Episode Transcript — Facebook Messenger & Beyond for eCommerce w/ Caroline Klatt from Headliner Labs
00:01 Announcer: You're listening to The Spend $10K a Day podcast. Brought to you by the performance marketing experts at MuteSix. This is your source for cutting edge insight into the world of online advertising. From the team with more Facebook case studies than any other agency on the planet. Here are your hosts, Steve Weiss and Stewart Anderson.
Steve Weiss: Hello everyone, welcome back to another awesome Spend 10K A Day Podcast. Hope everyone is staying warm if you're on the east coast and I hope if you’re on the west coast you're feeling amazing because the weather outside is beautiful today, again. Sorry, sorry to anyone that's not on the west coast but it's beautiful every single day in L.A., coming from an east coaster. Today we have an awesome Podcast, we have Caroline Klatt from Headliner Labs. Headliner Labs is a very interesting company, they are in the Facebook messenger space when it comes to direct response and I'm excited to pick Caroline's brain on some tactical execution stuff that e-commerce business can do to really drive performance, specifically on Facebook messenger. Caroline, thanks for coming on. Tell us a little bit more about yourself and Headliner Labs.
Caroline Klatt: Yeah, my pleasure. Thank you for rubbing in how nice and freezing we have it here in New York.
Steve: Yeah, always.
Caroline: At least we have chic outerwear that we can wear.
Steve: That's true, we do have that over L.A.
Caroline: But you never get to wear it.
Caroline: I am the founder of Headliner Labs as you mentioned and we work in the chat marketing space, Facebook messenger and beyond. And a tiny bit about me for where we come from and how what really shapes our vision of the company and I am from the e-commerce world so my background is traditional retail first, at Steve Madden I did shoe design and marketing. Then Ralph Lauren and then a company called Collective Brands, which owned Payless so, I spent a lot of time in Topeka, Kansas.
Steve: Oh Topeka, that's one of my favorite cities.
Caroline: Fine destinations, yeah.
Steve: Tornado alley.
Caroline: So spent a lot of time there but after that I worked at McKinsey for a few years, focused on retail and consumer package goods and then I was at a big startup called Fab.com, which was one of these unicorn companies based in New York and I was the director of strategy and operations there so again, it was actually my first time doing pure play e-commerce with a big focus on marketing and that's really my background leading into what we built at Headliner, which is a chat marketing automation platform. So, leveraging chat to do marketing automation for e-commerce customers.
Steve: So what headliner is, is you guys are solving the problem in the market specifically around how do you do automated ... just like you do with email drips so like you know, just like we work with the Klaviyos, the Brontos, you guys allow e-commerce brands to actually also do automated drips through messenger, correct?
Caroline: Yeah, through messenger and through other channels as well so, SMS, we've played around with a lot of other platforms, the most prevalent being messenger and SMS but exactly that. The problem that we're solving and it was something I took away from Fab was brands and retailers are spending a huge amount of money, as you know, on an initial Facebook ad, the initial traffic drivers.
Caroline: When you get folks to your site and then how do you really optimize that traffic? Other than getting their email and starting to send emails to them, which is it's own challenge given email open rates right now. What else can we be doing, what other tools can we leverage to enable you to connect with customers who are really anonymous web turn at this point and so, we do outreach through other channels, the channels that are registering the highest engagement so, chat. And mostly, Facebook messenger because pretty much anyone who hits your site is going to have Facebook. And so that means that you can reach back out to them with the drips.
Steve: Do you both chat as well as messenger? Onsite chat as well as Facebook messenger, correct?
Caroline: Yeah we do we onsite chat, Facebook messenger and then any other chat channels so like-
Steve: So, we've done a lot of messenger both with our own internal brands as well as the partners we work with. You know, one of the biggest issues, obviously is chat is very ... automating chat is one of the hardest things and whether it's on live chat on your website or even Facebook messenger because people respond. They don't just click away as they go by, they're actively going back and forth asking you questions, blah, blah, blah so, how does one solve the need of making chat work without having to hire a million customer service agents and being able to really make that a big part of their business?
Caroline: Yeah, I mean you're asking, obviously, a really big question, this could take us days to discuss and it's a topic I love because it was something I thought about a lot at McKinsey and at Fab where we were thinking about introducing live chat even more automation was a consideration. Live chat is important right? Customers are not going into stores anymore so, they're coming onto your website and they kind of want to have that experience that they had in a store where you can look at a salesperson and ask question, right?
Steve: Yep, totally.
Caroline: Makes sense that that would be the expectation and that brand would be striving to replace that experience digitally. But to your point, shoppers ... if I'm online at 4:00 a.m., and I'm shopping at the equivalent of a boutique, right? So a small brand, they don't have round the clock customer support, it is challenging when you don't have live agent. There's a question of bringing an automation so, we've experimented with automation, with natural language processing and with live agents and to make this long story short, what we see for the most part
Steve: Before you go further, you brought up a really good point. NLP, artificial intelligence, these are buzzwords that every e-commerce company has heard of at conferences and "We're using these NLP to do our customer service," and I think it's a really interesting topic because you're trying to emulate a real customer experience without using a human being, you know, less human beings so, let's dive a little deep into that. I know one of our partners I asked recently, I was like, "You know what, why don't you guys do onsite chat?" And his response was you know, "That would bombard, there'd be too many messages, I can't even get enough customer service agents." So, talk a little bit more about how do you use NLP or artificial ... what does that mean? I'm not a big expert in that area.
Caroline: So first of all, great problem to have if you're bombarded. By and large, we work with some really big brands and we work with lots of really call startups and by and large, you have enough manpower to serve your customers, right? That's the whole goal is to sell stuff, to serve customers, to build relationships and dedicating manpower to doing that is actually really important and really valuable in the long run. If you're a really awesome, fast growing startup and you have interested customers who you could convert to multiple times purchasers, like repeat purchasers, it's worth it to invest in having the right agent, spend five minutes live chat with that customer. The returns are ... they're demonstrated that you will have a return in the form of this customer making one purchase then over and over and over again so, that's my initial knee-jerk reaction to that comment. But that's true, the truth is a lot of retailers will be bombarded, right? If you walked into Macy's, you might not find a sales associate on the floor right away because there are a lot of shoppers and not a lot of associates and the same problem persists online and so, enter natural language processing and AI, two buzzwords that we all like use a lot. What they actually mean, for anyone who's listening who's not super familiar with them is two different things. Natural language processing is a computer's ability to process language, right? If the computer can have a conversation with me, it can understand what I'm saying now and respond to it.
Steve: How do you know I'm not a computer, I could be computer. No I'm just joking.
Caroline: You could be, you could be. I was wondering this whole time. An AI is really just machine learning, right? There's artificial intelligence that sorts your inbox, right? You can use artificial intelligence anywhere and that really doesn't necessarily have anything to do with robot conversation so, I don't want to confuse those two terms but natural language processing is a good thing because when a retailer says I want to get NLP to answer customer questions, you're really relying on natural language processing being pretty advanced, which I don't think it is. I don't think anybody's demonstrated that it's actually advanced enough to solve real human inquiry. So good examples are the Microsoft Tay bot, Microsoft Tay launched on Twitter, it was Microsoft's natural language processing bot and it was a massive racist within like what was it? An hour?
Steve: That's funny.
Caroline: It was crazy and the reason is because I really think we'll get there and I am a big believer that in the very near future, all customer support will be handled by bots and will be totally sufficient to do it but right now, if you asked a complicated question that was like, "I wear a size 8 shoe but my foot's really wide so do you think I should get this in the wide size? Or should I get my normal size so that I could take advantage of your expedited shipping?" You will be hard pressed to find a robot that can answer that and understand the different pieces of those questions. So, that's the general-
Steve: So the big picture is we're not there yet from artificial intelligence, NLP, chat bot perspective, you have to have ... the big takeaway, you have to have human beings that could help with messenger, you know with-
Caroline: So I think you need the last line of defense to be human beings. What NLP can do and can do really well is answer the first line of questions, right? So your first 60 to 70% of your questions are going to say, "Where is my package? What's your shipping times?" Asking questions that you can anticipate. You probably as a company have a lot of data, historical data around what people are going to ask and can answer. I think you quickly fall back on real humans.
Steve: So getting back to the main topic of execution. Like this is all great, like I'm sure there's a lot of listeners who are listening to this Podcast saying, "You know what, I'd love to take advantage of messenger I think. I haven't touched Facebook messenger as a retention or as an acquisition play." How do you whether it's leveraging Headliner Labs or another tool, how does someone start actually integrating and executing on Headliner Labs or any chat messenger? Obviously what we do is we get the pre check box in Facebook now, you have to check the box but you can have people opt onto your messenger flows from Facebook and most chat bots, whether it's Headliner or any of the others, they'll all actually be able to integrate with whatever e-commerce story you're using whether it's Shopify or whether it's Magenta. I'm sure you'd agree that you guys integrate with pretty much everything. Well the biggest question that I get when I'm like, "Guys, let's really get messenger a big part of the acquisition strategy," is, "Well I don't want people getting pissed off at me. I don't have the ability to answer these replies," and that's the biggest pain point I feel like, you know, is having that customer that serve as army to really start replying to these but I still think there's a way for brands to really start integrating messenger in both acquisition and retention without having an army of customer service people. Would you agree?
Caroline: Absolutely. So messenger marketing is a baseline at this point. If you don't have it, you're behind the eightball, you're missing customers left and right and you are only getting further and further behind so, listeners who are considering messenger. Stop listening, send me an email, send Steve an email, get yourself setup with some messenger solution fast because you're really missing customers every second that you're not implementing.
Steve: Okay, I guess it's also important to talk about from our product roadmap solution the way Facebook looks at messenger. Facebook, we both know a lot of people on the messenger team at Facebook. I know, obviously, three, four people, I was with them at CES but they think of messenger like we think of email. They think that messenger's going replace email so eventually, you're going to have all your automated drips coming through messenger. Whether it's, you know, I make a purchase and you know, the purchase confirmation, that's all going to come through messenger and I think with that said, it's really important to be able to start thinking of building different types of flows within messenger. Obviously there's messenger rules, which I'm sure we could talk about, which are more or less you have to get a response within 24 hours. Facebook doesn't want you to fully take advantage of their messenger platform but let's talk about number one integration, how do you play within the rules and what are some of the strategies that have worked for the brands you work with?
Caroline: So, first of all, I agree with messenger. I think that messenger will be email of the future. When you think about a kid who's in college right now, they email. Nobody uses email, there's a 14% of industry in retail so, what's the future? It email's not the future, where do you have access to pretty much 100% of your customers and the answer to that is Facebook. Facebook, Instagram, What's App, they're all owned by Facebook. Access to customers in a frictionless way so, that's the baseline, right? Someone comes to the site, why would I trigger abandon cart emails? If abandon cart emails require the friction of getting an email and I know the open rate is really low. The alternative is messenger, where I'll get the same number of customers, the open rate is around 80% so much, much higher open rate and much higher engagement rate because it's optimized for your phone, which is where your customers are spending all day, everyday. So, that's my thinking, that's why I believe so much in messenger and I believe in this coming from an e-commerce background where people really aren't opening email anymore. The strategies for- there's a couple strategies that are universal. The universal strategies are universal to messenger and they're universal to email, right? You want to touch a customer who's likely to convert. So if a customer comes on, takes an action, like they put something in their cart, they're obviously considering that pretty strongly. So, you want to touch those people, right? Abandon cart messages, welcome messages, these are really standard messages that today you should turn on and start doing. Then you start to get interesting. Thing and the interesting things you could do inside messenger come with intelligently mixing the artificial intelligence with messaging and understanding the brand and the brand goals. So, I'll give you two examples. One is we work with an amazing brand called OUAI Haircare, it's Jen Atkin's hairline. She is the Kardashian's hairstylist and so OUAI has this really interesting hair quiz where when you buy beauty products, and Steve, I'm looking at your luscious hair-
Steve: Whenever I just put the windows down and my hair blow in the breeze everyday, it's great.
Caroline: For anybody who can't see Steve, I don't think anyone can.
Steve: Thank God.
Caroline: It's all on top. But OUAI has, like all beauty products, there is a challenge when customers can't be in the store so, they don't have a hair person saying like, "I see your hair, it's fine, it's blonde, it's thin", whatever-
Steve: For those of you who don't know, Caroline has beautiful blonde hair so, you can't see her either but that's the reason why she's talking about that.
Caroline: That's how I came up with blonde, the beautiful part, I wish. But it's beautiful because I took the OUAI hair quiz, right?
Caroline: So who is the virtual stylist? OUAI developed this quiz where you answer seven questions about your hair type and it diagnoses you really, really, really specifically. I'm not allowed to give you the exact mechanics behind it but I'll tell you that there are many thousands of combinations of products that they put together for different customers based on your answers to their hair quiz and so, what we do with messenger is two things. One is re-target customers. I came to the site, I added the volumizing shampoo and then I leave, I should get a retargeting message, right? I'm obviously thinking about it. But if I come to the site and I don't do anything. OUAI can reach out and say, "Hey Caroline, we saw you visited our site. Do you want to chat with our virtual hairstylist?" Right? "Do you want to actually figure out what the right products for you are?" And then in messenger, on my phone, as if I'm texting with a friend, I could engage in this back and forth and get product recommendation. All coming off of the fact that I essentially walked into the store, right? I went on the site and I walked out. I left without converting and so, we have enabled OUAI to have the virtual equivalent of a really powerful salesperson.
Steve: That's really cool and it's all based on the questions that you ask so literally, you're going to bounce off the shopping cart then it's going to say, "Hey, would you like to speak to a virtual assistant to really understand which specific hair products you should have?" That's a really interesting way of reengaging. And how hard is that to build and integrate? Is that a really hard thing to do and how do you measure it to see if it's profitable or not profitable? How do you make sure it's not inhibiting the customer experience?
Caroline: Yeah, great questions. So the complexity of building it is really depends on the team we're working with. From our end ... so Headliner is a tech platform so, we provide the technology piece. The tech is all built and ready, plug and play, turnkey solution. If you had a quiz with ready, then you can get up and running today. Putting together the right questions and coming up with product recommendations is it's pretty brand specific so, you know, you want to have the right questions to get you the right products and we continuously optimize results for sales so, seeing our results are performing, seeing what people are converting on, we do all the ... that's where AI comes in.
Caroline: We've come full circle, here we go.
Steve: Oh no, AI again.
Caroline: AI, AI.
Steve: It's not real intelligence but artificial intelligence.
Caroline: Artificial intelligence. So, AI, we can track what people are clicking, what people are converting on, if people are actually responding to these results. We can continue to serve up these quizzes.
Caroline: But the integration is easy. Technically, it's very easy because we do all the tech heavy-
Steve: Cool. That's a really cool way of leveraging messenger. I'm sure they're getting a lot of conversions, they're really building re-engagement off prior purchase intent so, talk about the second way that people are leveraging messenger.
Caroline: So that's incredible and it actually moves into another way that we're doing it, which is post purchase recommendation. So, with the example of OUAI, if you came on, you bought your hair regimen, so I bought seven products in the volume category, OUAI can trigger notifications when it's time to repurchase each of those, right? So they know that they're shampoo lasts 60 days, 60 days later, I want a one-click link to check out on those products and we're seeing an incredible surge in repeat purchases, which is always a challenge online where customers are pretty much just shopping for a deal. If I get the trigger at the right time and it's a one-click mobile optimized link to checkout, it's essentially becoming what Amazon has with swipe to pay, the one click pay.
Caroline: I get this message, it's exactly when I needed it, "Oh, hey Caroline. Time for you to get your new shampoo." And I'm like, "Oh as it turns out, it is time for me to get my new shampoo. That's exactly right." Swipe and it's going to be sent my way. That's really powerful and we're seeing incredible results.
Steve: So question, I thought like when you opt into messenger, you only have 24 hours to engage. With your system, how do you give someone 60 days and keep messaging them once every 60 days? I thought that was against the terms of policy of Facebook.
Caroline: On Facebook messenger?
Steve: Yeah, I could be wrong. I'm just curious because that's how I always thought it was setup.
Caroline: So there are limitations in terms of messenger. They have what's called the 24 + 1 rule where you can message customers within 24 hours and then you get one additional after that 24 hours. There are set rules around how you can do other messages. If you talk to the Facebook team or you read all about them, there are different parameters, I guess we could call them, that enable you to send these additional messages when they are truly value add to the customer.
Steve: So it's a little bit of a bureaucratic red line of when we can message and when we can't. We've actually tried to really play around with that in our messenger flows. We actually try and get a response. Like we'll try and say, "Emoji, how was your day? Take 15% off, like if you agree," or something and we'll get them to respond and then that actually triggers another 24 hours.
Steve: So it's like a little bit of a game, the more you get them to respond to you and make it actually a real messenger experience, the more Facebook gives you but obviously the real issue is if you don't have a customer service team you know, you're sending out all these targeted messages without actually having anyone to really respond after two or three messages because everyone, you know, they're human beings and they respond like human beings and there's got to be a human being to respond. So, I think that like every company ... and the reason that messenger's complex, I'm sure agree with this Caroline is because there's not a one size fits all integration with your business. Your core business is going to be different, obviously OUAI Hair is going to be different than say another brand. Say if you're selling t-shirts. The integration is always going to be different and I think there's a lot of ... correct me if I'm wrong but there's a lot of strategy that goes into messenger unlike email marketing where you could build flows that are kind of generic in a lot of ways, correct?
Caroline: I think there's overlap. So I think that that's correct and incorrect. There's definitely those initial flows-
Steve: Whoa, whoa, correct and incorrect?
Caroline: I agree and disagree. I mainly disagree but I'm trying to be nice.
Steve: I love it when I get people to say correct and incorrect at the same time. You're right and wrong, you're a right or wrong type of guy, Steve.
Caroline: You ever watch Ali G?
Steve: Yeah, yeah.
Caroline: And that's always a thing that one of his characters does, Bruno.
Caroline: Where you get easy to get fashion people to say two completely contradictory statements at once.
Steve: I think that's right but I'm wrong at the same time. I think you're tall and short at the same time.
Caroline: You're wrong and you're right. So you're wrong and right.
Steve: Okay. So tell me how I'm wrong and right? Because this is interesting.
Caroline: So you're wrong in that I think you can automate a lot of this stuff. We have a plug and play platform with certain campaigns preset that work for everybody. And so, if you're a t-shirt company, or you're OUAI Haircare, or you're a sneaker company or you're selling tickets to a show, you want to be doing these campaigns regardless. It takes you no time at all to get setup and I quite literally mean that you can be setup within 30 minutes and sending out messages to your whole site. Where you become correct or where you can get personal is with the example I gave of the OUAI Haircare quiz, right? So, if you want to take it to the next level and make this a real virtual assistant, which I believe in very strongly and we see the results in direct sales. That's where you need to train somebody, quote on quote, right? If you were hiring a sales associate for your store, they'd have to get to know your inventory, they'd have to know where things are, they'd have to be able to answer customers so that much you have to train. On day one though, you can be set up and running with automated campaigns like the email drips we talked about and those are generic. Those work across brands, categories, products, sites, et cetera.
Steve: Alright cool, interesting. So let me take the topic to another place. We have a lot of both very experienced e-commerce marketers who listen to this Podcast. We also get people who are just starting out in e-commerce and they value every ad impression they serve on Facebook, every impression is meaningful and to really maximize the value out of every ad impression, obviously, you got to convert each user that comes into your funnel so, the more people we convert with messenger, obviously the more the better the campaigns look on paper so, if you're going to give advice to a new ... you know, not eve just a new e-commerce brand but a brand that's really just getting their feet wet in Facebook marketing, what advice would you give them on how to integrate messenger because messenger obviously it's not for ... not everyone's an enterprise level brand but if I'm a three or four person e-commerce brand and I'm selling my shirts, how are you going to integrate messenger? What is the first thing you do whether it's going through Headliner Labs, like give me the start to finish quick and dirty strategy.
Caroline: Yeah, I mean, first of all you go to Headliner Labs because what else would you do?
Caroline: First you start with MuteSix. You get setup with all of your ads and all of your platforms, then you go to Headliner Labs and you get setup with retargeting and retargeting should start with a welcome sequence for people who just hit your site, people who abandon cart, people who checkout, get all these sequences set up, like I said, it takes under 30 minutes to get the whole platform on your site and setup and start messaging everyone who hits your site because if you're spending a single dollar on Facebook and you're not reaching back out to the customers you're driving that way, that is a miss. You are losing, losing these customers.
Steve: That's great so start ... number one, sign up with Headliner Labs, setup your welcome series, cart abandonment series, all the different series and then from there, just get some data back and then start really evaluating which specific messages are driving revenue in. I guess the thing I always have a question about from an analytics perspective is how do I analyze messenger ad revenue or measure messenger ad revenue versus you know, Facebook ad revenue. I want to really understand which specific sequence is driving the bottom line, which ... or how do I continuously build a process to edit my sequences similar to email? How do I know which open ... obviously open rates are different on messenger.
Steve: But how do I know which messages people are engaging to and really evaluate it and try and continuously improve?
Caroline: Yeah, good question. I mean, I can say so what we do, which is we provide you a dashboard where you see each message and every campaign broken down so really, really, granular messaging and you see the open rates, the conversion rates, exactly the stats you need to evaluate what and how everything is performing and then the revenue information that we pull is provided by Facebook so through Facebook we can track everybody who gets a message, reads the message and whether or not they go on to make a purchase. So, much like Facebook, there is obviously some challenge around this data, right? Because I might say, "I served them an ad on Facebook, I served them an ad on Google, I served them an ad through messenger and then all of this happened on Monday and on Tuesday they made a purchase and where did that come from?" Right?
Caroline: So, that's a challenge that marketers deal with more than I deal with but I hear that a lot from our customers, which is how do we trust this data? We generally and notoriously are skeptical about Facebook data and I can't solve that problem for you.
Steve: Oh no, the one thing Caroline can't solve. Everything else-
Caroline: Everything else I can solve for you.
Caroline: I can tell you the data that we see, I can assure you that any data we're taking credit for, the purchaser had seen a message from you. I can't guarantee they didn't see other things as well but I can guarantee they did see a message and that part of the note.
Steve: Is there any last click? I haven't dug deep into messenger analytics but is there a true a last click that could be measure from a messenger ad? Can I click and actually make a purchase and really see real attribution on a last click from messenger?
Caroline: Yeah, you can see last click but it's unfair to disregard everybody else.
Steve: Spoken like a true Facebook marketer. It's the wide photo, it's just a bigger picture, it's the lift-
Caroline: It is, it is. I mean think about it, you might get a message on your phone but then you want to go to your computer because your chrome auto fills your credit card information.
Steve: Yeah, totally. I mean, I talk about it all the time. The great snowball effect of Facebook marketing. It's the full funnel, well they saw the ad on Facebook and that just disrupted their day, they need to make that purchase right away and then they went on Google.
Caroline: You're walking down the street and you get a push notification and that's it, game over, you're gonna head back to your desk and make that purchase.
Steve: Yeah, it's really interesting. Alright cool, just to wrap things up Caroline, before you head back into the cold and I head back into paradise-
Caroline: Alright we get it.
Steve: What would you say besides rephrasing, New York is the greatest city in the world, I'm from the New York City area so I just want to put that out there. What do you want to share with the audience wrapping things up? Any follow up thoughts? Anything that you want to that we didn't discuss that you want to go over, take the platform away.
Caroline: Oh wow. I mean, I'm excited about all the stuff that's happening on messenger, I think that we're in a really interesting time in e-commerce where we're seeing some, first of all, really awesome brands emerge really fast. For many years, people have been talking about e-commerce, e-commerce, big brands are going out of business, blah, blah, blah and I actually think that's true now.
Steve: Well Toys-R-Us just closed, I mean, for those of you that have been watching the news, you know it's the brand that we all grew up with as kids, you know, it's cool to be a Toys-R-Us kid. No longer cool to be a Toys-R-Us kid.
Caroline: It's sad, it really is sad. I come from the world of traditional retail.
Caroline: I mean, I don't know who's listening to this but at least when I was in school and taking marketing classes, it was still a time where luxury brands were opposed to the internet, right?
Caroline: And it wasn't that long ago. And really, really, rapidly we've seen this change where if you're not digital, you're dead. And if you're not thinking of the next app and understand customers and relating to each individual customer in their own way, you're losing, right? It's not the power of the brand anymore where customers are going to bow down to a brand and expect ... you know like, follow along and go out of their way to get to brands. I think we're seeing a really monumental shift where I as a customer have become so selfish that I expect brands to speak to me one on one, personally, at the time that I need it. And so, it's exciting to see messenger open up the flood gates where we can now take advantage of that because we have to meet customer expectations, we have to meet customer demands. Messenger is the best channel I've seen for doing that and I'm really excited to continue working with the platform because I think we're just skimming the iceberg now.
Steve: Yeah, I'm super excited too. I don't think this will be the first time you're on the Podcast, Caroline. I think you share so much great intel with the audience and I'm very appreciative of that and you know, I want to stay in touch. You're one of the few people that I met that really gets it. I feel like from both bigger picture e-commerce, it's not just about messenger, it's not just about ads, it's about providing the best customer experience. We're carving new customer experiences for consumers right now. We're kind of like the first people in the door of just building new customer experiences, ground floor so, I really appreciate your time and sharing all your great knowledge.
Caroline: Well the empty flattery is always appreciated but I'm always happy to talk about this stuff as you know.
Steve: Yep, cool. Well that's pretty much it for this Podcast. We are going to be on the messenger talk probably in the near future, we're going to dive a little bit more deeper into some analytics that we're getting from messenger from some of our campaigns but stay tuned for the next Spend 10K A Day Podcast, we have something amazing in store. Not as amazing as this, as having Caroline on our Podcast but almost as amazing. Thanks everyone for your time today and over and out.