Hello everyone. Welcome to The Age of Individualization:
How Identity is Transforming the Retail Experience. I am Joe Doran, the Chief Identity Officer
at Signal. In today’s webinar, we’re going to explore customer expectations in the digital
era addressing how brands are focusing on individualization to meet and exceed the demands
of an always on, omni-channel consumer. In particular, we’re going to dive into how brands
are uniting and leveraging their first-party data to really gain a complete view of their
customers, improve customer experiences, and optimize marketing spend by implementing more
effective ad retargeting strategies. We’re joined today by distinguished speakers,
included, Brendan Witcher, Vice President and Principal Analyst at Forrester. He’s joined
also by Dean Murr, founder of Programmai and a former Senior Programmatic Manager at ASOS.com.
Now before I turn it over to Brendan and really get into the webinar, I want to go over just
a few housekeeping items. First, the webinar is being recorded and this recording will
be sent to all attendees within the next 24 hours, so you’ll have a chance to watch
it again and again. It is there for you. All attendees have been muted. So if you have
any questions, please type them into the chat pane because near the end of our time today,
we will have a Q&A session and we’ll try to answer all of your questions at that time.
We also greatly encourage you to keep the conversation going on Twitter by following
our hashtag, #IdentifywithSignal. Now I’ll turn it over to Brendan to get this
webinar truly started. Today, Brendan will discuss how individualization is revolutionizing
the retail experience for consumers, online and off.
Thanks. It’s good to join you today. Glad to be a part of it, and excited to talk to
you about individualization and personalization. It’s a topic I’m extremely passionate about.
I find it a fascinating area, and it has been for a long time. We’ve talked about personalization
for a while. We’ve talked about what it means to personalize things. Back when I was a practitioner,
we used to think about personalization all the time. But since I’ve joined Forrester,
I’ve really kind of reexamined that and took a hard look at what’s causing the organization
today to say, “We’re doing lots of personalization,” but consumers may be saying, “That’s not really
hitting me. I’m not feeling like it’s personalized.” Brendan Witcher: We’re going to talk a little
bit about today what are the things that are causing challenges. But I want to start off,
real quick, by talking about what’s causing customers to make the decisions to shop with
the brands that they’re shopping with, talking about how this identity is transforming the
retail experience for consumers, and how that really is the benchmark for being able to
do these kinds of things, and the facilitator of being able to deliver personalized experiences.
One of the things that I like to do is do speeches, and I do quite a bit of them. I’ve
been doing this speech for a while, around a year, where I stand up on stage and I ask
the audience, “How many of you shop in a grocery store?” Of course, everybody says yes. I ask
them, “All right, I want you to imagine yourself going into your local grocery store having
to do an entire week’s worth of grocery shopping and they have no shopping cart.” And I don’t
mean they’re all being used, I mean there’s none there. You ask the manager about the
shopping cart, “Where are they?” And the manager says, “We couldn’t figure out the ROI of a
shopping cart.” Of course, everybody laughs a little bit,
and then I say, “Okay, well, how many of you will continue shopping with a grocery store
that doesn’t have shopping carts?” And I’ve never had any hands go up in the air.
And I’ve presented this to about 30,000 people at this point. So what is the value of a shopping
cart? What is the value of getting the experience right? For those 30,000 people, it was everything.
They were all willing to leave that grocer because they didn’t have shopping carts.
Now, you may say, “Well, what’s the big deal in that?” There’s two things I want to point
out. One, this grocer, a near grocer, has every product you need. Every product. So
when I try to point out to companies is that you feel like the assortment that you have,
you feel like the products you carry, the services you offer, you feel like they matter;
they don’t. The reason they don’t anymore is because they become table stakes.
We know you have products. We’re digitally savvy. We also know we have choice, whether
it’s a coffeemaker or light bulbs, a 529 college savings plan or motorcycle insurance. We know
that companies have products. We know that we can get them at a hundred different places.
So your assortment isn’t what’s going to keep me, it’s got to be the experience.
The other thing is, I’m completely willing to leave you if you don’t give me the experience
right. What I remind the audience when I go through this example here is that how many
experiences did this grocer get wrong and all of you are willing to abandon them? And
of course, everyone comes to the realization it was just one. One experience. That’s what’s
driving consumers today. The other thing that’s causing this shift
in customer behavior and why personalization matters is that each time a customer is exposed
to improved digital experience, their expectations for all experiences are reset to a new higher
level. What does that mean? Well, you can find the meaning in this, basically, in what
I don’t have on the slide, which is the word competitor. You don’t see that I say they
compare you to your competitors. An example of this would be, for you on the
call, you could go to a banking website, for example, and you could find what you needed,
and ask a virtual agent a question and, you know, that would be great. But if you go to
your insurance website, you can’t find what you need, you gotta call to ask a question.
Are you actually disappointed? Yes. That’s because, as consumers, we don’t compare
you to people in your industry. We don’t think that way. That’s the way we talk, people on
the call. We talk like that. Consumers don’t talk like that. They just say, “Well, it was
a great digital experience,” right? And we’re influenced every day by Uber, and Starbucks,
and USAA and all these other companies that are doing really innovative things. But they’re
not just influencing their industries; they’re influencing everybody’s expectations for all
industries. So the mistake I see organizations making
all the time is, “Hey, let’s just look at our competitors and see what they’re doing.”
Huge mistake. What you’ve gotta do is measure up to what customers are expecting from you
as those expectations are being driven by every experience.
Now, the last part I’ll talk about is, I’ve talked about the importance of experiences,
I’ve talked about where those experiences are being generated, but now I’m going to
talk about something that’s kind of interesting. Now, we have lots of stats at Forrester and
I could have thrown any one of them on the screen. But this one in particular I find
fascinating, 61% say, “I am unlikely to return to a website that does not provide a satisfactory
experience.” Some of you may find that interesting, some of you don’t.
But I didn’t actually put that on here because of the stat. I want you to ignore the stat
for a minute. I want you to look at the words for a second, because the words are the most
important part of this. Satisfactory customer experience. What’s interesting about those
words is to each customer satisfactory is going to be different.
Let’s just take retail for example. For a certain set of customers, to be able to buy
online and pick them up in store creates a satisfactory experience. For others, it’s
easy access for promotion. For others, it’s same day delivery. For others, it’s being
able to use comparison charts. The other one, it’s being able to look at a video of a product
before they buy. For others, it’s product recommendations. And on, and on, and on we
go. I know everybody wishes Forrester would get
on a phone call and say, “Here is the silver bullet.” There’s no silver bullet. There’ll
never be a silver bullet, unless all of us are starting to be created equal, which will
never happen, which is why there is no one solution for engaging your customers. Customers
are going have those shopping carts, all of their own personal shopping carts … I’m
putting little air quotes up … their own personal shopping carts about the experience
that they want, and they’re going to be individualized. They’re going to be unique for each customer.
All of us are built differently, right? Not all men are built the same, not all women,
not all people are in certain income level or in a certain zip code, or even the last
10,000 people that bought a certain product will all behave the same going forward.
This has led to a state of hyper-adoption, number one. So we all have Facebook accounts.
I know, if you admitted it, you’d say, “Yes, I probably do.” It leads to a state of hyper-adoption,
but it also leads to a state of hyper-abandonment. Right? Not many people maintain their MySpace
account anymore. This is an example of this flow of customers so rapidly from one thing
to another. It’s an example of why customers are the way they are today. They know they
have choice, they’re all generating on experience, it’s all about how the experience work for
me, and satisfactory is defined by what’s satisfactory to me.
Netflix, hyper-adoption. Blockbuster, hyper abandonment. Many people on this call probably
use Uber on a regular basis. This is how fast customers are moving. Now I say all this,
you say, what this has to do with personalization? What I’m trying to show you is, if you don’t
get that experience right you, don’t have time to sit around and test and learn, and
maybe we’ll work on this in 2019, 2020. I’ve shown you, you get one experience wrong, customers
could just leave you. It could be their shopping cart moment saying, “You know what? I’m out
of here.” What are companies doing? So 77% of customers
have told us at least that they’ve chosen recommended or paid more for brand that provides
a personalized service experience. A lot of digital professionals are hanging their hat
on that huge number and saying, “You know what? We need to do personalization more.
This has clearly got traction with customers when we get it right.”
I would agree. As someone at Forrester who sits here and looks at data all day long,
I will tell you that is a huge number. We don’t see consumers responding to something
like that, except for maybe lowest price on anything, like royalty programs or other things,
they don’t respond to that degree. So three out of four is a really high number.
What are companies doing? They’re delivering personalization everywhere. But back when
I was in the business, we used to say, “Well, it’s product recommendations.” You know what?
It’s not product recommendations. It’s more about content and experiences. Even in the
retail space, for example, we’re seeing a lot of personalization happening in the store.
Now why do I show you this? What I’m trying to show you is that, as consumers experience
more personalization, as more and more organizations start leaning in and offering personalization,
you’re going to start to see customers expect more personalization. It’s that high jump
slide that I showed earlier, that as consumers get exposed to something, they expect it more
and more. So you got to be out in front of this personalization thing right now because
it is very hot. But digital leaders today recognize that maybe
the way they’re doing personalization isn’t quite working, maybe there’s gaps in the way
they’ve done it and the way they’ve done it for years. So while 89% of organizations say,
“We are investing and personalizing the customer experience, 40% of consumers say, “Information
I get is irrelevant.” Only 40% say this is irrelevant to my tastes and interest.
So what’s causing the gap? I would even argue, is it personalization if you get it wrong?
No. It’s not really personalization if you get it wrong. If you get it wrong, you can
send them to anything. It’s only personalization when you get it right. So while 89% say they’re
doing personalization, I would argue only 40%, in theory, are getting it right or doing
personalization, because it’s actually more about what the customer experience is.
Here’s what’s causing some of the problems. Segmentation. Now I am not a naysayer on segmentation.
I used segmentation for years. But I want to show you some of the flaws of segmentation
that you may not even realize happen. I buy a toaster, you compare me to 10,000 other
people that also bought that toaster. So maybe 3,000 people bought toast tongs. Okay, great.
What about the other 7,000 people? Well, maybe 2,000 bought a blender, 1,000 bought a rug,
500 bought something else, 200 bought something else, all the way down the line.
Notice none of them were above that 3,000 that bought toast tongs, but they’re not a
majority. They’re not a majority, but they are major minority. In other words, they’re
highest number that we can get. So what we do is we show them toast tongs as a recommendation,
for example. Well what does this really mean? It means we got it wrong most of the time,
doesn’t it? Right? That’s the problem. I’m being generous here.
I’ve seen things where it’s like segmentation, the major minority is 5% of customers, and
10% of customers. That means you’re getting it wrong most of the time. This is why you
don’t see 30, 40, 50% less. You see 2% less, 5% less, 6% less, very small numbers. It’s
because you’re not getting it right. It’s not that you’re not doing it, it’s that you’re
not getting it right often enough. So you’re providing the wrong experience for
most of your customers. I would argue that most retailers don’t want that to be an objective.
Right? So I’m not just saying segmentation is broken and it’s a weak way of doing things.
I’m saying it actually provides the wrong experience. Yeah. I bought yoga pants. Guess
what? I’m not a soccer mom. Don’t talk to me like that. Right? You could offend customers
very easily by trying to use segmentation in ways that aren’t right.
The other thing we try to do is use single data points. Now if you have any background
statistics at all and understand statistics, single data points have a statistical significance
of zero; they tell you nothing. Again, if I buy yoga pants, it doesn’t make me a soccer
mom. It doesn’t tell you anything about me. Just because I bought an Apple cable, it doesn’t
mean anything, or drumsticks, or what have you. The things that I buy, the things that
I do, maybe it’s one thing that I dwell on, or look at your website one time, doesn’t
mean necessarily that you know me. So knowing one thing about the customer, statistically
speaking, math will tell you, you really know nothing about the customer, in that sort of
scenario. The last thing that we see with why personalization
isn’t very effective and consumers respond so poorly is that a lot of companies think
this is personalization. They say the day of the week, they’ll put my face on their
app, or they’ll say, “Hey, happy birthday!” Here’s the problem with that. Nobody comes
to your website to hear that it’s Friday, see their face, or get told happy birthday.
That’s not why they’re there. They get there to shop, they get there to learn about products,
to have a more efficient experience. What’s really interesting is that there were
teams that worked on this kind of stuff. I mean, this was a project for somebody, right?
CEO probably had to spend a week trying to decide the photo for this happy birthday message.
Guess what? I’ve never been on an earnings call, not once ever, where someone got on
the call and said, “Hey, we said happy birthday and sales went up 5%.”
There’s a reason that doesn’t happen, because this isn’t value oriented. Personalization
needs to deliver value to the customer. It has to be relevant and it has to deliver perceivable
value to the customer. It’s a waste of everyone’s time to do those sorts of things.
This is what’s creating the misalignment. So now you’re saying, “Okay. Well, you better
get me to the solution because, you’re right, I’m doing a number of those things.” So here’s
what organizations are doing. They’re moving to this next evolution of personalization,
which will rely more on the pillars of individualization rather than segmentation. Because they’re
recognizing, “Hey, the things I’m doing are flawed. I got to find a better way.”
How do we do that? Well, I’m gonna get to that. But before I do, it’s really, really
important for me to get to the unsexy stuff first. Why? Because it’s the unsexy stuff
that makes the sexy stuff work. To start, you can only truly personalized experience
for consumers you recognize. If you don’t know it’s me, you can’t personalize the experience.
Now that seems like a well, duh thing, but it’s amazing to me how many companies don’t
know who their customers are. They haven’t taken the initiative to say, “We need to get
better data about our customers. We need to incentivize them to log into our website,
to incentivize them to identify themselves in a physical location, or create tools that
they can self-identify.” The more you understand each customer the
more efficient your engagements will be. Your personalization solutions are just engines.
You know what goes into those engines? Gas, and that gas is customer data. You can have
high octane fuel, like the day of the week I like to shop, my favorite color, my favorite
store associate. If I buy shirts and pants together, I always buy two shirts together.
This is high-quality data. But my age, my gender, my zip code, these are not the kinds
of things that we need be working on. Those are not the kinds of things that drive good
experiences. So companies today are moving up this maturity
cover view and data. They’re trying to be data-led. They’re trying to say, “Hey, let’s
not make an assumption about all men. Let’s not make an assumption about people live in
zip code. Let’s let the data tell us the way people are behaving at the individualized
level and that’ll help guide our strategies.” You’ve probably heard the term “data-driven.”
Yeah, that’s okay. But usually that’s just to support decisions you’ve already made in
the organization. Data-led organizations are the ones that actually let strategies be driven
by what data is telling them in building their organizational go-forward plan around that.
Fortunately, we’re in a state where customers want to be understood. Over 70% of consumers
say “I know companies are collecting data about me. They just wanted to create better
experiences. Forget that Facebook thing, that was nonsense. I mean, nobody got upset about
the fact that Chase was collecting data, but they got upset about was how it got used.
And that’s absolutely true. Today’s organizations need to use data in the proper way, but fundamentally
they still need to collect it first and understand how to collect it and what kind of data matters.
What does it look like to get that 360 degree view of the customer? Well, unfortunately,
it looks like this. Quite complex, and it’s a lot. I would argue most companies today
focus on profile and behaviors when in fact so much good information and more relevant
information is about sentiment, content, affinity towards things, attitude towards things, and
even the context that I’m in. Are you personalizing experience because if
you know where I’m at, who you know who I’m shopping for, am I in a store, am I online,
what did I do last week, what day of the week is it compared to when I normally shop? These
are the kinds of things we have to understand about the customer to stay relevant and address
intent. How do you do that? Again, as I said earlier,
deliver perceivable value. Forget the happy birthdays. You’ve got to be able to say, “Hey,
we’ll do this for you,” and that’s the way you get customers to share data. There’s lots
of examples, which I’m going to show in just a minute, of how companies are doing that,
delivering perceivable value and getting good customer data out of it.
So it’s all individualization strategy, again after you’ve done that unsexy work of working
with data, is built upon key tech investments that have unique characteristics. One, customers
identified and treated as individuals using rich customer profiles. So their ability to
go in and say, “Hey, we know you. We know that you buy online, pick up in store. We
know that you talked to Joan, your favorite store associate. We know that you never shop
in social and always shop through email.” You don’t need personas. You don’t need segmentation
when you have that kind of information about your customer. The key is just to be able
to use it and use it well. Two, is dynamically calculating data in real-time
and figuring out intent. That’s terribly important. Why am I here today? If I came to your website
and looked at the toaster a month ago, that’s fine. But if I’m here looking at blenders
right now, nothing matters more than that. The right now matters most.
I mean, can you imagine store environments where everyone talked to you about the last
visit you had in a physical store? That’d be a horrible experience. But our digital
tools do that today, and we need to stop doing that. Our marketing, everything we do needs
to be at the speed of the customer. It’s terribly terribly important to stay relevant to your
customers today. This is nothing new and a lot of people say
this. Screens and channels has to be equal everywhere. I totally agree, I actually don’t
put this on here because that’s so eye-opening, but what I will say is these two walls that
you’re trying to break down are fine, and they’re good and you should do it, but I would
say there’s a third wall that exists that most people don’t even recognize. That’s this
wall of technology that you’re building. So, within its own channel even. For example,
I may go to a virtual agent and ask a question about, let’s say, Product A. The virtual agent
will do a great job talking about Product A. Then, I close down the virtual agent and
then I start browsing the website. What do you think the odds are that the product’s
recommendation engine is going to pick up on that conversation? It won’t. It simply
won’t. Why? Because there’s two different vendors.
They don’t talk to each other, right? This is still a disconnected experience for the
customer. They’ll even share with you, that is the voice of customer material right there
and you’re not dealing with them in real-time. So bringing these technologies together, this
is that invisible third wall that you really need to address to make sure that if you’re
truly delivering personalization, you’ve got to do it in an omni-channel fashion. I often
argue if you have an omni-channel initiative and a personalization initiative in your organization,
those things really are one and the same when you get right down to it.
I get a question all the time, who’s doing this well? Well, let me just say, first of
all, that nobody’s doing this great. But it’s an evolution, not a revolution. It would be
hypocritical for me to say, “Doing individualization should be a five- to seven-year roadmap for
you guys, if you’re gonna do at the enterprise level,” and then saying somebody’s doing it
well. Of course, they’re not doing it well because there are parts and pieces that are
being done well by certain organizations. But not everywhere.
So when you think about this pyramid that I’m showing you right here where you move
up the chain from no personalization all the way to individualization, remember there are
some companies that are doing this well in email. There’s some people doing it well in
marketing. There’s some people doing it well in stores. There’s some people doing it well
in different places. But their objective as an organization is to get to individualization
everywhere. Some of them, some companies are just really good at collecting data right
now. They’re not actually good at delivering the experience yet, but they’re building the
right base for delivering personalization going into the future in a way that’s more
individualization than segmentation. Can I give you an example? Yeah, I can give
you a ton of examples, but I’m limited in time, so I’ll only gonna show you a few of
them here. One is True & Co. This is one of my favorites
and most clever way of doing personalization. True & Co., obviously, they sell bras here.
What they do during the return process is they create a dialogue with their customer.
You’re gonna return something, so you say, “Well, why didn’t you like it?” “I didn’t
like it because of this and I didn’t like it because of that.”
Now they realize, “Hey, if somebody buys something from us online once and returns it, and then
we get lucky enough that they buy again and they get a bad product again, they’re never
gonna probably buy from us again.” Right? What they did is said, “Hey, when someone
returns something, let’s get the data and then suppress items on the website that are
similar to that item. No matter if they’re higher margin, no matter if they’re bestsellers,
it doesn’t matter. The point is the customer said they don’t like something. We’re listening
and we’re taking action that’s in the customer’s best interest.” I love this example. It’s
great way of delivering individualization through data and understanding.
Neiman Marcus. Again, doing great things through an app. Just through a simple app that they’ve
built on over the years. Again, it’s a journey and evolution using snap, find, shop. I can
learn about what the customers interests are by collecting that information and looking
at the pictures, the things they take pictures of, and finding the items for them in the
meantime by delivering value. I could connect them with the local store associates through
calling, email, text. Even FaceTime, right? It’s all about what the customer wants to
do. So I can talk to my local sources. I can talk to Joan anytime I want and ask her information,
because that’s what I want to do as a consumer. I can share, create a dialogue with the customers.
Again, sometimes companies just don’t create the input devices, but I can tell you things
I like and I don’t like. I can share with you things I’m interested in and not interested
in. This is voice of the customer stuff. Forget dual times and click data. You get me to look
at certain things on your websites. That’s not really good data. It’s soft squishy data
because you’ve gotten me to do certain things. Here, it’s actually my voice, pure and simple,
saying I like or dislike. Then, accessing things that I did in the store.
Again, that’s omni-channel and personalization tied together where I can see those personalized
looks, the things that I did in the store and accessed it through the app.
I like the Dulux example because they don’t even sell direct to consumers. I can paint
my walls in my house with the colors that are Dulux. Now why would Dulux do this? Because
they understand the customer journey. If you’ve ever stood at Home Depot or any of these companies
that sell paint and sit in the paint aisle, you realize there’s a lot of companies that
sell eggshell white. Right? There’s 15 brands. Dulux knows that, right? So they had to get
out in front of that aisle. They had to get before the aisle and deliver individualization
so that that customer can get that unique one experience for them before they make that
buying decision in the aisle. Sephora is a great company that does a great
job of this really across the board almost everywhere. They do it in the store, like
this beauty station. Now I’ve done this experience. I’m an autumn. If you saw a picture of me,
you wouldn’t know that, but I am an autumn. The women had a great time with me, but at
the end of it, they said, “Hey, would you like to have all this information for later?”
I said, “Yes.” They said, “Well, just give us an email address and you’ll become a beauty
insider.” And bam, look at that. They’ve got all my information. All this could have been
totally wasted. It would have been anonymous data. But no. They incentivize me to self-identify.
This is that creating value in exchange for customer data.
But they don’t just stop there. They compound that information with information they collect
through the apps. See there, I am a beauty insider, right in the middle of screen, you
can see right there. So I can do things like book a reservation or try on certain products,
again adding to my customer profile. Even on the website, there’s a specific place where
I can go and talk about the things that I like and don’t like, creating a dialogue with
the customer. What should companies be doing today? Well,
first and foremost, making the move towards individualization is not easy. It’s an enterprise-level
initiative. It’s not a tactical thing, it’s a strategy. So as you’re doing this strategically,
consolidate customer data from both internal database and external partners in a single
customer data repository. This isn’t required, but people who do this are 50% more effective
at delivering personalization. So, yes, it’s not sexy stuff, but if you want to talk ROI
in a business case, personalization is where you lean on that.
Two, identify and roadmap fixes for digital gaps and delivering personalization throughout
the customer journey. Look for places. I mean, you have a customer journey map, ask yourself
a question: where are we not collecting customer data; where the data we’re collecting is not
identified by a certain customer; and finally, where are we not delivering personalization,
and then go to fix those places. As you fix those places, design each digital
solution to both collect and use individual customer data – not just within channels,
but across the enterprise. A lot of companies like Fabletics, who’s collecting that dressing
room data, they don’t necessarily use it in the store, they use it in their email campaigns.
So think about the enterprise, collect data sometimes for the purposes of using it in
other channels. This is the golden rule for avoiding that
creep factor. Be overt in collecting data, covert in explaining delivering personalization.
In other words, if you’re going to ask me my body type, you better tell me right then
and there why you’re asking me for a body type, explain to me the value I’m going to
get, that you’re going to find the right products for me, that we’re going to curate the website
and it’s the right products that are gonna fit me, and then do it. But then when you
do deliver it, you don’t need to share with me that you know my body type every single
time you talk to me. Just give me the great experience. That’s all I’m looking for.
Finally, use personalization to sell pain points before trying to surprise and delight
customers. Why? Because we’re far more influenced by pain than we are by pleasure. Nobody drives
a car, gets out of the car, turns around and says, “Thank you, car. I love you so much.
I really appreciate where you got me today.” No. We don’t we expect things to work, right?
We expect things to be great. But if the car breaks down, we certainly have choice words
for that car. So that’s we’re trying to do. We’re trying
to create those shopping cart moments for customers so that they stay with us, they’re
loyal to us. They stay loyal to our brand and they continue to be customers we could
win, serve and entertain going forward in the future.
Thank you for your time today. I look forward to the Q&A session in a little bit later.
Thank you very much, Brendan. That was amazing. I think you gave us and everybody here watching
the webinar a great deal to think about when it comes to individualization. I just want
to say, personally, I loved the quote that you use that personalization solutions are
just engines and you need high quality gas to power your engine, and that gas is high
quality customer data, and that gas will lead to driving your individualization effort.
I really love that part of it. I thought that was great.
We have a lot more to go here in our webinar. One thing I want to tell the audience right
now is don’t forget that you guys can submit questions for Brendan through the chat pane
on your screen or join the conversation on Twitter using the hashtag #IdentifyWithSignal.
Let’s now move over to Dean to discuss how he has actually implemented an identity solution
at a leading e-commerce company to accelerate its individualization efforts across the digital
ecosystem. Thank you. Hello everyone. Gosh, that’s going
to be really difficult to follow. That was a really, really awesome presentation and
great energy from Brendan. So I’m up next. I think I’ve got about 20 minutes.
I’m Dean. I founded Programmai after leaving ASOS. I was at ASOS for four years; they’re
a fashion retailer that sells clothes online. I’m going to talk through my experience of
using identity of using Signal and how we leverage that as a tool to unlock a bunch
of different things. My strap line here is “the importance of identity in the quest of
personalization.” So, I’m going to follow on from Brendan.
This is what the world look like at ASOS prior to having an identity solution. It’s very
disjointed, it’s inefficient, and we needed to solve it. This is a typical journey where
customers would come to ASOS through different marketing channels. Actually, one of our biggest
pain points was that we have a very young demographic and those customers are on multiple
different devices and they took to that trend very quickly.
So everything that we were doing was cookie based and you can already see the problem
with that. We were seeing more than one customer many times. Here is just one example of how
we were spending a load of money on trying to retarget them with the products that they’re
looking for to bring them back to the site to convert them.
One thing that I always kind of say here is that one of the the biggest flaws of marketing
to this date has been the fact that the only way that we’ve been measuring it is on last
click. Even today, most brands and retailers will give credit to the last click that drove
the channel and, therefore, you create this ecosystem where everyone is just optimizing
to be that last click and you end up with kind of a really weighted end of funnel.
Part of my role at ASOS was to try and come in and solve this challenge. The challenge
of multiple devices and multiple cookie IDs, and equally, we had a bunch of really awesome
data. Coming off the back of what Brendan said, we had that gas. We knew how people
behaved on-site, how they dressed, how often they transacted, and so it was kind of up
to us to solve that. And we did solve it and there’s a bunch of
case studies that you can check out, they’re in the public domain, about how we use our
data with machine learning to produce predictions that we then activated across both AdTech
and MarTech channels to really go after some very big business challenges and customer
challenges. It’s not about that last click or any click, it’s about how do we incrementally
drive value for the business that we otherwise wouldn’t. All of that was underpinned by the
identity asset, which Signal provided. Step one in that journey was creating identity.
This is a typical funnel. From the left to the right, you have your AdTech and MarTech
vendors. We were doing our own trading in-house that we were buying programmatic ads. It wasn’t
good enough to be buying ads against the cookie ID. We needed to target customers when we
knew those customers were. For the reason of being able to control how much we spend
on them, what the creative message should be, and how we measure success.
To the bottom there, you have a retargeting vendor. They were just focusing on the lower-end
funnel. They were showing answer to customers who have been on flight recently and they
were playing in that game. There wasn’t much we could do there, apart from suppress.
Then on the right-hand side, we traded in-house again across social platforms and again we
were able to use identity to upload email and telephone numbers and speak to customers
that we knew. But the point of this is that what we wanted
to create was an identity graph that ASOS owned. We didn’t want to go into the DMP side
of things where we were leveraging a bunch of third-party data. What was really important
to us is that we fixed our first-party data problem first.
The ID graph needed to stand the benefit, ASOS and ASOS only. What I mean by that is,
we weren’t too concerned with trying to go after reach with no kind of thought to our
own pool of customers that we might perhaps be offering up to other people trying to do
the same thing. The ID graph was owned by us. The way that
that works is that whenever a customer comes to the site and logs in or transacts or opens
our app or opens an email, we would expose the customer ID, and Signal would track all
of the cookie IDs and device IDs that were connected to that customer. The minute that
they do that, we can obviously see that customer, where they are in the funnel, and we can choose
to do branding in mid-funnel stuff or lower-end stuff. So that’s the first pillar.
Once you have, so you’ve solved that known and unknown problem, well, the next thing
you want to do is onboard all of that rich data that you have about them. The way that
we did that was we started to on board session data. This is how customers behave when they
interact with the website, both on desktop mobile and app.
What are they browsing? How often do they come to the site? Whatever different nuances
in the way that they browse. Some customers will come and check out your new in section,
other customers go straight to the category that they’re looking for, others are led by
offers and deals – and we’ll get back to that in a second.
Then, sales and CRM data. How many times do they buy? What do they like to buy? How many
times do they open their email? All of that kind of CRM stuff that you have. Really when
you kind of combine those two data sets with what we were doing at ASOS, which was using
data science and using machine learning, we started to look at the behavior as an event
that correlates to certain outcomes that we’re trying to optimize towards.
So when I talk about no longer caring about last click before conversion, what we really
care about is marketing and advertising doing things like driving customer lifetime value
up so that the next time this person thinks about clothes they think ASOS, not just a
dress, and purchase propensity. What are the behaviors and events that correlate to purchase
and is this customer showing that on this right now? Have they just hit a page and left?
Or if they hit a page, you had a couple of dresses but there’s one dress where they’ve
gone on all the products, they checked the product sizes. So, here’s the building up
intent and journey churn, everyone understands that, based on how often you come to the site,
how often you open emails, how often you transact. If we’re seeing a decline in that, then we
need to do something about that. That’s the way that we approached it.
Once we have that capability, we took it to CFO and we said, “Give us a really big business
challenge to go after with this.” Now that we can see our customers and we have data
about them, then we can predict whether they’re likely to buy, not buy, or what their future
lifetime value is? What is one big problem that we can solve? The answer that came back
was reducing one-timers. We actually coined this as hit-and-run in the UK, but some people
call it one-and-done. But it’s basically when someone buys once and never again.
Marketing and advertising has done such a great job to acquire the customer, but the
experience hasn’t lived up to scratch. They bought once and they’ve not bought again.
Now some of that stuff is okay. People buy at Christmas time their gift, but some of
those experiences, we wanted to really try and get under the skin of why those people
aren’t buying. The first thing we have to do is understand
the problem. This is a graph that illustrates the total number of second orders over 12
months, so this is a rolling twelve months. The blue bar that fills up is all the cumulative
second orders that happened over the year. The key thing to point out is that the red
line is when those second orders have taken place.
You can see that 25% of the entire 12-month, the entire year, second orders have happened
within seven days of the first, and then by about six weeks out, you’ve got half of those
orders already. So at this point here, at 75% or anywhere to the right-hand side, there’s
diminishing returns there. Anything that you try and do is unlikely to succeed.
The way that we tackled this was we said, “We have an email welcome program when someone
joins and buys their first item, we want it organically at that kick in and finish. Then,
from 14 days onwards, from there, we’re going to target customers to try and influence that
second purchase.” So, that looked like this. These are the types of creatives.
It’s kind of one thing I should point out, actually. A lot of what I’ve been talking
about is about sophistication in data and technology. Actually, it’s quite a common
mistake that people in the data game and programmatic channel, they kind of forget about the actual
deliverable. So we came up with some decent creators and all we were doing here is we
were leaning on the fact that we knew the customer ID. We knew their gender, we knew
what language they spoke, we knew how long it was since their first purchase. On the
right-hand side, you can see that the messaging gets a bit more desperate as you get further
out from that first order. We were populating the ads with product recommendations that
our data science team were giving us. That’s something that Brendan mentioned. You
don’t just want to keep showing someone a dress. Actually, in fashion, it’s really difficult
to get that right. You have to look at the correlation between users and customers and
their tastes and trends and, equally, the correlation between products and what products
go well together. Then at some point you’ll get a good recommendation there.
Then, the other thing to point out is that, because we had identity, because we onboarded
our data, because we were doing quite sophisticated machine learning, we started to bid in a very
strategic way because we could look at future lifetime value of the first purchase and risk
of churn, we were putting ads in front of people that we thought needed that ad in order
to stimulate that second purchase journey, and the results were fantastic.
The reason why we were bidding smartly is because we didn’t just want to waste a load
of ad spend on seeing if this work. We had a limited test budget to try and prove this
case. What I can tell you is that, on month one, we only spent a very modest 6,000 pounds
in programmatic ads and yet we drove 2,000 incremental second orders.
What I want to point out here is that the control methodology we used is that we had
a 20% hold up group. Once we’ve matched all of our customers when were buying ads on a
customer ID, if you were in the control, we just didn’t bid for you, we put you in the
control. When you’re in the expose, we bid, we win the ads, and we show you one of those
ads you’ve just seen. Then when the finance team sit down with us, we look at the incremental
number of second orders within that exposed group.
Every single month we were improving the difference between the exposed and the control, so that’s
a good indicator. Cumulatively, when you look out, we were driving 1.3 million of incremental
second orders that we otherwise wouldn’t have taken. The results were were fantastic and
because of that this test, it made my team grow at ASOS. We grew from two to five people.
It became an always-on activity trying to nurture that second purchase, became something
that was always-on across all markets. Then we took that capability, the identity,
the data, the predictive stuff and we went off to other use cases. I can speak to a couple
of those. One of them was that we decided that incrementality was more important than
just conversions. We didn’t want to just keep spending all of our ad spend on that kind
of end of journey, the end of funnel kind of goal. We wanted to understand where we
could put money within the funnel to influence those propensity to go on and drive sales
we wouldn’t have otherwise drove. One of the things that we realized was that
one of our channels was actually taking up a lot of our ad spend unnecessarily because
they were showing ads to customers that were loyal and likely to come back and buy anyway.
All we did there is that we repositioned that budget and spent it, still in a retargeting
fashion, but on customers that were perhaps still sat on the fence, we predicted that
they’d been to flight, they did some things that they weren’t quite sure about what they
wanted to buy. Then on the right-hand side as well, we didn’t
just focus on stimulating second purchases, we also looked at improving frequency. This
is very common in the CRM space, how do you drive frequency up. We said we did that successfully
and we also mitigated churn where some of our lower customers were showing signs that
they might need nurturing before they go on to leave us.
I’ll finish with this. This is kind of what the result was. It was now a world that was
driven with identity whereby ASOS really controlled the audience, the data, the decision-making.
It was pure across device, we could see where the customer was in the journey, how likely
they were to buy. We could target them across multiple devices. We could see things such
as how they downloaded the app, and because of that we might not need to be so aggressive
with our spend. This is kind of like the way that I left it.
Yeah. We had a we had a really good trend. So, that’s me. Please keep in touch. There’s
my email. I’m sticking around for the panel for the live chat, so I’ll speak to you in
a second. Wow. Thank you, Dean, for sharing your successes
at ASOS. They’re very, very impressive and it seems like your efforts to drive individualization
has made you a much better marketer and a much better understanding of your customers
as you drove that. It also sounds like, regardless of whether the retailer operates in an e-commerce
or brick-and-mortar or across digital and physical channels, it’s really necessary to
centralize that rich customer data around a single identifier so that you can maintain
control of this proprietary asset in order to remain relevant and competitive.
For the audiences that are out there listening and watching, these presentations you just
saw from Brendan and Dean, who sparked a number of questions from you. Let’s dig in a little
bit further. But before I really get into the questions, I’ll just remind the audience
one more time that you can submit your own questions through the chat pane on your screen
or join the conversation on Twitter using the hashtag identity with Signal–, or #IdentifyWithSignal,
I said that incorrectly. My first question, I’m gonna go back to Brendan.
From your research, Brendan, that really sounds like individualization is foundational to
how retailers will market to their customers in the 21st century and really kind of like
how they need to do it today. I have a a two-part question for you, really building on your
presentation. One is, where do you think most retailers fall on your personalization to
individualization maturation pyramid? Following onto that, the second question would be, how
should marketers in this industry be thinking or behaving differently to climb on top of
that pyramid and really satisfy their customer expectations for one-on-one engagement?
Well, I would say, as with most of the things that I’ve seen trend, whether it was the customer
session trend or the data-led trend or the individualization trend, usually these industries
are led by about somewhere around 4 to 8% of the market. I don’t think individualization
is any different than that. I think that’s where we’re heading.
Then, following down the pyramid there, probably for an advanced segmentation, you’ve got maybe
20% of companies. These are the ones that aren’t seen as innovative, but they’re seen
as fast followers. I would argue, and I know this is kind of a weird thing for an analyst
to say, but I’m actually a big fan of being a fast follower. I think it’s pretty high-risk
to be so innovative, but now that we have people in the market that are releasing proof
of concepts and showing how it’s done. Again, I showed some examples in my webinar,
but I could have pulled from a hundred of them. We’re not in a state where doing individualization
is now so innovative, but it’s just really operationally hard. I think organizations
today need to figure out how to do that. Sorry, I didn’t catch the second part of the
question there. Yeah. The second question was basically how
should marketers in this industry be thinking or behaving or acting differently to climb
on top of the pyramid and really satisfy? Yeah. It’s kind of an odd question, but as
an analyst who’s been looking at this for a while, I will honestly tell you, you can’t
look at this as just a tactic. It has to be a strategy. When you would say it’s a strategy,
it means the idea and concept has to be bought in at the highest levels of the organization.
You have to take it and try to fundamentally shift your organizational way of thinking
about personalization and saying, “You know what? We went through this webinar today.”
These are broken models. These aren’t just models that are, “Well, we could do individualization
but we’ll just keep doing segmentation.” No. You need to literally show them, “This is
creating poor customer experiences and the health of our customer files is at risk here.”
That’s where I think a lot of organizations they fail. They fail to address the non-technical
challenges of addressing culture and organization and not really getting them on board, and
that’s why they don’t get traction. Not long-term. You can do anything for a week, right? I used
to work at retail. You want to be a different company next year than you are this year?
It takes buy-in from the top. Absolutely. We hear that a lot from a lot
of our clients when they try to drive against these strategies. I want to build upon your
answer, Brendan, and throw it back to Dean. Dean, you were in this situation where you
had to go and drive towards delivering a personalization or an individualization experience. How did
you learn how to do this? How did you learn how to start? Was implementing that identity
a process of trial and error? Tell us about that journey that you took to lead to this
great success in this investment, in individualization. Yeah, journey is the right word. It wasn’t
smooth at the very beginning because we didn’t really know what we were doing. It was a journey.
I think one of the biggest problems was the fact that when you start trying to solve for
identity, it matures, doesn’t it? Because you need customers to come to your site and
authenticate with you to create that kind of link between your customer ID and these
other devices or these other cookie IDs that are out there. So it just does take time.
One tip I can think to give is that the minute that we included Signals identity pixel within
our email, we saw that lift, because ASOS has a good email open array. So people will
open their email on their phone, on their device, on their tablet and that exposes the
customer ID and gets us to that device. So, that’s the time.
The other thing that’s swirling around in my head here is that marketers tend to think
in this very linear way and you see that with most lifecycle programs. You know, we want
to acquire a customer, then we want to push them for the second order, then we want to
give them a recommendation, then we want to get them to download our app or subscribe
to our loyalty program. It seems to be there are these steps that we kind of nudge people
towards. I think one thing that we did quite well at
ASOS was we understood when the right time to do that. The way that we did that is when
you turn to data science for the solution and you ask it to train a model on what has
worked in the past for certain individuals with particular characteristics or attributes.
Once you’ve trained that model, you can start to identify people who are on your site now
behaving in a certain way. You can leverage those types of predictions and pull them back
in to understand when is the right time to push a second order. When is the right time
to get someone to subscribe to a prime delivery offer or download the app?
Because if someone is showing signs that they could go on to be high lifetime value, then
it probably is okay to fast-track some of that lifecycle stuff. You don’t have to wait
two weeks until you’re welcome program is finished before you can communicate with them
with some other messaging. But at high level, this is all test and learn.
I think there isn’t, as Brendan has already said, there isn’t a silver bullet. As long
as the organization has a culture within it whereby they encourage testing, they encourage
iteration, and everything is looked at with kind of a control holdout group to measure,
where this individual strand of activity is pushing people towards the goal that you’re
running towards, then I think that’s a good organization to be with.
Thanks, Dean. Dean, that was brilliant. I want to build on, because I think you gave
a great answer around the science and the iterative process of that journey, one thing
that you mentioned in your presentation that I want to ask about is kind of like really
how marketers can drive a philosophical change within the organization. I think one thing
I heard from you specifically is how you went to your CFO and your other partners to get
buy-in on how you would actually drive the individualization and the value were to drive
for them. I was just wondering, can you give us more of the art side of that, of driving
organizational change? Sure. Yeah. I mean, I was lucky to be within
an organization that was very forward-thinking, innovative by nature, so convincing the people
that were taught to work in this way wasn’t very difficult. I think, speaking to experience,
one thing that worked really well was, four years ago, the way in which ASOS approached
marketing was quite disparate. It was very fragmented. You had brand marketing, you had
trade marketing, you had performance marketing, and they all kind of had different goals.
Performance marketing was about acquisition and driving conversion. brand was about awareness
and how do you measure that. Trade was about what was going to be on the site, what kind
of deals and promotions. When they combined those channels together and really looked
at it from what are we trying to do here? We’re trying to acquire a customer, retain
a customer, increase frequency. promote loyalty. Then those pillars naturally lend themselves
to certain business organizations, certain functions within marketing.
I believe this comes back to when we spoke earlier about identity with known and unknown.
If you know and recognize a customer, then you can pull in information about them and
make a decision about what is it you’re saying to them. When you don’t know them well, then
you’re kind of inferring a lot of stuff. Yeah. I think, just to come back to your question,
I don’t have the magic answer for how you go and convince a CEO or CFO, but I think
if you go to them with a very clear business case around what you’re trying to do and that
business case works like it did for me, then you’re in a good position to go and ask for
that always-on budget and I think they’ll let you work in that way.
Excellent. Excellent point. One thing that both you, Dean, and Brendan brought up today
was really around speed. I want to push this question back to Brendan. When we talked in
regards to the technology investment, one of the things Brendan talked about is customer
data is assessed in real-time and dynamically calculates intent. When Dean talked about
his retargeting example, and he just reiterated it again in the previous question, the window
to convert one-time shoppers diminished greatly after 14 days.
Back to, Brendan, how the marketers test and maintain the freshness of their data to develop
the right individualization or individualized experiences at the right time?
Well, I think you needed to … I see so many marketers turning into button pushers these
days. They really need to be able to be the kind of people that are comfortable getting
into data and saying, “What don’t I know?” Challenge yourself as a marketer and say,
“I don’t want to just A/B test things and push the winner all the time. I want to be
smarter about my customer. I want to learn what data do we have that’s working and try
to understand what data we don’t have that we could use given that.
So if there’s a scenario where you learn the day of the week is a very important thing
about shopping, it’s like, okay, what about hour and maybe it’s channel specific. Maybe
it’s not, right? There are things you could build off of, think of like a Lego block,
right? When you find something that works you start to build Legos around that and say,
“I’m always optimizing,” but it’s iterative. Right?
Again, it’s not just doing it. It’s not just saying, “Well, this works, so this must work.”
No. You must iterate and say, “This worked. Test if this worked. Did it work? If yes,
move forward. If no, go back and say, “Well, maybe something else will work on top of that.
To me, it’s about being smart about your approach to the marketer. Don’t think of yourself as
just a coupon pusher or an A/B tester. Think of yourself as a person whose primary job
is to say, “Do I know my customer? Do I understand my customer? and then take that and say, “Okay.
What data don’t I have to really get me to that where I do know the customer,” not on
a persona level, not at a segment level, but as a profile.
You know what? This is something that anybody can look in the mirror and say, “I’m a consumer,
too. What would I like? What are the things that are important to me?” Then, start to
ask yourself, “Do we have that information about our customers?” To me, that’s really
critical from a strategic standpoint, is to make sure that you’re always taking a proactive
approach to your marketing. Excellent. Excellent advice, Brendan. It looks
like we’re getting up to the top of the hour, so I definitely want to thank our presenters
today, your speakers today, Brendan Witcher from Forrester and Dean Murr from Programmai.
Thank you guys very much for some very, very insightful content and answers to the questions
for the audience. Thank you very much for joining us today.
To the audience, thank you guys for being here and listening to us for the last hour
on individualization. We hope this webinar has really motivated you to think about the
effectiveness of your brand’s customer experience and how you can leverage your own first-party
data to individualize a customer journey at every touchpoint. Of course, like we said
before, look for the webinar recording to be sent out to you guys in the next 24 hours.
Thank you again for attending today’s webinar. That will conclude our presentation today.