The session, Commerce in the Face of Amazon, at Techonomy 2018 in November 2018.


The following transcript has been lightly edited and condensed for ease of reading. 

Speakers: Michael Krakaris, Deliverr | |Ben Parr, Octane AI

Moderator: Josh Kampel, Techonomy

(Transcription by RA Fisher Ink)

Kampel: Homestretch. It’s 4:30 now, Monday and how could we go any further without having a conversation about Amazon. It’s been touched on throughout the couple days, but this is really going to focus on Amazon. Here with me today, I have Ben Parr, cofounder of Octane AI and Michael Krakaris, the cofounder of Deliverr. Session’s about what we’re calling “Commerce in the Age of Amazon” and both these guys are people—you know I’ve known Ben for a while, but really that I started talking to over the past couple months about their companies and how they’re applying AI and technology to help merchants, especially in this weird time where companies like Amazon has amassed so much power. So Ben, tell us a little about Octane AI so that everyone here gets a sense of what you do.

Parr: Hello, everybody. I’m Ben. I’m the cofounder of Octane AI and so what we do is Chatbot for e-commerce, so merchant stores, they communicate with their customers in really sophisticated ways over email and over these other kinds of platforms, but many of their customers have moved towards messaging and so we have built a platform to help merchants both big and small communicate directly with their customers for everything from like helping them to pick out a product, reminding them when they forgot to complete a purchase, to help using actual artificial intelligence, to help understand like what kinds of questions customers are asking and create automated responses back to things like “Where’s my order?” or “What’s your return policy?” until a complete solution for the entire customer journey. So we work with everyone from like L’Oréal and GoPro to over a 1000 merchants on the Shopify platform.

Kampel: And Michael, tell us a little about Deliverr.

Krakaris: Yeah, so Deliverr is essentially an Amazon-like fulfilment service for e-commerce merchants. And so we handle the warehousing, the shipping for a merchant whenever an order comes in and essentially what we enable merchants to do is offer free two-day shipping anywhere they sell. So if they sell on Walmart, they sell on eBay, they sell on their own Shopify store, letting them offer free two-day shipping, so that it’s not just, you know, on Amazon where you can get their products for free too sort of thing. We work with some of the largest Amazon sellers today. There’s actually a good chance a lot of people in the room have bought from a few of our clients, but a lot of them tend to be resellers so you might not actually know some of the names. And we actually don’t own any of the warehouses or the carriers, anything like that. We are more like an Uber for fulfillment so it’s an asset light model and we orchestrate the different assets so that we can affordably offer, you know, free two-day shipping across the country.

Kampel: So how many people here are Amazon Prime members? Ooh. And how many of those people do it primarily for the two-day shipping? Right. I mean, they have other services, video and all that that are—but really people are using it for the two-day shipping. So you know, have we gotten to that point where customers expect it to be there in two days and if you can’t do that, that they are going to move on to the next brand or merchant?

Krakaris: I think—you know Marketplace Pulse, which is an e-commerce analytics company just ran a really interesting analysis looking at the number of products enabled on free two-day shipping, and the number one provider is Amazon Prime, so right now they come out with 100,000,000 products enabled with free two-day shipping on Amazon Prime. If you look at how many Walmart has right now on free two-day shipping, which is the closest, it’s around 2,000,000 products. So to say that free two-day shipping is table steaks, I think is really—even though buyers you expect it—but when everyone’s a Prime member, obviously you’re going to expect it, but outside the market, it’s a huge challenge. And I think—you know one of the really things that we’re excited about is, you know with our partnership with Walmart, powering free two-day shipping for their merchants to really kind of help drive that number up so that, you know when you go and you look at, you know you’re looking at should I buy on Walmart, should I buy on Amazon.

Well, not needing a subscription to buy on Walmart is actually a pretty big value idea. Right now, I just renewed my Amazon Prime subscription. It was $129. Initially I thought it was going to be $120, but I got another $9 tacked on and really for me, I’d say like the main reasons now is to watch like “Jack Ryan” or “Top Gear” or things like that. But as Walmart adds more and more selection on free two-day shipping, I think it’ll be harder and harder to retain the Prime user base until they make big, big investments in video and groceries and those other initiatives.

Kampel: So Ben, you know obviously people can choose to sell their products on Amazon. They could choose to go to direct to consumer using platforms like Shopify. You know, are you seeing a lot of the merchants that you’re talking to, trying to again—I think we talked about controlling distributions, controlling destiny, controlling your data. Are you seeing a resistance to wanting to work with Amazon, you know the merchants that you are working with?

Parr: So yes. But I think some historical context is super-helpful here.  So even ten years ago or even a little bit more, like the main way in which you as a creator of a consumer brand reached consumers was through these distributors who had the stores, not just an Amazon but a Walmart, a Target, a grocer, or whatever it might be and then you ran television advertising, you did other things like that. Then the Internet came and then e-commerce came and then the capability to easily sell things online came. And it made it easier to ship things than it was ever before, and it became a more normal consumer behavior, in part due to Amazon. And now you have options. And so you have to—like when you are selling through a distributor, you are giving a significant cut a lot of times. To them, versus keeping all that for yourself which obviously is something that you want to do.

But it wasn’t really that much of an option until in recent years when now are you seeing brands like Away, like the suitcases creating really powerful direct to consumer brands. And so Away is a great example. They’re one of the fastest growing e-commerce companies in history and they grew super-fast partly because they just built this really direct to consumer brand and they do not distribute through anyone. You’re not going to find it through an Amazon. You’re not going to find it through most things—at least directly to them and they don’t even sell in other people’s stores. They only open their own stores.

Kampel: Is that a—I mean is that a trend we’ve seen? I mean obviously you can talk about Warby and Casper and you know all of that emergence of these direct to consumer brands. I mean are they just really trying to protect that relationship with the consumer and it’s all about that loyalty and building trust with your consumer?

Parr: So in the super short. The loyalty with the consumer, keeping the margins, but also when you’re on an Amazon or something, you are giving them your data as a merchant and we have seen time and time again like—and intelligently for Amazon—creating private-label brands and being very successful with them and so if—your best bet to de-risk yourself is to go completely direct to consumer. It is also the hardest path.  You have to like be great at advertising at like using a Clear Bank and go and buy lots of ads in order to reach these consumers but also build enough brand value, brand equity for them to be repeat consumers. But it is safer than relying on a distributor who can make one decision that could destroy your company.

Kampel: So you know Michael, it’s funny. When we first, you know started talking, I’m like two-day shipping, that’s easy right? Like FedEx, UPS, everyone offers two-day shipping. How hard could it be? But then we really started talking about what your model is about, right? About really using AI and algorithms to optimize where you’re warehousing products, so you don’t need to pay those high costs of shipping which is typically associated with two-day and you could maybe do it via ground locally. Talk about the technology that you had to build so that if I have a 100 widgets, you know you would tell me “This is where you need to distribute your products in our warehouses.” Talk a little about the technology that you had to build to optimize that.

Krakaris: Yeah, I mean if you buy from Amazon today and it’s coming from a warehouse, the chances are, it’s not coming from very far. Versus if you buy anywhere else, from like Warby Parker or something like that, there’s a good chance that package is travelling across the country to get to you. And what we realized is that fundamentally there’s a discrepancy in the market between the way Amazon runs fulfillment and the way everybody else does and that discrepancy was only getting bigger and bigger. I mean Amazon’s now opening up, you know free one-day shipping for—with Prime. And the way we approached the model is that let’s say you have an item and you—two different items. Let’s say we have jackets and we have surfboards. What we do is we create a demand graph for that item. So we know jackets are going to be bought, you know in the Northeast, Midwest. Surfboards are going to be bought in the sunbelt, so California, Texas, Florida. And based on that, we’re going to figure out where to position the retailer’s inventory.

And so Amazon runs a very similar model. The difference though between our models is that, you know Amazon had to sink a lot of capital into these warehouses and today, even though they have a 100,000,000 products enabled on Prime, it does lead to certain constraints because you’re constrained by the number of warehouses you can have in the network. And so fundamentally we made the decision when we started deliveries, that we didn’t want to own the warehouses. And that enables us to expand capacity very quickly. It enables us to add warehouses very quickly in locations that are strategic for our skews. So for example, if there’s a lot of buyers coming in Manhattan, we know that we need to find warehouses that are very close to Manhattan. And when an order is placed, we figure out what is the optimal warehouse to ship that order from.

And there’s a lot of different data points that we use. We look at historical data of carriers and their delivery times and for the most part, we do end up shipping ground. It’s very rare that we ship on a two-day method and that’s why, you know we can offer two-day shipping at a rate that really no one else can do outside of Amazon today.

Kampel: So your margin is dependent on you being able to ship at the cheapest way possible.

Krakaris: Yeah.

Kampel: But getting it still there in two days.

Krakaris: And the big part of why retailers love it—because in the end to a retailer, we are a warehousing shipping company. And it’s very simple for them, but in terms of what they pay, it’s a fixed rate. And so for like an iPhone case, it’s like $5.25 nationwide, two-day shipping. And so that predictability enables them to know, okay, this is what my cost-structure looks like and so then exactly am I going to be making or losing money on this item on day one.

You know, prior to that if you were to work with like a typical third private distance provider, you’re going to have an upfront contract, you’re going to have an account management fee, you’re going to have a receiving fee which is hourly, you’re going to have storage fees, you’re going to have shipping fees that depend on the carrier, how many zones it’s going across. There’s a million different factors to a point where you’re like “Okay, I’m a five-person business. I have no idea what my cost-structure looks like.” You know, not everyone is like a Nike where they can have a team dedicated to logistics and dedicated to operations.

Kampel: So you know Ben, you talk about Messenger and you guys using Messenger to engage with consumers, primarily around abandoned cart, right? So is—is that sort of one of the areas that you’re working in?

Parr: Yeah. And then re-targeting like welcome series, re-engagement, these kinds of different kinds of campaigns and like kinds of flows that merchants are running every day over email and on apps.

Kampel: And how are consumers feeling about that engagement. Obviously, you talked about all the different channels. You know, we’re starting to get more retailers communicating with us via SMS. Obviously email traditionally has been the way they’ve touched us. In Messenger something you’re seeing both high engagement and ROI?

Parr: So high engagement and ROI—like, well the answer is yes. So kind of some of the numbers to like think about. So like on average our merchants are making anywhere between 7 to 25% in lift overall on their revenue by adding Octane and part of it is because we can connect with customers earlier in the customer journey. So like for our software, think of it like email where like you opt in for emails from, from a brand by giving them your email address. You opt in for Messenger or—like by giving them permission by using—like clicking a check box or messaging them first or saying yes or clicking a button—Send A Message button—and then you have that different kind of experience.

And so I think of it as like—just like email—if you’re sending really crappy emails, you’re going to have high unsubscribe rates and you’re going to be looked at negatively. Same thing with messaging. If you have really bad messages, if you are just promoting stuff all the time, if you’re spamming, it’s not going to work. And even Facebook has rules against that. But if you are building like high engagement kinds of things, you’re giving them like—there’s just unique things you can do over messaging, you can’t do over email. Like a back and forth conversation. Learn more about them, personalized in a really strong way. We see that works really well. Open rates of 70-80% or higher. Really heavy increases in revenue.

Kampel: How much of that data does Facebook see and get and understand?

Parr: So one is like—so Facebook doesn’t give us that much data which I think is part of the price of things. You only get what as a Messenger platform first name, last name, profile photo, time zone generally. The very basic info. Which I think helps protect privacy. But like they’re having that back and forth conversation. Facebook can see some of it, but a lot of it they can’t see the structure of it, right?

So in a conversation, for example, like L’Oréal’s Kiehl’s bot is powered by us. They’ll like—the Kiehl’s bot will ask questions like what kind of skin do you have, what kind of product are you looking for, and so they can see people sort of answering that, but they don’t have the structure around like what is that entire list of people, what are they—like-and then we target against that or understand for example, that like these people also chose this different option in this different platform.

So they see some of the data, but not in the same kind of structured way that we do. And they don’t see data across like—so for example, they don’t have the connection to Shopify, so they can’t correlate that with purchase data which I think is a big key to making something really relevant, making a chatbot really relevant for consumer of a brand.

Kampel: So you know again, I think as brands and merchants, like you know is Amazon a friend or foe? Right? I mean we talked about, you know earlier today 100 plus private label brands using the data to actually compete with their sellers, but on the other hand obviously a way to scale and distribute. You know traditionally, how are you—well how are you starting to see brands think about Amazon? Are they starting to pull back? You know, you talk about Walmart marketplace, you know seeing an influx of—of merchants. You know, are people looking for alternatives and alternative platforms just because Amazon had amassed too much power?

Parr: On the Amazon side it’s—I mean it’s one of the best channels to be on. I mean it’s incredible when you look at selling on Amazon in terms of the film and fees that they give you, the Amazon referral fees and then just how they’re about to give you ads product that’s really strong. To get your product out there and acquire new users. It’s an incredible platform for that.

Krakaris: And so my recommendation is never—don’t be on Amazon. I think for a lot of merchants, Amazon is a really critical piece to their business, but you shouldn’t only be on Amazon and I think that’s—that’s one of the things that has been a really big challenge for our sellers. I mean, we work with some—you know some of the top ten Amazon sellers in the world today. These guys are pushing well over $150,000,000 revenue in Amazon annually, but Amazon makes up 95 to 98% of the business. And that’s just not a sustainable business model because-

Kampel: Well, we’ve seen Amazon literally look at those businesses, right, and say okay there’s so much through-put there, we need to be in that business.

Krakaris: Yeah, you’re playing in their world, right? And you know what happens is they’ll go, and they’ll source certain products that they—that you find that are selling well and then you have to go and buy new products. And that’s actually part of the Amazon engine. They use third-party sellers to find out what’s selling well, they’re either going to private-label it or they’re going to go and source that product and purchase it in wholesale one key relationship and then third-party sellers are stuck having to find new items. And you know diverse defines a huge piece of any business and I think for a lot of sellers, Walmart is—a lot of the big Amazon sellers, Walmart is emerged as one of the really big next channels. Shopify has definitely emerged as a big channel. And there’s new ones too. I mean I don’t know if anyone here is familiar about Google launched theirs, Shopping Actions Marketplace. We’re obviously at Deliverr really excited about that and—and we’re trying it out, seeing how our sellers are doing fulfilling those Google orders. Facebook just launched their marketplace for B to C Brands and so there’s definitely a lot of activity coming out. I think also the nature is going to change from what it is today to what it will be by the end of next year.

Parr: I know we held—like we’re almost out of time but—it’s a very frenemy kind of relationship where like—it really depends though on what vertical you are. If you’re selling a commodity—like an Amazon makes a lot of sense especially because like a commodity it’s not about the specific brand affinity. If it’s a very unique kind of product, a unique kind of brand affinity there is a good argument to be made don’t—don’t be on Amazon or another platform. Be on just your own platform built on top of Shopify, really build your own distribution channels, you know. Because they could see like—this is our cute little mascot Octy [ph 0:17”43.6] if you’ve been wondering. So if they saw—if Amazon sees that Octy’s popping on—on Amazon, even if it’s only a small portion of your business like, it’s logical for them to go and start building that part of their business as well. Because you are giving up something by being on that platform and so you can only really—it’s a focus thing too. Like you could focus on building like your own channel or really focus on being on the Amazon channel but when you’re at the super-early stage, it’s hard to build both and you kind of have a decision to make.

Kampel: Great. Well, we are out of time but let’s thank Michael and Ben for joining us today.