NYC Personalization Pioneers Summit –

Read the full transcript

So we’re Nuts. We are about 300 people, nine year old company, and we’re trying to do something pretty unique. So in order to understand sorta where we’re goin’, I’d love to just give you a sense of our heritage. We’re a family-owned business, a family run business. So right here is Poppy Sol who founded the business during the Great Depression. And this here is Jeff, our current CEO who brought the company online in around 2000. The most important thing to know about the business we’re trying to build is that we want to do it with humans in mind. So first of all, for the people who work for the company but also for the customers as well. So we’re trying to build everything that we can around them, starting with the products. This has been the growth engine that’s really driven the growth of the company. We’ve scoured the world, and Jeff and others all the way back to Poppy Sol were obsessed with finding the best products they possibly could. So whether it’s removing the dyes from our strawberries, or working with our goji berry manufacturers to get the perfect moisture content, scouring the world for the best dried mango, it’s pretty awesome. We have way more than just nuts. That’s actually a quinoa pasta, frica, and an Office Snacks box as well. We’re expanding quite a lot, speaking to people in many different channels. So I can’t read this slide without sounding a little bit like Dr. Seuss, so apologies. We’re online, on air, in the air and in Union Square. Somehow that all just kinda came together. So Union Square, we now have a long-term lease to every freestanding structure, so we’ll be opening up two sort of small footprint retail locations in the structure that’s on the southeast and the one that’s on the northwest. And there’s sort of a company wide passion project, we also have the lease to The Pavilion as well, which is a 215 seat seasonal restaurant which we’ll be opening up in mid-May. Not branded but we’re still extremely excited about it. We’re on United Airlines, sold snack. Online is really our bread and butter, and we’ve just started opening up QVC as well as a channel. So what do these things have in common? These are all of our customers, so although our boxes are recyclable I do not recommend eating them. But we serve families with specialty ingredients and pantry staples. In case you’re not familiar with the thing in the middle, that’s a sugar glider. We serve a sugar glider breeder, I was pretty psyched when I found that out. And on the right is the best restaurant ranked in the world, Eleven Madison Park, we serve them as well. So if you ever want to know where they get their Japanese bar peanuts, we can sell you those too. Dynamic Yield, thank you to everyone, big Office Snacks customer of ours, so powering the office. But what I hope you’re taking away from this is that we have an extremely heterogeneous customer base, because we’ve been very product-driven in the past. So extraordinarily wide range of product, and because of that we’ve attracted a very diverse set of customers. And until about a year ago we had one experience for every single person who came to us. So everyone got the same email, everyone had the same site experience, and there was no way that we really spoke to people as human beings. So what do we wanna build? What we think we’re building is the new school of CPG. So I love this slide, the top 25 largest manufacturers, and granted this was in 2015 but still think it’s true, aren’t contributing to category growth. So only $1,000,000,000 of the growth in the category is attributed to the largest manufacturers. All of the growth is happening for the small manufacturers, the folks who are really disrupting. And where we think there’s an enormous amount of potential is to really hit big CPG where it hurts. Go after the smaller segments of customers that aren’t served by the products that they have, and aren’t served by the broadcast-level messages that they’re putting out there. So thought this was a really good representation of just how much they are under attack. My favorite was when I read a New York Times article on Yoplait. After the failure of Yoplait Greek and the Greek yogurt wars, they rolled out Oui by Yoplait to be more authentic and deeply connect to their customers. And the article was actually really flattering, and it was saying that they learned how to fake authenticity, and that was going to be the way that they would succeed. And we just don’t see that. What we want to do is actually be authentic, and actually have unique and authentic relationships with our customers. We really, really want to get to know them, we want to speak to them as individuals and we wanna grow the business based on providing value that other companies can’t. So product innovation is really our heritage, this is something that was really important to every person who’s run the business. How do we find the best dried mango, or partner with an African co-op to find the best dried jackfruit the world has ever seen? But now we wanna pair that with really deeply human marketing. So going far past broadcast to the audience level, personalization and eventually predictive, so we’ll get into that in a little bit. So our job as marketers as we’re building out the team is to turn this, how many of you have seen the LUMAscape or the Chief Martek ecosystem? So, crazy, right? How could anyone possibly understand the interactions between 5,000 different vendors, all of whom I’m sure are cold-calling you, into this. So I thought this was really funny. I’ve never heard of these before until I looked for them, they’re called the Stacky Awards. These are for folks who have built the best marketing stack, which I thought was really fun. And the visual metaphors here are really incredible. So you see a rocket ship up on the top right, you see a tree growing, right, a bridge being built towards the future. My favorite, for those who have sharp eyes, and I’m sure you could guess the one that won, it’s this one, because who doesn’t love marketing tech falling from the heavens and delighting children everywhere? So I know this sounds a little nuts, right? Why are we in a position to do this, and why do we think that this is so foundational to a CPG company that we want to build? So over the past year we’ve really been focused on putting this into practice. So gutting our tech stack and setting the foundation for what’s next. I wanted to sort of boil all of this down to something very practical that, even you’re not a CPG company you could take away and think through. So especially with the first one I’m gonna be a little forceful, but I was just hoping that folks would really, really consider this. I think you should fire ESP. I almost don’t care who they are. So unless it’s something that is designed for the next 10 years, they don’t really serve any value. So picture, I don’t know if you know the amount you’re paying them, picture the CPM in your head, picture 90 to 95% off from a CPM basis, that’s what you should be paying. The opportunity, and it’s huge, is to move to the new-school model. So technology platforms where they have to make a lot of decisions, like how enterprise ESPs are built. They’re built to push a lot of data into them to allow marketers to segment, to dynamically target content, change send times. And in reality, not only don’t you need that but it’s actually not preferable. So if you were to ask a bunch of people, they would say oh hey, I’ve got lots of email triggers set up, and that’s really the value that they provide. And on one hand that’s great, and I’m sure those are really high performing, but on the other hand those are locking you in to holding the triggers within your ESP, and preventing you from coordinating those with all of the other opportunities you have to talk to people. So I wanted to lead with this because from a cost perspective what I’m gonna suggest next might seem prohibitive, but you’re gonna save a ton on your execution channels. And it’s also gonna unlock a lotta value as you start to connect the experience between email and all of the other touch points that you have with the customer. So I think you should hire a CDP. This is, I think, a very, very transformative moment in our space, and we’re now able to do the things that were only promised in the past. So what does a CDP do? There’s no agreement anywhere, and every vendor will say they’re a CDP, everyone’s trying to be a CDP, this is the sexy new buzzword so you’re gonna see this a lot. But the really good ones, they’ll perform the functions that are in green. So ETL, not the sexiest thing in the world. This is all the data pipes, having those pre-built into every single execution channel. Getting the data out of whatever is sending email for you and down to the raw data level. So for Matt, did he open, what time did he open, what emails did he respond to? Getting all of that into one place. The thing about ETL though is it’s not always real-time. So you need some real-time data, for example did someone purchase, to be able to stop some of the automated and triggered marketing that you’ll build. So if, you know, you’re serving retargeting ads and you were just relying on ETL and someone already bought, you wouldn’t know to stop serving those. So really important to have a real-time component as well. Data modeling, this is where things get really exciting. So I think in the future a lot more people are gonna own more of their own data science internally. It’s going to become more commoditized and easier for folks to able to own that. Everything from who are your highest value customers to who have you predicted has lapsed, and when should you speak to them? For us, one of the biggest challenges that we’re working on right now is who’s the business? A different team internally deals with those folks. We have a sales team, we have customers that spend hundreds of thousands of dollars a year, so we need to know that that’s a business, what type of business they are so that we can create a really human experience for them downstream. Data warehouse, all this data’s got to sit somewhere. This is a great opportunity to layer on a BI tool. So really excited about the direction Dynamic Yield’s headed in, where they’re pushing some of their data to BI to be able to really understand and report and dive in, same concept here. The guts of this tool are not just collecting the data, but really doing something with it and taking action. So downstream, you’re gonna have all of your execution channels plugged in. You’re gonna have display, email, on-site, and anything else you dream up. For us, a dream would be plug it in to our order management system so we can trigger gifts to people, or add things to peoples’ orders that we really want them to try. Direct mail on demand is a huge opportunity, voice. So for us, if a customer stops buying from us who’s spending hundreds of thousands of dollars, don’t send them an email or a display ad, send them to our sales team to follow up, to stop by their place of business or give them a call. So then it’s not gonna do execution, it’s just going to plug in to best-in-class execution channels and then report on that. So the analytics are really, really, really important. You wanna be able to understand if you’re taking action on a win back flow or even a one-off campaign send, what effect did that have on the customer cohort? So with this, you can do really fun stuff like longitudinal holdout groups. So you have an idea for a win back, you want to show up at someone’s place of business. That’s really expensive, really difficult to pull off. So allocate 5% of the traffic for the show up at their place of business strategy, 50% for an email strategy, 40% for a phone call and then 5% for a holdout group across every downstream touch point so you can really measure the lift and justify it to the team. Alright, cool. And then from there, connect that to best-in-class execution partners. So they have to have great APIs. If you take this path, you will actually save a ton of money because you don’t have to have the partners do as much. I think, sorry, Dynamic Yield is an exception. They have to make a ton of real-time decisions on what to show people on-site, there’s just an enormous amount of complexity associated with your site experience, so that one’s really valuable. But SendGrid as an example, all they specialize in is deploy this message, in this template, to these people right now. So you’ll just pass in the data that they need to be able to trigger that email, and they’ll send it out for you at a extremely low CPM. And in fact they’re so good that they’re who Dynamic Yield uses to trigger all their emails. Lob is a great example for direct mail on demand. So direct mail as a whole is really expensive, there are going to be some people who you can’t reach through display or email. These are great people to try a physical mail piece, something that’s more special or personal, certainly personalized, and get it to them at exactly the right moment. So as a P.S., one huge opportunity, I love to see the McKinsey slide on Scrum and Agile teams, it’s something we’re moving to internally. Continuous improvement, so we’ll get there. But the challenge is that if your marketing team was heavily channel focused, if you had an email marketer you might need to restructure the team so that folks are more, they’re thinking about strategy, and planning, and automation, not on the channel level, they’re thinking about it globally. And for a lotta people this is actually a huge opportunity. They can take those same skills that they built building automation flows in email and unlock the potential by building in other channels. So just to make this really real, the current state of our win back is, has anyone ever read this quote before? If you love someone, set them free? That was sort of our approach to win back, right? If you left us, great, if it was meant to be you’ll come back. We would never actually reach out to you with anything. So maybe good dating advice, but not great for our business. So I’m not gonna go through every single one of these, but just to show you some of the ways that we can iterate into a win back flow when we’re using the CDP is start really simple. You know, we have a basic definition of what lapsed mean, right? Hasn’t bought in 90 days. We aren’t connected to our downstream channels yet, so you know what, we can segment a list based on everything we know or might know, and actually pass that to our inside sales team to be able to call on those customers and measure the effect of that. And for everyone else, we’ll put them into a drip campaign through a SendGrid implementation. As we get smarter, we can predict who’s gonna be B2B and B2C and fork the experiences that they get. We can integrate our CDP into Hubspot so that we don’t have to manually pull the list every time, it’s just populating the agent’s queue, the rep’s queue, every day with the freshest people who have lapsed, so we can reach out to them in real time. We can sync audiences to Google and Facebook. So I think this is huge and really easy to do. It’s wonderful to test speaking to people who aren’t responsive to other channels or maybe just prefer if you speak to them in a different context. So trying targeted creative there. And version two, and this goes on to infinity, right, we can predict customer lapse. So if someone buys every week versus someone buys every year, one’s a gifter, maybe one’s filling their pantry, we can understand who’s who so we can put them into the appropriate win back flow. Cool, so for us we see this as a, maybe this isn’t as elegant as Amazon but it’s something that we feel pretty strongly about. It’s as we get closer to human marketing, right, as we identify the audiences, and the people, and what they really care about, we can use that to drive real product innovation, use that to create content around that product and then use that to drive better marketing. So as soon as we know that a huge audience for us is, you know, top-tier restaurants or sugar glider breeders, or offices who want delicious Office Snacks, we can then understand what those folks are buying, what they really care about, talk to them as human beings, collect all their feedback and work with the team to bring on new and better product for them. So I will cashew later at drinks, sorry I had to say it out loud. But try to leave a little time for questions, or we could just break and grab some drinks and sausage.

Matt Butlein, Chief Vision Officer at, discusses how he structured his product, team and technology stack to deliver the most engaging experience to the customer online or in the middle of Union Square.