Data-Driven Business: Curating the Customer Experience - Heidi Zak + Jaya Kolhatkar + Stewart Rogers

Now, I'm one of those lazy motivators that likes people to introduce themselves. So why don't you tell everybody who you are but more importantly why is the co-founder and co-ceo thirdlove and I would say that has a really important to our business because it's the thing that really separates us apart from both our competitors in the space and the incumbent Victoria's Secret. Of course, I think the most important thing and I think it is it is the differentiated and the Strategic Advantage. If you can leverage your data in a way that

enables you to have the best customer experience because customers tell you with their actions every single day what they want from you and being able to understand what Is a key to your success like buy into the idea that we have to be tired of driving these days. I think if that's an absolute given but I'm always interested in how people utilize. Data they collect how to use it. What kind of competitive advantage that that gives them. So like maybe we can get everyone some real examples of what you're collecting what

you're using how you using it and what that does for your business idea. Maybe you could take my pills over 16 million women have done our bit find her on our website. And so I think even taking it at a bit higher level I think as a company as a founder and when you think about data, I think in order to collect data, you have to provide value to your customer and I think sometimes people forget that meaning, you know, your average person doesn't want to give you their data unless they got something from it was a great example of

that as well. Third love for us. It was really about how do we help women find a better bra size while we can do that through asking certain questions the collection of data and then we can allow her to easily find something that's going to work better for I'm so the fit finder is the best example of that in our business. We ask about things that are really unique that most people don't ask about like fresh shaved fit issues their bra size and then we leverage that to not only make the size recommendation but to make everything more personalized m in her journey and if she comes back to visit us

what she sees what she experiences of the product is a good fit. But what else does that do to do for you about the journey? So for us it's about where's the customer on their Journey from the very very top of the funnel where she's coming to thirdlove.com for the very first time and knows nothing about us. What do we show her? How do we understand what she needs so things like talking about her fit guarantee free returns and exchanges. Why are browser different that's very different than how we approach the Journey of somebody who's maybe in the

middle of the funnel where she's a repeat a repeat visit her but hasn't purchased and then down to the bottom of the phone also for us. It's really about every step of the way understanding who the customer is where she is in her journey and what types of content we should provider to help her understand third love in the benefits we offer I'm trying to think we collect enormous amounts of data. We collect the 2015 billion event the day of people watching content and context

for different needs. Some of them might be continuing to watch certain shows other people might be trying to discover new things very similar to how are you thinking about it? Where are they in the journey? And how do you make sure that your product is actually matching? Of course. I told her we'd do pretty much everything or what kind of content should we license or make and things like that or to marketing to advertising spending a lot of time over the last year or so is leveraging these

fifteen billion events a day to understand what sort of products do we need to make for our customers. So for example, last year we were doing our customers actually binge watch at least once a month and you're watching two three episodes at a stretch and this goes into them question. Is there a different add product that you could offer these customers so that it Does not actually distract from what they watching and so the team went off and build a bench add product which allows us to a predict who's going to binge and then work with Our Brands to say what kind of advertisement

could we showcase that would allow people to have a better viewing experience. We also have a millions of people pause while they're watching and you know, that's kind of interesting. Why are they passing? I mean are they going to say is that a solution that you could build out? Can you show a static and if you well and I've been absent, you know, I drink sides and things like that, which you don't play to that moment. So we need a lot more now in thinking about the product.

Experience 444 a content and showcasing that the first time and it tells me exactly when pizza and beer experience. We all quiet blinds ads now at Ackles allowed amount of people live in stool. Blockers on that browsers cuz they're just so sick and tired of them. I'm going to take a lot of sleep. Comes from the fact that appetizing has just being the throwing up against some of it sticks model. How close are we to Turning adverts into something that is more of a kid

into a kind of Google assistant. I've experienced where it's the the right thing when the right place on the right device at the right time, but it feels like you're actually helping people as opposed to appetizing to them. I mean how close are you to that? Started this. We had not really structured the wavy set up a data in a way that allows us to understand context relevant in a way that is easy. And so now that we have said that up it makes it easier for us to look at the various contacts that someone's watching

understand some aspects of mood if you will end with my father and what you do and then how does it flip out? If it does huge part of what we do I would say marking is a core competency. Obviously, we leverage Hulu as a customer at third love for us. It's really about leveraging dated understand the right marketing mix. So you think building things like our internal teams felt like our marketing mix model. I'm to really understand if we want to spend one more. Where should we spend it? What does

that look like, but for us for ads in particular, I think when you have sort of that captive audience, like when you know what they're doing, it makes it a lot easier to have a targeted message. And so that's really how we think about it. I mean, obviously we do things like TV linear TV, but again, you don't exactly know who's watching or what message is going to resonate what times do you have to do something that's broader. Where is with Hulu? We can be more targeted or things like podcast where we really understand what what somebody's listening to you can be so much more targeted

with that message. So I mean one of those things that strikes me is that whilst you're both getting an awful lot of stator in a little bit later is extremely private yours in particular of the date of your collecting is extremely personal to your customers out, of course around the world. We have little Souls of different regulations to think about, you know in Europe the coast is gdpr and there are many Ava to front types of regulations similar to that. I'm as you go around. How are you balancing collecting that data using that data, but at the same time note folding fan of

any of the regulations and staying completely mindful of the privacy of your yeah. I mean, it's always privacy first for us everything we're doing with the data is anonymized data sets right that we're leveraging to understand segments so customers so we're never focusing on you know, the one woman who answered the question a certain way, but certainly privacy and I'm putting more of an effort on that internally and making sure our data secure something that our teams are doing and thinking about everyday. I think it's very similar to how Heidi talks about privacy is first Frost in

compliance with the law, of course is where we are at and on Friday as well. So we do not use any data and not algorithms or even that is really protective and makes it easier for us to abide by the law. Juicing customers to personalization because you know again, it's been a challenge for marketers for many many years to get to that one. The one personalization level. Everyone's been talking about it for seemingly decades and I'm pretty close to a lot of that. But at the same time

it is very easy and we've seen lots of really famous examples of where people have fallen foul of that. Something may be too personal somebody much to personalized. It's cool sorts of issues and problems sometimes even national news. How do we again introduce people to buy personalization experience as a case of showing the vanilla looks like and then trying to motels in the last looks like and giving me the opportunity to see the value or how to operate more interesting ways because I start the old way of doing it. Is there anything more

interesting in terms of how we actually get people into accepting personalized? New streaming service is coming out everyday. It seems like and we'll enjoy and so for us position is key and making sure that you are showing them things that they would be interested and I think our customers expect that is how they see that particular new discoveries and they also wanted you to take them to the old and familiar that they want to watch a certain time. So very personalized experience. Our

expectation is that that's what they would expect that they come to the side and we deliver that Now it's become expected us older folks are equally interested in the position. So how much does personalization play a thousand years and how do you introduce people to that? I mean, I think it's from the very beginning because we started a relationship like that's how I look at sort of, you know, Adidas CP company, right? You're the whole benefit of anything that's d2c is we

now have talked and I know something about you and you shared something and so how do we make sure we continue that conversation in a way that makes sense? So from the first email that we sent at send after you do fit finder. We're talkin about you. You know, we're talking about you are what we recommend for you what you told us. We're getting you content that makes sense to me. We do it obviously a ton of testing as I'm sure who does as well and I mean, that's the easiest way to learn new things like putting someone recommended bra size in a subject line has much higher likelihood. She's

an open an email than just some generic, you know email until I think really it's about testing but sometimes things don't work as well. So I think that's also what you learn is that just because you know, what somebody what you think somebody's for for us what their body shape is doesn't mean all the time that they necessarily need to see a model who looks just like them. So there's just Nuance to it that you need to be thoughtful about and and always be testing these I think like most pass if you're running a test 50% of test should fail your most of your test

shouldn't work and so have all your tests are working. That means you're being too safe and what you're trying and you're not trying to do something like different and so I think that's We really learned over the past few years Your Business Leaders and then see what other people are doing really really well and then you do matching re-engineering not see if I apply to all I'm just outside of your organization and state your business in Tustin lights content in helping the customer

experience and journey, but it's really wild you and you so yeah, that's wonderful. We could we could do something with that areas that I think and this is not in the streaming voice and the way that tore through all of the irritating things when you're in the park and try to make them simpler with the magic violin, Orlando. Which allows you to do things that you don't have to take your credit card with you you can just pay for it with your magic bands, you know being able to have your hotel

Key is everything in one area. Of course. It's amazing from a perspective. But I find that it's it's very convenient and I think about you know, how they have talked through the problem the operational aspects of the problem to try and take some of those invitations out. I kind of find that fascinating flow in hot spots in and everything else. I'm any good examples. You say Good Egg has really leverage customer data. I feel like they know me and they know what I need and I think that's again like the best user experience is something

where you feel like the the the service or the products helping you and your daily. Life until I feel like what they serve me the email reminders. They know what I meant when I haven't been around and I may be thinking about turning or I've gone somewhere else because it's been awhile like their marketing is very targeted. And so I think that that's a great relationship with a company is leveling the playing field between startups and larger organizations that have been around a long time and that's partly

because we thought UPS have a Greenfield site. You got to decide what you're going to collect you don't have all those years and years and years of Legacy systems in changing between the LP's and supply chain platforms and so on and so forth and strike fear into my heart. So, you know, you have that situation. Maybe she'll do you have a more mature system and you have your own challenges, but the data has to be right and especially as we move into a world

where were using machine learning and AI to get a leg up on the competition until it's a win the day how important is it to make sure that your data is from the get-go going to ask you a different question. But how important is it to make sure from the get-go it is as good and clean as is humanly possible. And how do you maintain that? I think that's a really really great time 10 years ago was a big public retailer and I went to pull some data to took me three weeks to get data from some team it team and it came back

and then I couldn't even use it to do the analysis I want to do and so, you know when building thirdlove, I'm absolutely it was like how how can we leveraged the data that we're getting and make sure it's accessible but even for us you know it 7 years ago. You know, we're shifting now from redshift to Snowflake and we're redoing kind of the platform that were using right for data storage, but it gives us a really good time to take a step back something like we use Flickr as a lot of businesses do and what we came to find out was and I was telling the story actually before we came out

on stage with I was in a meeting and we were talking about a certain tastin and conversion rate and one person was like conversion went up and one person was like conversion went down and I'm like, okay. Well like, you know, that can't be obviously right so clean data leveraging the same data and making a very accessible to everyone in the business is incredibly important, but it never ends. I mean, you could probably attached to this I think you constantly like anything you're doing it's an evolution and you have to be checking in at any moment in time to make sure that data still make

sense in the way you're thinking about it makes sense. And so we're doing that today. So I think yes, we have a much. Your system anyone at thirdlove can access the data that's relevant to their business unit. So they're not dealing with a three-week like like like like I did 10 years ago, but I'm certainly the evolution of how you think about your data. I don't think it ever end make a start up every single person in the company has access to everybody had apparently their own definition of what a subscriber was

going to meetings in the first 15 minutes for be like who's that has correct step back and said you'd really have to bring inconsistency. I mean Democrat is data the basic philosophy that needs to be accurate and it's going to be accessible. Analysis or come up with business decisions. Everyone should be able to say 18 months in governance data quality making sure that there is training where we talked to folks about. What does the definition of subscriber mean? You know, how can you have different

definitions of the different contacts? How should they all rolled up? All of these things? I think takes a lot of time and effort and older company is I think the harder it gets so one of the things that I will have advice for startups is, you know, spend a little time and just making sure that you're defining data and as it changes make those changes so that you know where you stand at that point. Average data does not equal so you should be spending time on hero governing every single piece of data is important for

you and then just emphasized that data and take care of something else for you to choose, you know, we shouldn't just develop things because we can and we could we technically we can do it. We should develop things that fit the business purpose and because we should do it rather than just because you can do it. So, you know, that's why I think an important message for a referral startups and I remember interviewing jet.com with actually let the very first interview on station and they said to be best very first hire with a data scientist. That was the first I was the most important

thing to him because without that data without that being structured correctly without everything being available in the right way. They were never going to succeed at a cool safe famously went very quickly to us when I condition final question. You got a time machine. I would love to have a time machine that be great. But imagine you go to time machine. You got back to the start of your career and in need of daily driven marketing and then customer experience but you get to take all of your knowledge with you. What's the very first thing you going to do to set

yourself up for success? That's a good question. So if I could go back to the beginning of third love. I think the biggest thing would be. Probably bringing on a date ahead of data in a data engineer earlier than we did. So we did a lot of sort of data capture in the early days, but weren't really leveraging it. I'm going to sort of the chicken and egg right like new new no, really strong data scientist wants to come work at your company when you don't have enough data. So there's

just there's that trade-off point, but I do think if we would have brought in some in a little bit earlier actually some of the challenges that I'm talking about that we have today. We could have media ever ever heard of those. I think if I could go back in time, sometimes when you are in the dataspace you think so lie about the data and you don't really spend time thinking about how the business can leverage what date are you have being able to like really understand what the business purpose is just

a much better day. Burson way back in my career and that comes through time learning the foot. Your building is not date of a data sake but it is for a purpose and figuring out how to connect the two would have been a good skill to have right off the bat the best fitting bra in the world is a Woman's Fed love. What's her first and second loves is about creating a third option for women. So this idea that You were either a b cup or C Corp, we created being a half based on the data. We collected right and so the idea that you cannot be comfortable bra or the pretty Bravo we

created both so creating a third option a third love for women in Champaign.