Episode 19: While Amazon, Walmart dominate retail media there is room for every retailer

Amie sits down with Nikhil Raj, chief business officer at Moloco about Retail media's emergence as a significant revenue source for major brands like Walmart and Amazon, and how smaller retailers might be missing out on the boom. Recorded at Shoptalk Fall Chicago.

Transcript

Intro:

Coming up, on this episode of Retail Untangled.

Retailers keep things on the shelf that don't expire. Advertising inventory expires the instant you leave the website.

Your eyeball is gone, the ad impression is gone. So it's important to optimise in real time.

Amie:

Welcome to Retail Untangled. My name is Amie Later and this is the podcast where we speak to retail industry experts and find our business hacks to help you succeed. You won't find these gems anywhere else and we have some superb stories from the coal face as well as helicopter insights from retail industry leaders. This week we're bringing you insights live from Shop Talk Fall in Chicago.

Amie Larter (00:00.11)

Retail media and retail media networks were the hot topic at Shop Talk Vegas earlier this year. And it's no surprise that it's still top of the agenda at Shop Talk fall.

Today I'm joined by Nikhil Raj, Chief Business Officer, Retail Media at Moloco to chat about how what was a relatively new concept just a few years ago is capturing a bigger portion of digital ad spend and looks primed to become a significant revenue source for major brands. Nikhil, you've been within the retail media space since the early days where you were involved in the launch of one of the big two retail media networks, Walmart Connect.

I'm keen to understand how retail media emerged as a significant revenue source for major brands like Walmart and Amazon and what factors have contributed to its rapid growth in recent years? 

Nikhil:

Well, thank you for having me on the podcast. 

Amie:

You're welcome. 

Nikhil:

Really appreciate it. I would love to say that I was involved in the launch of Walmart Connect, but that was much after my time. 

Amie: Oh right. 

Nikhil:

I was there at the beginning when we launched a thing called Walmart Exchange, WMX in 2011, which evolved into Walmart Connect, which is a lot bigger and a much better operation than I could have ever imagined. I'd love to take credit for it, but I can't. 

Amie:

The early days of Walmart Connect then. 
Nikhil:

The early days of Walmart Connect might be more appropriate. That's right. Yeah. So what makes it so interesting? I think there are a couple of things. Or interesting from the retailer's perspective, why is it getting so much attention? The most important thing is the retailers have access to the consumer sales data. That's where it starts. That's how it started for Walmart exchange when one of the largest advertisers in the world, Procter & Gamble, a customer of a vendor of Walmart, wanted to understand the impact of the media they were buying in terms of in-store sales. 

And that's kind of how this whole thing got started saying, can you guys build us a report for the advertising that we do? So we started with sort of giving them a little insight into the shoppers that were exposed to the media that they were buying and the impact of sales, which really was not available anywhere else. Even today, if you want sales data from a Facebook or a Google or someone else, you can't get it unless the retailers share the data with those platforms. So that fundamental difference of knowing who saw the impression and who bought the product, that connection is what makes retail media so unique. 

And with that unlock, or that data being unlocked with all kinds of technologies like the cloud and the machine learning, all that stuff, advertisers started paying more and more attention to this channel. And now it's really exploded in the last few years. 

Amie:

And you could say that it's really grown, right? 

Nikhil:

Mm-hmm, mm-hmm. Let's see, when we were doing Walmart Exchange, of course it was at the start. Yeah. There's very little revenue on Walmart's ad business. And I think last year they announced four billion dollars, which is incredible. But at the same time, Amazon is now a 50 billion dollar advertising revenue business. That is just incredible if you think about the growth in that industry, rivalling that of Google and Facebook. 

Amie:

Yeah, I think Amazon's dominance in this space is undeniable. What strategies do you think they've implemented or employed to establish itself as a leading player? And how can other retailers, I mean, would be hard, but learn from their success?

Nikhil:

Actually, the playbook is not that hard. Okay. If you think about it, there is a couple of things that are driving Amazon's ad business. Number one, it's a marketplace. So when you think about advertisers, you should think about them not only as the billion dollar advertisers such as Procter & Gamble and Coca-Cola and Pepsi, you should also think about the hundred dollar advertisers like the long tail of sellers that Amazon has in their marketplace. 

That's driver number one. So $30 billion of Amazon's 50 comes from $3 million, $3 million sellers buying $1,000 of advertising each year. That's $30 billion or $10,000, right? So that is the large part, 60% of Amazon's ad revenue is in that marketplace sector. Every other retailer, if you look at it, Walmart has leaned into marketplace, other retailers are looking into leaning into marketplace. So leaning into a marketplace business model enables this long tail of sellers to participate. That's number one. 

Number two is monetising your own ad inventory versus monetizing other people's third party or offsite as it's become known in retail media. What does that mean? It means make sure that every impression that you show on your own website, in your own app, is maximised for advertising. Which means the consumer sees the best thing they can see and they click on it. So if you have 100 impressions to sell, you want to get 10, 20, 30, 40 clicks out of that 100 instead of 2, 3, 4. So you could 10 times your advertising revenue of your own website and your own apps rather than taking that money and going off site and buying ads on Facebook or somewhere else with your data. So that's strategy number two. 

Amazon doesn't do it. Walmart doesn't do much of it. They do some of it and controlling your data and using that to personalise the advertising experience on your website and in your app is much more productive than using that ad revenue to buy ads elsewhere where you can't control. That's strategy number two. And strategy number three is related to the marketplace. You want to enable this as a self-serve and you want to give them the performance. So what you want to do is let them buy ads on your website. Don't try to manage it. Don't try to put minimums. 

Let them do it on a self-serve basis. So what that means is it unlocks many, many, advertisers. So when you bring in lots of advertisers, then there's more demand for ads on your website. So you open up more ad slots that you can fill with more relevant ads because you have more advertisers with self-serve. So these are the things that you want to unlock. You want to deliver highly, highly relevant ads that consumers click on, when you have good relevance, you want to open it up and get many advertisers in. 

That drives up the demand and the prices. When there's enough demand, you open up more ad slots. That's the playbook, Amazon's playbook. And now on my Amazon feed, almost everything is an ad. That's where they've gone to. 

Amie:

And the playbook makes a lot of sense, but you see in the market there's very much two players doing really well. And then I think everyone else there doesn't seem to be the growth there that Walmart and Amazon are seeing. What's the roadblock here? 

Nikhil:

It's the first step. So the playbook is in that order. What was the first step? The first step in the playbook was make the ads relevant using our first party data on site. That's where most of the retailers are stuck. Because the ads are not relevant, consumers are not clicking on it. When consumers don't click on it, you hit a ceiling with the amount of ad revenue you can make. So you take that money and go offsite, which doesn't help your business, or you just can't get enough money from your advert. The budgets just don't come to you because you don't have performance. 

So we at Moloco, we feel like this is where everybody's stuck, where they're unable to deliver a personalised, relevant ad to the consumer because you've got to take care of the consumer, without which the advertiser or the ad platform don't go anywhere. 

Amie:

There's just no engagement. 

Nikhil:

There's no, there's no sales. It doesn't matter if you have 100 impressions or a thousand impressions or a million impressions, nobody's clicking on it and buying. And if they don't click on it and buy, the advertiser doesn't get a return and the consumer gets a bad experience. And as an ad platform, you make no money. That's where the first step, the relevance, the advertising relevance, that's where it all starts. 

Amie:

Okay. And so the way to fix that is really with a platform like yours.

Nikhil:

Absolutely, that's what we believe. That's where we made a bet. Unlike most of our competitors and other vendors who've gone on to build really, really sophisticated but cross-platform products, on-site, off-site, in-store, we've said no. We're going to focus on on-site. We're going to break this logjam that our customers are in by providing them technology that has only been available to date inside of Google, inside of Facebook. They don't sell that ad tech.

That is their proprietary advantage. So what Moloco has done is gone and hired those people. We have 600 employees in the company, almost 700 now. 400 of them are ad tech engineers from Google and Facebook and Amazon. These guys and girls built those ad systems. So we've brought them into Moloco. They've built that kind of technology at Moloco. Now we're bringing that to our customers as software so they can unlock that entire first step in that value chain. That's what we're doing. 

Amie:

And so we've spoken about a lot of the roadblocks and challenges, but can you give us an example of some retailers potentially that aren't Amazon or Walmart that are doing well, that have kind of crossed that first step that you've spoken about and are really using, stepping into retail media in a big way? 

Nikhil:

Yeah, I think some of the very large advertising businesses have started to have realised that and have started to go in that direction. I don't know if I can name any names, but certainly if you look at the leads of engineering at some of the large retail media networks that came from Amazon ads and Google ads, you'll know which retail media networks have figured it out and which ones haven't. So that's the key. The second thing I'll say is that this is not a new thing for marketplaces. Pure player marketplaces like Wayfair, DoorDash, Uber, Uber Eats business, and many of our customers are in food delivery. 

The digital native businesses that started off only digital, Expedia, these companies have figured out this model a long time ago. And if you look at their advertising businesses, they're run by sophisticated ad tech people that have built this type of capability for their advertisers. And they have hundreds of thousands of advertisers. Hundreds of thousands, not even thousands.

Our largest customer has 85,000 advertisers, one company. It is at a different scale that you operate when you bring them self-serve and let them opt in and deliver performance. So some of the largest food delivery companies around the world, some of the largest marketplaces around the world have figured this out. Without actually spoiling any one specific flip card in India, Mercado Libre in Latin America, Coupang in South Korea, Bold.com in the Netherlands, DoorDash, UberEats on the food delivery side, not on the grocery side, on the food delivery side, have all figured this out. And they're doing extremely well. 

Amie:

And I mean, it's profitable business, isn't it? Because when you look at Amazon, their advertising made up 68 % of the total profit, did it not? I think I was reading that in one of the webinars that you did. 

Nikhil:

Yes, I mean, that's incredible, isn't it? The core e-commerce business is at 2 or 3%, 4%.

But the advertising business comes in at an incredible margin, and this is why onsite is important. You don't get that margin offsite. So if you get a dollar and spend it on your own website, that dollar, almost the whole of it, marginal dollar, is 100% margin. Because you've already spent the money to build the website, to put the content in front of the consumer, you've already got the consumer there, you've done everything. Now you just show them an ad, it's 100% margin. 

Because you've basically incurred all the costs of putting that impression in front of that consumer. That's why it's such a high margin. And that's the problem with offsite. Because if you took a dollar and did not spend it on your own website,

Now you have to pay a Facebook or a Google that dollar or a part of that dollar, which is mostly all of it, most of it, and you keep like a 10% margin. So you can't build a 6 % margin business if you're trying to build a retail media network with offsite advertising. It has to be on site. It has to be personalised machine learning AI, very sharp talk. We're talking about personalisation a lot. If that technology or capability is not part of your advertising impressions. 

Amie:It's just all white noise then. 

Nikhil:

Yeah, it just comes back to the first step of our three-step formula or playbook. People don't engage with it. Exactly. 

Amie:

And then you don't make money. 

Nikhil:

Nobody makes money, right? You don't get repeat clients and 

Nikhil:

And your consumers are like, why am I seeing this irrelevant ad in front of me? 

Amie:

And they're actually mad, and then the advertiser is mad, and then everyone's mad. 

Nikhil:

Exactly. And then you have this situation we call ad blindness that starts to develop where consumers are like, that's an ad. I'm just going to scroll right past it because I don't want to click on an ad. Whereas if it's relevant, everything can be an ad.

Amie:

And it makes so much sense and I'm intrigued in terms of what you're seeing from a growth perspective because there's obviously such high margins involved with something like this, Amazon's advertising business. Are you seeing this take a greater share of that, you know, the non Amazon and Walmart, or is it still people struggling with finding that scale up opportunity? 

Yeah, so this is, we at Moloco strongly believe that Amazon and Walmart are getting an unfair share of the advertising opportunity because of the technology gap that they've built between themselves and everybody else. They started a long time ago. They started a long time ago and they invested in the right types of technology, specifically Amazon, Google, and Facebook. Walmart realised the playbook only a few years ago and they started to go in that direction. 

Amie:

Yeah. 

Nikhil:

And our point of view is you need, as a retailer, that kind of technology to close that gap because if you look at it from the advertiser perspective, I give a dollar to Amazon, I get $4 back. I give a dollar to someone else, I get $2 or $3 back. So other than the dollar that I have to give them because they're my customer, I have no further incentive to give them any more money. So this technology gap has created an unfair market share gap where more money is going to Amazon because the ads are more relevant over there, they're getting more return and therefore the advertising revenue for everybody else has kind of stalled out. 

Amie:

Yeah, and it makes sense. So if you're not on the scale of your Amazon or your Walmart and are looking to justify or, well, essentially maximise your investment in a retail media network, what factors should you be considering? 

Nikhil:

Yes, look, like I said, we don't think it's a scale problem. Let me give you very simple maths on this. You've got a hundred impressions to sell when Amazon's got a million impressions to sell. You would think it's a scale problem, right? But if you sold 20 impressions of your 100, that's a 20 % click-through rate. And let's say Amazon had 1,000 impressions to sell because they're 10 times bigger. And they also sold 20 impressions. Basically, both of you made the same amount of ad revenue, but Amazon used 1,000 to do it and used just 100.

Unfortunately, the situation is reversed right now. Amazon is the one that's getting the 20% click through rate and you are the one that's getting the 2%. But that is the problem you have to fix. Again, it comes back to the same point. If you have a limited scale like you're mentioning, it's even more important for you to be even more productive with that limited amount of scale that you have. And once you are more productive, which means you're getting the return for the advertiser the dollars will follow. 

There is no question it has happened over and over again. It happened in the 80s with Nielsen when they launched measurement. A lot of the money moved. At that time, most of the brands were spending 3-4% of their sales in media. After Nielsen introduced measurement, it became 8 or 10%. And now with retail media, it's becoming even more for brands spending on advertising as a percentage of their sales has gone up every time there's been an inflection point in the measurement and performance of advertising. 

I think we're at another inflection point now with retail media, and those that do it well will command a higher share of the brand's advertising spend than those that don't. It's very simple. 

Amie:

It makes sense. So looking ahead, and you've obviously been following this space from very early days, how do you think new technology, you know, like generative AI, will impact this space? 

Nikhil:

That's a great question. The generative AI has two parts to it.

The first part is the G, the generative. So making creatives, making text, making audio, video content, all of that generation helps you create a much better and much more personalised consumer experience. The other part of generative AI is the other end of the thing that's called chat GPT, the T. That's a transformer. A transformer is a new kind of neural network, or a machine learning or artificial intelligence technology that was developed in Google in the late 2016, 2017 was the paper that enables you to do ads personalisation at scale. So on the one hand you can generate things or on the other hand you can personalise things. What we know today as recommendation systems, Netflix movie recommendations, TikTok feed. The TikTok feed is so engaging because underlying it is this thing that recommends you the very thing that you think you want to watch, right? 

That whole recommendation system technology comes from the tea of the chat GPT transformers. That is a technology that Moloco is using to enable personalisation of advertising on retailer websites. So when you do that, you end up creating a much better user experience going back to the beginning of our playbook. Better user experience, more advertising clicks and revenue, more platform ad sales.

It's a very simple flight. That's how the AI and the machine learning, the generative technology and the transformer technology is really transforming all industries, not just retail media. 

Amie:

I was about to say, that was my next question, is this just the retail media space? It sounds like there's so many applications. 

Nikhil:

Yeah, so at the end of the day, you're making a match. 

Amie:

Yeah. 

Nikhil:

And the more data you have to process to make a match, the better the match is. But you have to process that data. And the more time it takes, the more computational resources, the more data storage, the more data centres. So, doing that match is a very simple thing to say. Actually, getting the technology to do it is very, very hard. It's expensive. So this transformer technology unlocked the ability to create this matching, whether it's a person to a product, whether it's a video feed to a person on TikTok, whether it is an image that's customised to a person. 

All of these things are like literally infinite types of combinations of things that simply couldn't be processed at that scale before the convergence of transformer technology with cloud technology and machine learning, all of it coming together and the costs of hardware dropping down. It enabled all of these things to come to life. And it doesn't matter. You're trying to match one thing to another, wherever you're trying to do that, this technology is applicable. 

Amie:

Nice. 

Nikhil:

Okay?

Amie:

Yeah. Very good. Okay. 

Nikhil:

So I hope that was clear. 

Amie:

That was really, you make it sound very clear and you also make it sound simple. 

Nikhil:

I hope so. 

Amie:

Yeah. It is not. Now just in terms of this podcast, is there anything that you feel like, you know, is a little bit of a golden nugget for retailers that we haven't spoken about? And I can ask a question on that. 

Nikhil?

That's a great question. Look, retailers, particularly the larger ones, have to consider the cost of building something like this. Yes, the technology is out there. Transformers are now invented. But the reason for Nvidia and its stock price going up so much is because a lot of people are trying to do this. And so it's gotten even more expensive. It is not easy to build these systems internally. 

So what I retailers should do is think about their customers, think about their consumers for the customer of their retail operation and think about the brands which are the customer of their advertising operation. The retailer should focus on building that relationship, making sure that their customers or consumers are getting a good experience and their advertising customers are getting a great performance out of the ad platform and leave the technology to people that have optimised these highly expensive GPU, TPU operations and made that cost effective so that they can leverage that and deliver that experience and that capability at a much lower cost than if they were to try and do it themselves. The stuff is not cheap, it's not easy. 

Amie:

You also have to have whole teams and experts, I'm assuming, to consume all of that and then use it properly. 

Nikhil:

Yeah, we have a whole team of engineers figuring out how to optimise GPUs and TPUs. Why would a retailer want to do that? They should focus on the consumer. They should focus on the advertiser. And they should enable us to build- companies like us to build on top of their first party data, make sure it's private, make sure it's siloed. So your data as Nordstrom is not being used to help your competitor, I don't know, Sephora or Macy's, right? So you want to make sure that these sort of rules are in place and within those rules or those boxes, you allow the leverage of this technology.

Amie:

And as well, it's sort of a new, it's sort of a, it's a new customer for these businesses as well. It's not who they're normally selling to. And so you would assume that that itself is something that needs to be built up and the relationships need to be. 

Nikhil:

Look, you're opening a whole new part of the conversation we haven't even touched, but just to take a step back and give you a little bit on that angle, there are many, many new things here, including, I'll come to your point in the end. Number one, retailers are a B2C business. Advertising is a B2B business. Number two,  retail is a very thin margin business. And on-site advertising, as we talked about, is a very high margin business. Number three, and this is the most important point, or maybe not, number three, retailers keep things on the shelf that don't expire. Advertising inventory expires the instant you leave the website.

Your eyeball is gone, the ad impression is gone. So it's important to optimise in real time. But the most crazy thing over here is that the retailer's vendor is now the retailer's customer. So you literally have to flip the roll across the table. Hey, Procter & Gamble, let me buy your shampoo. I'm your customer. Change your hat. Hey, Procter & Gamble, let me sell you some ads because I've got some consumers who want to buy our shampoo. 

Now that is a really difficult thing for retailers to do, to flip from being a merchant buyer with a big budget to buy shampoo, to a seller of advertising. Now you're trying to tap into Procter & Gamble's marketing budget. That is a very different shift, it's a different team. And so this is a brand new business build for many of our customers. So prior to Mocolo, I was at Bain & Company, a consulting firm, advising clients around the world on this very issue. 

It's a new business for them. It's a different product, has a different margin profile, has a different expiry profile, has a different go-to-market approach. All of those things. And most importantly, you're now selling to the person you're buying from. And that is sort of a concept that is really difficult for a retailer to get through. That's what makes this unique. 

Amie:

But doable. 

Nikhil:

100%. Look at Amazon, look at Walmart. 100 % doable.

Outro:

A big thank you to Nikhil from Moloco and we hope you enjoyed this episode of Retail Untangled. To hear more insights as they land, feel free to like and subscribe wherever you listen to your podcasts.