Search the website

Albina Kehoe, Head of CRM at Stitch Fix, speaks about the evolution of marketing, at Optimove Connect 2017

Video Transcript

Hello, everyone. So as you just heard, Stitch Fix is a very innovative company. We’re actually a data science company that just happens to be doing fashion. We just started in retail industry and so far done it very successfully. Many of you might know, we just recently decided to file initial set of papers to go IPO.

So, huge milestone, there’s a lot more public information about us now out there. Prior to that, we were pretty secretive. At Stitch Fix, in every meeting, when we start a meeting, if everybody is bothered by something or something is weighing on you heavy, we do a minute of clearing just so we can go on in a more constructive way. So my clearing for right now is that I was very humbled by Blake presentation, not only by his ability to deliver it, but by his deck.

My deck is not quite as polished and I hope that next time I’m presenting in this similar forum, I can do the same. With that being said, I’m going to move forward. Hopefully. Yes. Okay. So I will say that marketing and technology always been hand-in-hand. My training is actually I started as an engineer.

Prior to being an engineer, I studied statistics. Then, I went into information system, and now, many years later, I’m in marketing. I will say that marketing and communication have always been hand-in-hand. From the time of the prehistoric man, when communication was based on stones and rocks and things of that nature to tweeting, advancement in technology advanced way people communicate.

You can argue that Twitter maybe did not really advanced our ability to communicate, but it’s something new and we need to embrace it. So what I will do is walk you so quickly through a few stages of what happened prior to internet, and then maybe we will talk a little bit about internet marketing and where we’re going to go forward. Hopefully, tying really well to your last presentation about where we’re going to be in 2027.

Okay. Got it. Okay. So I will argue that in the last 3,000 years, there really only been 5 major development that advanced marketing and changed the way marketing works. Okay. The first well-acknowledged piece of marketing material could be still found in Museum of Britain and it’s basically inscribed in papers with a very, very outdated technology.

And the content of that marketing was for a landlord to advertise that he lost a slave and the amount of money that he’s willing to pay to bring this person back. About 2,000 years went by before something truly changed.

And what changed is invention of the first printing press. Prior to printing press, very few people were able to read and write. It was actually privilege of very high class. With development of printing press, books developed. With development of books, people end up learning how to read.

And so direct marketing actually took off in 1440. After that, again, about 500 years went by, industrial revolution started and printing technology became significantly more sophisticated. With that, Montgomery Ward printed one page.

That one page changed how we do shopping forever and is foundation of direct marketing as we know it today. Very shortly, his competitor… All of this is a little bit of American piece but hopefully well enough known brands. Sears printed another catalog. Just few years later, Sears catalog was 500 pages.

Montgomery Ward stayed in business for about until 1985 and Sears is still somewhat in business. Both of those companies started as a catalog company, which is the perfect example of initial stages of direct marketing. Then, industrial revolution really took off, and what happened is now you were producing significantly more products that were not really differentiating that much.

So your goal as a marketer became not just to put it out there, but ability to tell people how is it different than the other product that’s seemingly very similar. And that’s where true marketing as we know it today, advertisement in marketing, became a discipline. Companies started to invest in departments of marketing.

Advertisement agencies started to be a thing of nature, and advertisement as we know it today began. Again, a long time went by, but as you can see, with every change, the change became faster, and faster, and faster. If the first development in technology and marketing took about 2,000 years, this one only took 40.

In opening remarks, Amit brought a picture of Don Draper. Don Draper is actually not a real person, a one that you all know, but based on real person and that real person was Bill Bernbach who basically was one of the founders of DDB. What happened in 1949 is, three very creative people started the first advertisement agency that changed the world.

The reason it changed the world is because marketing became less about hard sell and a lot more about emotional connection. And that maybe is a more significant thing that happened in the last century in marketing, and sales, and advertisement. Majority of people are here very young, but you might or might not remember Volkswagen commercials that look exactly like that.

It doesn’t say what it is, it doesn’t say how much it cost, it doesn’t say what it’s going to do for you, it does say how it’s going to make you feel, right? So Volkswagen, it makes your house look bigger. The Avis was not a really important and interesting case. Avis was number two rental company, not number one but number two.

And instead of trying to say that they’re cheaper or you can get them in more places, they said, “We try better for you because we’re number two.” This was extremely innovative and completely changed the way people think about product. One of my favorite quotes, and I’m a huge fan and have a huge crush on Don Draper, is, “Getting your product known isn’t the answer. Getting it to want it is the answer.”

Years went by, not that many years, and the next most innovative thing happened. That thing is why we’re here today. The database technology evolved or actually became a thing. In 1960, the IBM researcher started the first, actually, truly database. It was the first navigational database.

Ten years later, there was a relational database and SQL. And as 10 years went by and there was desktops and spreadsheets. And as the marketers, who uses spreadsheets today as your main marketing analytics tool? Don’t be shy. Majority people still use spreadsheets. Later, other things get developed.

There’s a lot more database technologies, very different and innovative. You can process different type of data with new channels and a new type of information like unstructured data. SQL is not the thing to use anymore, however, the foundation of database is the foundation of modern marketing. So, a few minutes ago, we talked a little bit about the main milestones in marketing from 1,000 B.C. until ’90s.

I would argue that TV and internet, while definitely changed the world, they really didn’t change the world of marketing dramatically. They are just another channel or another multiple channels. It’s not about how you market, it’s just the places where you market.

So, with that, I’m going to walk you a little bit through how it actually developed in real life through my own career. So as I mentioned, I started as an engineer and now I’m doing marketing. I realize the font and the screen doesn’t look very easy to read, so my apologies.

Many years ago, I started my first CRM project at Gap. It’s a big retailer in U.S. I believe they’re all over Europe as well. At that point, Gap had 20 million customer records, 4 brands. It took us three years to implement CRM, and it took about 36 hours to run a single campaign. So it wasn’t really usable.

It was hours and millions of dollars without really being able to deliver value to a business. A few years later, I joined Sephora, also in U.S. At Sephora, we try to do something very different. We launched first truly omnichannel loyalty program, which is a CRM as well. It took us about two years and it was pretty simplistic.

I will show the example of what we were doing to that. Years later, I went to Walmart which is still largest retailer in the world. They have 70 million customers, this is a little bit outdated but it’s probably not. They basically nail almost everybody in United States and beyond. When we tried to do CRM, we hired an amazingly innovative consultant firm.

They spent months and months and months working in the closet. They came out with 130 segments and they were completely not actionable. The reason they were not actionable is because you cannot develop content marketing for 130 segments without really having a very, very different infrastructure.

So, again, very expensive failed effort. A few years later, I went to Shutterfly. In six months, we developed an amazingly different cluster in segmentation that basically changed the paradigm from a product to a customer. We talked a little earlier today and yesterday, I believe, about what does it mean to be customer-centric company.

One of the fundamental rules of that is how your company is structured. If your company is structured by vertical or by a product line, it’s unlikely that customer-centric paradigm will work for you. So, exactly what happened there. There was a strong belief that there is such thing as a photo customer and a photobook customer.

In reality, that’s not true. Every customer sometimes buys prints, and sometimes buy photobooks, and sometimes buy greeting cards. So even though the data was supporting the hypothesis that customers are not that simple and they don’t align with our very clean organizational structure, changing executive management mind about that was much bigger problem than creating a clustered model.

Years later, I went to eBay. Probably that is not considered innovative anymore. That being said, they are very innovative in data. eBay has the biggest installation of Teradata in the world. They have 180 million users. Think about that. It’s enormous. They can figure out anything about any customer at any moment of time.

And actually, when I started at eBay, their approach was 100% automation and they brought in marketing team, CRM marketing team, to go away from that paradigm. And then, we’ll give you few examples of why that is where I believe we’re going to all end up and why they decided at that point to go away from a full automation into combined approach.

A hundred percent automation with a catalog that’s not even driven by you but driven by a user can create some really interesting experiences like combination of electrical massager and a kid toy in the same email. Algorithms are only as smart as the rules that you put in them and there’s only so many rules you can do.

So, in two years at eBay, moving from 100% automation to more intelligent solution, we doubled our GMB. GMB is the way eBay thinks about revenue because it’s basically commission-based business. And improve our customer retention, improve churn turns, and improve new client onboarding.

So a few examples of how that actually worked in real life. So in 2006, I mentioned CRM at Sephora. The CRM was extremely simplistic. The profile that you see right now is actually not from that time, but it not that different.

There’s basically five questions that we asked customers. Your eye color, your skin tone, your hair color, your birthday, basically. That was as much as that. And the system we used for that was basically omnichannel, it was connect in our POS or our Point of Sale and our e-commerce. And these simple rules like if you have green eyes, we’re going to recommend purple eyeshadow, we doubled the value of e-commerce channel.

Just think about just one simple question and that’s one simple rule. It’s personalization for sure, but it’s not sophisticated. Sephora, right now, invested significantly more into Beauty Insider. It’s one of the most successful loyalty programs in the world. When we started, we thought about maybe 15% of people will join.

It became so big so quickly that we had to figure out what to do next because now, it wasn’t exclusive anymore. In about two years, 85% of Sephora customers were part of Beauty Insider and we had to figure out how do you hold Beauty Insider so people who are more special feel more special. At eBay, we spent about a few years building out infrastructures that truly connected all of these different channel.

It’s a very complex business not only because eBay is driven by eBay users, right? eBay actually does not have a product catalog. The product catalog photography description, product attributes are 100% driven by customer. If you want to sell a car on eBay, you upload picture of your car and you say whatever you want to say about your car.

That car does not exist anywhere else in eBay’s systems, right? So that’s one extreme. The other extreme, there’s lots of new fashion brands who are coming to eBay to try to get a different type of customer, offload inventory that’s maybe not most current anymore, or just small businesses who do not want to build their own e-commerce and marketing capabilities and come into eBay to use them as e-commerce and CRM platform.

So, lots of different businesses, lots of different goals. We spent few years trying to build this system that basically will bring all of this together and it will enable marketing to do this combined approach. While the creative of these emails are not amazingly beautiful because, again, eBay does not own creative component of it, years went into trying to figure out how to send them.

And so I will go, maybe, one by one explain why this was so interesting. So the first email on the, I guess your left, browsed but did not buy. So we spent probably about six months trying to figure out how long does it take for a customer who looked at a product to buy it, and what we figured out through series of tests that if you did not buy the product in seven days, you’re not going to come back and buy it.

So at day six, the trigger went on that says, “You looked at this product. It’s still here. And here is all other products that are similar to it.” It drove about 300% conversion from this particular email. So once we figured out that, we thought, “My God, what else?” There is thousands of other opportunities like that.

So a lot of people watch item on eBay, they’re not necessarily ready to buy, they’re not ready to bid, but they add it to their watchlist, right? So, similarly, if somebody added item to their watchlist and did not do anything for 48 hours, the chances are they’re not going to do anything. But if you remind them that it’s still there, and not only it’s still there, there might be other items like that, maybe even different price.

That’s the beauty of marketplace. It’s up to you to do that. All of a sudden, conversion of this email went up dramatically. The next one is probably my favorite one because it’s very aligned with what Optimove is doing, is basically, eBay commission is on average 8.5%. So how do you do 10% coupon that’s ROI-positive?

Logically, it sounds like it’s impossible, right? Because you’re automatically giving away more money than you’re making. So the only way to do it is to drive incremental purchase and to drive incremental purchase not just by pulling forward what people would do anyway, but truly incremental. So, the way we did that is a little bit of how we talked about Optimove testing and self-optimizing campaigns and basically suggestions of what subsegment campaign is working for.

It wasn’t using Optimove, but it was a similar methodology. It’s, let’s send the campaign and then let’s see which subsegment of customer is actually bringing in incremental value based on that, and can they bring enough incremental value to offset the cost of 10% coupon?

And guess what? It’s possible. eBay audience is extremely interesting and there is tons of outliers. We talked about outliers a little bit yesterday as well. You can buy a 99 cents item, you can buy a car or a house. So speaking about normal distribution, it doesn’t exist in an environment like that.

We actually have, at that point, statistical professors from different universities come and try to figure out how to solve problem like that because datasets like that just don’t exist anywhere else, right? There is no other place where you can buy a kid’s toy and a house in the same platform. So basically, what we were able to figure out is that if you cut data in a specific way, we found out that there is some customers who only come to eBay maybe once a year or maybe once in a lifetime, right?

So they’re searching for something very unique, they found it, they pay for it, they never come back until they’re searching for something unique again. And there are customers who come into eBay and buy with eBay every five days. Think about it. Every five days. On average in retail, people shop with a brand 1.5 to 2 times.

So every five days is huge. So basically, what we were able to figure out that that’s the customers who should get coupon and they should get the coupon at five-day period. If they have not shopped at 5 day and we send them a coupon, and we basically continue to hold this holdout group for another 30 days to make sure it’s not a pull forward, we see incremental value and we see that that’s not profitable.

And the last but not least is the new customers, and basically the difficult part there is you don’t know anything about these people yet. How do you know what problems are you having? So through basically conversion analysis, what we figured out is that people who don’t convert in first 30 days… Only 40% people convert, period. People who don’t convert after 30 days will never convert.

So if we start thinking about, like, why? Why isn’t that converting? What’s going on? And we forget there is some very simple problems like they don’t know how to connect PayPal account to eBay and they think it’s required, right? Simple problem to solve, instructional email, educational email. But there are other reasons. They don’t fully understand brands.

They don’t really understand how to bid. They’re a little bit scared and worried. But we don’t know yet at that moment of time unless somebody tried to do specific action and we can act on it, what their problem is, right? So what we developed is a nurture series that basically touched on each one of those points. And the goal there was we don’t know yet what you’re interested in, but we can learn it from one or two emails.

We know that there is four major problems, we can touch on four major problems, and based on what you’re going to click on, your next email is going to be touching more on that specific problem. I love eBay example because as a marketer, it’s very easy to think that the only way you can change customer behavior is through promotions. eBay is not a promotional business.

Most of the businesses I worked at actually were not promotional businesses. And you can definitely increase revenue and profitability of your marketing campaigns without doing a single discount ever. I truly believe that. Which brings me to Stitch Fix. Stitch Fix is a very transformational company as we already said. Our tagline, at least internal tagline, is that we are transforming the way people find what they love.

Notice it doesn’t say fashion, or shoes, or anything else. It’s definitely very wide-open and it’s about we are going to help you to find what you love. We know, in some ways, better than you do what is out there that might be perfect for you. Stitch Fix the box, or what we call “Fix Actually”, it’s not a subscription business.

You can order whenever you want. You can put it on auto-delivery but you can also order. There is no subscription fee, you can order on demand. Five items come in the box, you don’t know what they are, you cannot preview them until they come in the box. When you open the box, it’s actually what a lot of people love about it and there is a ton of videos on Instagram and YouTube about the surprise and delight moment of that.

For a lot of women, specifically, busy moms who don’t necessarily have time to go treat themselves, it’s a moment of fun. People have rituals about opening their box and taking videos, and posting the videos, and soliciting advice of what they should keep and what they should not keep. And that’s really a human element of that. But what’s behind that is that it truly is 100% personalized.

No two fixes are ever the same, ever. It’s all based on data science, so definitely combinations of science and art. Between profiles that you fill out as a customer and our metadata about each product, we find the algorithmic merchandise and find a perfect fit between customer and items and narrow down millions of skews down to 50 to 100 items.

Then, a human stylist who understands how things actually work together and that polka dots and stripes don’t always work, and that if you live in Florida and go on a ski vacation in Colorado, you probably don’t need a summer dress but you can use a puffy and other way round. Narrow it down from that algorithmic selection down to five items.

Each box and each fix comes with a personal note from a stylist. We all style, all of the Stitch Fix employees style, I style as well. Not a lot, I don’t have a lot of time, but I do style a bit. And you definitely spend time describing each item and how it should be worn, and that note goes into the box. Also, in the box or in fix is going a style card, which especially our male client are fanatical about, which basically shows you with these five items how you can build an outfit.

Super-helpful. And the last but not least, but this is what really the problem I’m personally trying to solve, is, with all of this sophistication, our marketing until now have always been one-size-fit-all. We send a lot of emails that go to entire database. We have a decent number of clients.

And a lot of times, they all get the same email. When I started at Stitch Fix a year and a half ago, I was shocked that this is what we’re doing and this is why it became a really big passion project for me to figure out how to bring personalization and sophistication of our business into our marketing. So why is that? Why are we not personalized?

So if you look at this data, it basically tells you a little about who is our average customer is. And this is a very dangerous word. Average is a very dangerous word because if you just glance at it, our customers what you kind of would expect for business like that, right? It’s a busy mom who is a little bit older, has a little bit more money, willing to spend money on herself but doesn’t necessarily have time to go shopping.

A lot of homemakers or stay-at-home mom, lots of education, medical professionals, again, super-busy people. The height and weight is not that interesting. Just an average woman. And you’re like, “Okay, that helps.I can kind of build a prototype of what that person is.” But once you start looking farther, you realize that in that average is anybody from millennial to baby boomers.

We have clients who are in their 80s and we have clients who have barely finished college and starting their first professional job. If your approach in marketing is one-size-fits-all, most likely, there’s going to be only a very small percent of customers who are going to be happy with that. When we started and we only served particular type of customer, that’s basically early adopters, that approach worked.

When I started a year and a half ago, this approach wasn’t working anymore. What we saw that the conversion for marketing was going down on a daily basis. When we start asking questions of, “Why? What’s happening?” We realized that the core client is not core anymore. A lot of different things changed. And we start thinking about, “Okay, how do we approach that problem?”

How do we figure out segmentations that work for Stitch Fix specifically? This is one of my favorite quotes, “Essentially, all models are wrong, but some of them are useful.” George Box was one of the marketing statistician professors and I think it’s very, very true. You can build a lot of segmentation models, but it only works if it’s actually useful to you.

There is no such thing as a right model. So this is our first, basically, attempt to figure out, “Okay, what’s going on?” Why all of the sudden our conversion from marketing specifically is down? And what we learned is what you would imagine, right? People initially came to Stitch Fix for a very specific reason and that’s segmentary easygoing. They’re not against fashion, they’re not fashionistas.

They want to look good, but they don’t need to wear Prada. They’re busy, but not so busy they never shop. They don’t have ton of money but they have a little bit of money. The early adopters of technology says they feel comfortable with something like that and there’s not a lot to say. They’re just average people. Once you saturated that segment, you have to be able to grow out of it if you want to grow as a business.

And this is where you’re starting to see, “Okay, now we attract a lot of different type of clients.” We have people who truly do not…they only get dressed in the morning because it’s socially required to get dressed in the morning. We have people who truly hate shopping like more than just, “You know, I don’t like to go to store.” Truly don’t like shopping. And we have a good amount of people who are truly fashionistas who spend more than $500 a month on clothes.

Their wallet is very distributed. We will never be able to be the only place where they shop, purely because this is not how they think about fashion. But we have a lot to offer to them. So if you look at this, all of the sudden, the problem you’re trying to solve becomes very, very different, right? You’re not trying to solve one problem of convenience and lack of time.

For different segment, you’re trying to solve a different problem. For example, for somebody who is fashionista, the problem you’re trying to solve is we have 85% of Stitch Fix merchandise are unique to Stitch Fix. So what you’re adding to them, it’s something they cannot buy anywhere else. So you will never have their full share of wallet, but you can be in their mix.

For people who are style rejectors or shopping haters, you’re solving a totally different problem. They need a very specific direction and explanation of how to get dressed. They’re not interested in trends, they don’t want to know about ripped jeans or embellishment on new sleeves that are very in right now. The more you talk about that, the more you turn them off because basically, impression of the service they get is, “This is not for me. This is a fashion-forward company. I don’t care about fashion.”

And there’s basically everything else in-between. So what would really help us to understand we need to change the paradigm. We need to speak to our clients the way that they want to be spoken to. And most importantly, we need to understand what is the value proposition of this service to each group of these clients. A lot of models also go only transactional or behavioral.

And so we always had, you know, the kind of a typical segments. New, active, flops, dormant, and things like that. With Optimove help, we went much, much deeper in figuring out what does it actually look for us in a specific business. But more importantly, we’re combining those two things. We’re combining psychographic clustering and transactional and behavioral components of the customer behavior to try to figure out not only who they are, but how to change their behavior.

Why this is super-important is because it’s easy to say, “I want to reduce churn by 1%,” or, “I want to increase new customer conversion by 2%.” But how you going to do that? So if 5% of your customers are fashionistas, you have to focus on fashionistas who are new or you have to focus on style rejectors who are churn to be able to figure out, “This is my goal for this specific microsegment, and this is how I’m going to be talking to them to change their behavior.”

So, once we figured out that this is what we really need to do, we realized we have a huge problem. We don’t have infrastructure to do that. While our company is more engineers in data science and marketing, significantly more engineers than data scientists and marketing, we also have very unique problems that they are trying to solve that’s specific to our business.

Nobody else is doing what we’re doing, so everything that we build in-house specifically serves our needs. Marketing is not a unique problem, so we’re not necessarily solving this problem internally. What we realized with our creative marketing process, what previous panel was about, is extremely manual. We approached each email as a piece of art, which is great, but not sustainable.

You could not bring segmentation and make double or triple, or increase your production by 10 times if you reinventing the wheel every time you’re creating piece of content. All of our data was extremely distributed. We couldn’t really bring client and marketing data together.

And we really do understand multichannel behavior, how to optimize multichannel campaign. Or how to optimize multichannel campaigns. So we went on a journey and basically tried to figure out where to find the CRM that work with our business or built in-house.

We consider both built in-house and looked at pretty much every CRM solution out there and decided to go with Optimove for a couple of reasons. We are data science company so we know that we speak same language. We’re probably mostly going to be using models that are developed internally while combining that with Optimove to some degree. But the philosophy of testing, that one was really core to how Optimove looks at things and how we at Stitch Fix look at everything.

If I’m liking one way is that I don’t have to sell idea of control group to anybody at Stitch Fix. Everybody knows that you have to test, and that even though there’s the opportunity cost for having a control group, it’s worthwhile to have that opportunity cost to learn something from it. So this is example of what it would actually take to go from where we were initially to a much more segmented approach.

So basically, take customer tenure, their last experience, and the fact that they have a fix in progress or don’t have fix in progress, and then trying to introduce any additional variables like the psychographic segmentation or price level. You basically go from 2 to 4 versions to 200 versions really quickly.

In our previous environment, we were not able to do that. So what we did last year is basically…it’s going to sound funny, but we actually did not have templates before. Because, again, we approached this as art. So we went to templates and dynamic content. In combination with Optimove, it enabled us to really produce a lot of different content that’s specifically serving a purpose for the customer subsegment that we’re trying to serve.

Where we’re trying to go with that is basically what I was talking about a little back at eBay example. We don’t want to be 100% automated. We want to combine art and science just like we do in every other part of our business.

We would like to get to a place where there is inspiration and recommendation. So, for example, we know everything customers bought and kept and why they keep it. So if we know that you bought a pair of blue skinny jeans three months ago and you loved them, and now there is a new trend in print blouses, we can tell you, “Hey, do you remember those jeans that you bought? Here’s five new tops that just arrived. Schedule your fix now.”

And that’s basically, again, the combination of art and science that’s core to Stitch Fix and who we are, and that I truly believe, as a marketer, is the way to go and the way of the future. So, where are we trying to go as a company, and probably where the industry is going as well, is transforming the way people perceive us or experience us while we’re transforming the way they find things that they love.

Seth Godin is one of my favorite marketers, and content marketing is the only marketing that’s left. I do truly believe that. The hard sell promotional marketing doesn’t work anymore. When we’re talking about loyalty a little while ago and the statistics that they may share it in terms of 80% of clients are not being loyal anymore is because you can’t compete on price.

And that’s always been this way, but that’s probably more so now. It’s always easier to find something cheaper and that’s going to come to your house quicker. So the only thing you can compete in is in content emotional connection and relevance. So what we’re trying to do at Stitch Fix is to go from promotional marketing to content marketing.

We would never have promotional business, but you can argue that in our content strategy, we were somewhat promotional. It’s always been, “Here’s new stuff. Schedule a fix.” Where we’re trying to go is like, “Here’s new things that are relevant to you and here’s why they are relevant.” The second really big part of it, and that’s where definitely Optimove was helping us, is going from campaign type to contact strategy.

So we have talked a lot about today about big email sends, triggers, and transactional. I would argue that you need to take it a step further and stop thinking about campaign types and start thinking about contact strategy for each customer or for each customer group. What I mean by that is the same campaign could go to multiple people, but it doesn’t have to be in the same time and doesn’t have to be in the same sequence.

Imagine the world where you’re testing not only the content or the type of email you send but when or how frequently. Think about newsletter that may be the same newsletter but it’s basically personalized to each customer group, and each customer group gets a different time in their journey. Because today, the most relevant campaign may be newsletter for this one group, but for the other group, today, most relevant campaign is something else like a birthday.

But tomorrow, that same newsletter might be relevant to them. Frequency is also a very important piece of that is some clients are totally fine getting their emails every other day and some would rather hear from you once a month. So where we’re going with this is not thinking any more about content type or even email type and thinking about what is the journey that each customer segment is going from.

And the most important one for me is moving from the attribution to incremental value. I’m a true believer in control groups on every level. And I think the only way you can truly understand that marketing is actually worthwhile of doing and if your job is actually contributing anything is to know if you’re delivering incremental value and not just moving one customer from one channel to another or from a month from now to today.

So those are probably three goals that we have and probably applicable to a lot of other businesses. What I’m going to leave you with today is three parting thoughts. It’s about where marketing is going. First, embrace, preach, and believe in data-driven marketing.

It’s the only way to go. The outdated one-size-fit-all marketing doesn’t exist. You, as a marketer, have to be a believer for your company to believe. And basically, data cannot only be there only when it’s convenient. It’s going to be inconvenient, and sometimes, you’re going to learn things that you would rather not know, but that shouldn’t stop you from learning.

Embrace technology. Today CMO’s technical budget is bigger than CIO or CTO technical budget. That was actually really amazing write-up in Harvard Business Review in August about that. You should definitely check it out. There’s over a thousand marketing software companies in the world and we all using them, and we all making these decisions all the time.

So, as a marketer, you’re now in techno…is a technologist, 67% of companies saying that they will double their marketing technology budget this year. So marketing data and technology is now one discipline. There is even articles and conversations around a new title of chief marketing technology officer.

Think about that. And last but not least, the only thing that will never change about marketing is the storytelling. No matter what you do, is it data-driven or not, what technology it’s powered by, in the end of the day, people buy not with their heads but with their hearts. It’s an emotional process, and it will always be about the stories that you’re telling them.

The truth isn’t truth until people believe you, and they cannot believe you if they don’t know what you’re saying. And they cannot know what you’re saying if they don’t listen to you. And they will not listen to you if you are not interesting. And you would not be interesting unless you say things in imaginative, original, and freshly. Thank you. So it looks I have about four minutes for questions if you have any questions.