have to do? The first thing we need to do to deliver on that promise of real time personalization is centralize our decision making. Arguably, the top barrier to getting this whole personalization equation is our tech stack.
Being a tech conference, you may have seen this before. But what this is, is all of the solutions that are available to us right now, across the customer engagement space, there’s literally 8000 unique solutions out there at this exact moment, which we can use to engage our customers. And on the surface. That sounds wonderful, right? But in reality, what we see is that’s about 8000 too many because that’s 8000, disconnected siloed applications, each of which has their own brain, their own rules, their own data models, their kind of unique ways of understanding and engaging with customers.
And even if these things are from the same vendor, we all know that none of them were built to work together, right? They were each built to help businesses like ours, you know, sell products to customers, in big batches, large segments, right? They weren’t designed to be agile, and help us solve customer problems during times like these.
So, let’s dig into that for just a minute. Let’s double click down. Now, obviously, you’re not gonna have all 8000 options, right? But as we’re out there, we’re engaging with clients. Here’s some common buckets, if you will, some stuff that we typically see. And obviously, this isn’t exhaustive, right? This is just meant to be an illustration. But what’s key about this, and really, what I want to highlight is the fact that there’s nothing that combines these things together, right? Nothing that stitches, all of these disparate systems with this approach, we’ve got data all over the place, right?
Our systems, and even our people, they’re not connected, there’s a lot of complexity here. You’ve got inbound over here. Outbound is over there pays off on some Island somewhere, nobody knows. Not to mention, anything else that touches our customers, like our service department, maybe billing, right. And when we do engage with a customer. That learning often gets lost, right? Or even if we’re able to capture it, from a technology point of view. We’re just not in a position to really take advantage of what we know and what customers are telling us in real time.
So we’re going to fix it. And here’s how it has to work, right? There can only be one brain, one central decisioning authority, the power, all of our engagements. And what this brain does it sit at the center, all of these channels, right, all of these different applications, and the clicks data from them, like, what’s the customer doing on the website? Or did they call us? Did they tweet us today? Do they open an email? And then it combines that data with all the historical information that we have about them? Right, that’s in their profile, it’s in their interaction history.
Now, what it does is based on this data is look at each customer in each unique moment and determines what’s their context, right? What’s their situation? And is there anything that we can do to add value? Is there a next best action? And again, that might be a sales offer, but it could also be something like a retention plan, a service task and nurture effort, right? But whatever it is, we figure it out in real time using things like AI adaptive models, machine learning, and then deliver it back out across any channel inbound, outbound, own, paid, agent assisted, you name it. And that entire process needs to happen in less than 200 milliseconds.
So, alright, you just heard me talk about how, you know, a centralized brain enables us to engage our customers using this concept of next best action. And what I think is really key about that is the fact that it’s his next best action, right? It’s not just next best offer. And when you think about kind of what’s at the heart of personalization, it’s a person, right. And if we’re going to engage each customer, as an individual, it really requires us to get beyond that traditional product, push mindset, and instead be able to pivot to things like service or retention, in the moment as somebody whose needs are shaped are shifting and changing.
Let’s talk about for just a minute how we’re going to do that. Once we’ve got that centralized brand in place, we’re able to do things like this. Instead of just looking at sales offers, we can actually start with a list of all of our potential actions, things like service, retention, nurture, maybe a hardship message, but whatever that may be, right, we’re starting with everything we could possibly do for this customer. In the moment, this could be five things we think customers do 500, 600, 1000 different things to talk to somebody about. But once we’ve got that list, here’s where the math here’s where the AI comes in.
So how exactly do we know which one of these is going to be the next best action? Well, the highest level what the next best action is, is the one that best balances, what’s best for the customer, with what’s best for our business, all within the context of a given situation.
So let me show you how that works. Once we’ve got the list of all of our potential actions, the first thing that we want to do is determine how likely is each one of those actions going to be desirable to the customer, right? That’s the key. That’s the propensity score. And then we want to look at, okay, what value is each of those options going to generate for us? Meaning if they read the content, or they accept the offer? What’s the impact, right? Does it give us more revenue? Does it reduce our cost?
Then we want to give ourselves some levers to adjust things up or down, based on our current business situation. Like, if we’re behind plan, right? Maybe we ramp that one up, or if we’re low on inventory, we can ramp it down, something like that. But ultimately, the action that we take, right, the next best action is simply the one with the highest total, P times v times L. But here’s what’s key about real time personalization.
It’s the fact that every single time a new piece of data comes in something like browsing history, or content clicks or event data that’s coming off the network, or maybe a mobile device or a call center, right? It can be anything, right? But what happens is the instant that new data comes in, we automatically re score that person’s entire profile and calculate a new next best action. So good example. Right?
So, Amanda, here is on our app during the lunch break, and she’s clicking on a dozen different pages. And what happens is, those clicks are streamed right into that centralized brain that we talked about earlier. And we’re using them to rescore her profile, and see if that activity is telling us something new, like, maybe she’s trying to cancel her contract, right?
She becomes a retention risk. So, what we can do is instantly recommend our next best retention action and surface that right on the mobile app, right as she’s online, right? We don’t have to wait for two weeks for our next batch campaign to run and she’s already left. So instead, we’re triggering that action right then. So it gets to her in the moment when it can actually do some good.
But let’s say you know, 10 minutes later, she swiped her credit card and makes a purchase. What we’ll do is say, okay, does that purchase, give us the opportunity to add some additional value, right, maybe there’s something that pairs really, really nice with what she just bought. So again, that purchase data comes in, were decision her. And now our new next best action isn’t a retention plan anymore, right? We shift to a rewards offer based upon that new context.
Or if she calls in, right, she talks to a rep and they have a conversation. And she actually rejects the offer. But you know, she tells us, hey, post COVID, I’m planning a trip, that new data and it opens up an entire suite of things that we can talk to her about to make sure that trip is awesome. You know, for example, we can send our tips and tricks about how to use her products internationally. Or maybe we can nudge her to avoid service and rent. But while these are few, you know, just a few basic examples.
What I think is key about this is that it’s not just about selling anymore. If we’re doing real time personalization correctly, we’re doing things like this, right, where we’re able to determine is now a time to sell, or serve or retain, or maybe we haven’t seen him in a while and we just want to nurture the relationship. Or maybe it’s a hardship message. Or maybe you know, now’s not the time to engage, right? It’s really hard for us to add any value in this situation. But whatever it is, right?
We’re able to pivot between all these options in real time and pick the best one for them in the moment with every single customer that we engage with. But it’s not enough to stop there. If you’re going to personalize the customer experience in real time, it’s not enough just to know what to do. We’ve got to be able to get that decision out to the customer during their moment of need. We’ve got to make sure we can adapt, right, we can shift this experience as theirs, as their needs change.
So let me show you kind of how that would work. To go back to our previous example, if you remember, our P times v times our next best action screen, let’s play out the scenario here that our friend Amanda is currently going through a hurricane. If we were going to use kind of the traditional approach, right before, we’re leveraging that centralized brain and doing real time, personalization, this is what Amanda would see.
Even though there’s a hurricane going on, it’s really hard for us to move and build out a new strategy, and figure out, you know, who that applies to, and make sure that they’re the only ones who see it. So instead, we’re probably just gonna stay with the status quo, right, we’re gonna default to a sort of static experience, which means, we’ll show a credit card, or maybe another major offer up in the hero banner, because as a business, that’s something that we want to promote.
And then down below we’ll show, a whole other series of offers in our secondary containers. Usually things that there are other top sellers, stuff that we think. Hey, this is something people typically want something typically need. But organizations, we’re trying to personalize this, but sort of like we talked about throughout. They’re basically doing that with segments, right. Rather than doing anything one to one. And they’re updating this in batch. So it’s not really about Amanda, right. It’s sort of geared for people that kind of look or kind of feel like her. It’s showing them content that they may have been interested in whenever the last time they updated the data was.
So the way that we need to think about this is that every aspect of this experience needs to be personalized, to hurt right to Amanda. And what we need to do is look at each section of this page, just as a blank container, right? In other words, what this really is, is a series of smaller experiences, each of which we need to populate with the next best action. So what happens is the moment she hits the website, right, it calls into that centralized brain, we give recommendations about how to fill all of these containers. Again, what’s relevant to her, what should we show based on our P times V times L.
So in this instance, since Amanda is in a hurricane location, those messages are critical, right? They’re very relevant for her right now. So that’s what she’s going to see up in the hero banner, right? We’re not going to show her any offers. And instead, we’re going to focus on helping her get through this any way we can, like, you know, directing her to the mobile app or self-service, right. So she can start a case there, and then transfer over to a rep when one becomes available.
We can do things like show our safety tips, because that’s the right thing to do, right. And since in the past, we’ve noticed, you know, customers that are in a similar situation, pretty frequently they get separated from their wallet, or maybe their purse, so but they still have their phone. So we want to make sure that she knows she can get cash from any ATM even without her card. And well, that approach, right works tremendously well on inbound channels like the web.
It works the same way for outbound too. Take email, for example. We simply treat every area of that experience as a container for a next best action decision. And unlike traditional campaign management, or journey mapping tools, which pretty much just use business rules to say, you know, if customer does x, then do y and send this message at some arbitrary point in time. Instead, what we’re doing with this is we’re constantly monitoring that customers context, right?
So every time there’s an interaction, or a new piece of data comes in, again, score them through all those propensity models to see, is there a relevant next best action for them? Is this an opportunity for us to add value, and if it is, great, that’s when we’ll engage. And we’ll do so with that personalized next best action.
And beyond email, right, this works the same way in places like the call center. Where being able to adapt our approach delivers massive amounts of value. When we can ship what we’re going to say to that customer based on what they just told us. Half a second ago, or on a paid platform. Right directly connecting to channels like Google or Facebook or Instagram or YouTube.
So we can engage folks one to one without being beholden to a DMP, or a DSP, or on a mobile channel or something like a chatbot or an IVR. But honestly, guys, what I what I hope you’re getting from this is that at the end of the day, the channel itself really shouldn’t be as important to us, right? Because they’re all using that one brain, that one central decisioning engine to power all of these experiences, right? Keep them connected, and ensure that we’re giving our customers the best and most personalized next best action experience possible.
At this point, you’re probably thinking to yourself, okay, Andrew, who’s doing this right, who are some of the companies that are actually seeing the benefits from this approach? And there’s lots of them. You’ve got guys like Nat West, formerly the the Royal Bank of Scotland who have transformed their entire organization around this concept which they call Personology, which I think is a pretty cool term, but it’s really about moving out of that purely sales driven culture to using data using real time decisioning to make sure that every time Talking to a customer.
It’s about what’s right for them, not just what’s best for their business based on everything they know about that customer. And as a result, they’re actually engaging 35% less. Now, typically, if you tell a marketer, hey, Andrew. You’re going to engage one third less with your customers. I’m going to have a heart attack and pass out. But honestly, because they’re only talking to folks when there’s something relevant to say. And when they can add value. They they’re seeing things like a 6x increase in response. A 10, to one ROI, or folks like CBA, Commonwealth Bank of Australia. Who are doing these next best action decisions 24 million times every single day, across 18 channels.
That level of personalization has led to a 12 point, lead and NPS over their nearest competitor or folks like sprint, right? They went live with this approach in 90 days, reducing their churn to the lowest levels in company history, while also increasing upgrades by 8x! 800%!
Or even in industries like healthcare, you’ve had folks like Australia and the Netherlands having an 85% save rate, which is absolutely massive. So, guys, these are transformational results. And what’s key about this is that you’re going to start seeing value quickly. And then you’re going to earn the right to take on more now. A lot of times after we walk through this, people have a lot of questions, right.
So what we did is we went ahead and wrote a short paper that breaks all this down. And a little bit more detail, getting into a deeper level. So if you’re interested in learning more. You can download this right from your console. Or right from the URL you see on your screen here right from pega.com. And with that, I just wanted to say thank you again. I hope you guys enjoy the rest of RevTech.