← Behind the Numbers: an EMARKETER Podcast
Behind the Numbers: an EMARKETER Podcast
Measurement and ROI in Commerce Media | Behind the Numbers Special Edition
00:30:14
Loading summary
Transcript32 lines
- [00:07]
Marcus
Welcome to a special edition episode of the eMarketer podcast, behind the Numbers. I'm Marcus. And today we have another special episode from the eMarketer Summit, Commerce Media Trends 2025 held at the start of May. We have for you a panel discussion that examines sophisticated approaches to measuring commerce media success, maximizing return on investment. Our experts discuss attribution approaches, cross channel measurement strategies and KPI selection for optimal performance tracking. Emarks Principal analyst Sky Canavis hosts Christine Gramier, Vice President of Global Measurement products at Liveramp and Mike Glasser, Vice President of commerce media of PepsiCo and Michael Campi, Senior Vice President Marketing for Mave. Enjoy the panel.
- [00:50]
Sky Canavis
For those of you who are just joining us now, welcome. I'm Sky Canavas and I'm a principal analyst here at eMarketer. I'm I'll be moderating this session made possible by Liveramp on measurement and ROI in Commerce Media. It's one of the most complex problems facing advertisers today, so I'm very glad to be joined by an expert panel of speakers who have deep experience and will be sharing some terrific insights. So before we begin, let's do a round of introductions. Can each of you tell me who you are, what you do and how, where you're joining us from?
- [01:26]
Christine Gramier
Hi there. I am Christine Gramier. I lead the insights pillar at the Liveramp product team. We have the pleasure of working with most of the media networks out there across all the different ways that they might be working. So I'm really excited to dig in today and I am joining you from the beautiful city of Los Angeles.
- [01:50]
Marcus
Hey sky, thanks for having me. My name is Michael Campi calling in from Austin, Texas where it's already summer and excited to be with you guys. So I lead the marketing team at Maeve. Maeve is for dog lovers out there. It's the best possible food you can give your dog. We're a digitally native company so we sell primarily online. We make human grade raw dog food, treats, supplements and seriously, if you have a dog, check it out. It is the best. I've been in the CPG industry for about 20 years. The first 10 I was in what I call big CPG. I was in market research and brand management. And then about 10 years ago I made the switch over to growth stage companies where I'm really excited by the opportunities that growth stage companies have in partnering with founders and trying to create insurgent brands and take advantage of some of these things. We'll talk about today in terms of measurement. So yeah, Happy to be here.
- [02:50]
Mike Glasser
Hi Sky. Hi everyone. My name is Mike Glasser. I lead the commerce media team at PepsiCo. Our team's responsible for end to end strategy, planning and execution of media in support of the full beverages and foods portfolio wherever and however our consumer prefers to shop. So that could be grocery or e grocery convenience aggregators, Direct commerce. We do that in partnership with our retailers, their media networks as well as some great media and technology partners out in the industry. And I am joining here today from New York where it's not quite yet summer Michael, but we are holding out and excited.
- [03:28]
Sky Canavis
I'm with you Michael, here in Austin and it's great to have you all. So now let's dive in and I want to start with one of the foundational issues and that's that as retail media has grown, the landscape has become increasingly complex and fragmented. The and comparisons are hard when standards vary across channels and platforms. According to a survey research from the association of National Advertisers, more than half of retail media advertisers identified the lack of standardization as their biggest retail media related challenge. So I'll start with you Christine, since you have like that big picture view of what truly needs to be standardized and what are some of the biggest obstacles you're seeing today and what, where's progress coming from in this area?
- [04:18]
Christine Gramier
Okay, well I'm excited to kick off with a big challenging question. Really I think that when we look at this ecosystem we see, we see companies who are not technology companies necessarily having to dive in and become technology companies, measurement companies, advertising companies, all of the sudden. And so really the lack of standards we see as part of the maturing process of this industry. And so when we took a good look at the challenges that we're facing, suppliers in these, in these retailers and the retailers themselves, we identified a place that we could really help was in what we would consider the architecture. So the architecture of how data is being shared across companies, how measurement is being executed, how publishers, retailers and suppliers are all working together. So we do think it's solvable sky to, to address that part of your question. And we really see that by creating some standards around the architecture, it helps legal feel more comfortable with how data is maybe moving between companies, legal and compliance. It also creates the money flow in a more streamlined way. So a lot of these budgets are coming from different places inside of retailers or inside of suppliers. And so by creating an architecture in a way of thinking about it for everybody, for how money and data flows, we've really seen a lot a Lot more commonality, start to grease the wheels and help data move between companies. We actually built a steering committee of publishers, retailers and suppliers to help this over the last six months. And really it's to complement many of the other things going on out there. And we've run a lot of pilots to show everybody how easy it can be. So that's how we're investing time and energy. We are supporting lots of different types of measurement specifically, but we see the architecture as the place that really we could help everybody move the quickest First.
- [06:24]
Mike Glasser
Christine, I'll build on that, maybe offer sort of the brand perspective. We're navigating broad portfolios across a whole different number of retail partners. Again, we kind of want to be wherever our consumer prefers to shop. So standardization for us is critical, right? Standard ways to build audiences to bid and buy on, media to target. And certainly from a measurement perspective, I think the sooner and the more progress we can make there, a, the more ground we can cover, we can reach our consumers in more places and engage them. But B, it allows us to adopt learnings and apply them to different parts of the industry. And so I think where sky you asked about where progress is being made. Christine, you have a great example on sort of data infrastructure and frameworks there. I think the work that the IAB and MRC are doing around accreditation is really good and we're generally excited about that movement. And even outside of that, just the more open and honest and transparent conversations we're starting to have with our media network partners about measurement methodologies and attribution, what, what brands and SKUs get credit, I think that's how we make progress to this, this space where it becomes easier and, and more efficient to, to be doing media planning here.
- [07:41]
Sky Canavis
Michael, anything to add on that front on what you're seeing or interested in?
- [07:45]
Marcus
No, that's great advice. I think that I'm coming from a company where we have about 20 people. We have to make really difficult decisions around where we're putting time, money, resources, effort. And when it comes to standardization, it is really difficult for an earlier stage operator because a lot of times retail media networks, they may have their, they're, they're, you know, a very specific attribution window like Mike touched on, like how do they actually attribute sales to certain products? So then there's nuances to each and it can become overwhelming for, for a small team to feel like you have to become an expert of all these different methodologies or attribution metrics. And so standardization would for us would be that we get to a point where we can more quickly make decisions. The data analysis is much more simple for us and ultimately at the end of the day we can just deploy resources much more efficiently.
- [08:46]
Sky Canavis
And as we've been talking about, there's been today there's been this huge proliferation of retail media networks, just dozens of them. And with this long tail of retail media networks now competing for the attention of brands, I want to ask what are some of the attributes or features that a retail media network can offer that would make it attractive to you and more compelling for you to spend your ad dollars? Maybe Mike, I'll start with you since you work with a lot of retail media networks and then Michael, you can jump in because you work with one and what you would be looking for to grow your presence on retail media.
- [09:24]
Mike Glasser
Yeah, thanks guy. And I think this, this flows nicely from the standardization question because I do think that's a big part of what we look for is a good understanding is this a common way to build audiences. Think audience targeting and display or even search, bidding and buying models. Think second price auction for search. And so that gets us part of the way there. But from where I sit, where we generally are very pro retail media networks, again we want to invest where we see our consumers shopping. For me, we're managing a portfolio of thousands of SKUs, dozens brands, thousands of SKUs. And our team is largely hands on keyboard. And so any day there are millions of decisions that could be made. Obviously no human or a team of humans could, could do that. So we do lean on automation capab and we're fortunate at PepsiCo to have an in house product and engineering and data science team that builds out capabilities. We also partner externally with some of the best of the best technology platforms that allow us to operate very, very complex scaled marketing campaigns. And so I see them almost as two sides of the same coin sky you've got. Standardization allows us to operate sort of at scale and speed and understand what the opportunity is. And then automation allows us to make decisions to really deploy our resources in the most efficient, efficient and effective way.
- [10:44]
Marcus
Yeah, that's great. I think, you know, if I could call the person in charge of the retail media network we partner with. I think the, the things that are most important to me at the end of the day is, is just, I mean the highest level thing is simplicity and predictability. So when, you know, when we're investing in a retail media network, we're doing it because we want to drive trial, get our product know in more households and Sometimes the link between the advertising on the retail media network as it then shows to like scan throughs at, at a checkout, they don't, they don't always sync correctly. And so there's a level of understanding predictability. So like if I put in X amount of dollars, how much sales am I going to get out? And that like simplicity of then taking the data extraction from that. Like we, we use a lot, we use a lot of the native tools within retail media network reporting, dashboarding and sometimes retail media networks make it kind of complicated to pull data out in a way to manipulate. And so some of the ones that I've worked with in the past that are most useful to us is just when we can take that data and put it into the way our system and the way that we're looking at the data to do analysis. And the last thing which I think a lot of them seem to be lacking this is that there's really a lack of ability to test and iterate within a campaign and do these like small tests in a structured way where you have a control on a test and you can a b test to treatments across both. Maybe you're testing different ad creative, maybe you're testing different bidding strategies. A lot of them really lack that. So if you want to test something you're kind of exposing your whole business to the test and that's risky for smaller businesses like my company.
- [12:37]
Sky Canavis
And then I think apart from the retail media networks, the theme of this year's summit is commerce media. Because that's what we're talking a lot about today. We've seen the rapid rise of commerce media networks that are non retailers, the financial services, travel and hospitality players, transportation and delivery intermediaries. What would you say is the value that they would offer your brands that say distinct from what retailers have on offer today? Michael or Mike, we'll start with you.
- [13:10]
Mike Glasser
Yeah, I'll jump in. It's a space we're really excited about. I think clearly whether it was Covid kind of accelerated the behavior, we see the rise of what we'll call aggregators. So the doordashes and Uber Eats and the instacarts of the world again I think this is the consumer voting with his or her wallet on where they prefer to shop. And those platforms are growing and because they're growing we want to make sure we meet our consumers there. So they're an important part of our marketing approach and our media investment and I think will continue to be, I think the financial services part of it. Sky is really Interesting. Truthfully, it's not an area we've gotten into as much, but I think what intrigues me about that is we're often looking at a narrower opportunity to engage with these consumers. Obviously the Chase Media Networks or the PayPal ads are seeing a much broader journey before shopping and after shopping that could truly help you understand what does a consumer want or need. And so I think it'll be interesting as that opportunity grows and we find ways to bring more engaging and more relevant advertising to our, to our shoppers that becomes an interesting opportunity for us.
- [14:19]
Christine Gramier
I think this movement of really I've been in the industry for 20 years and data quality is a foundational issue that we've all struggled with for a very long time. And digital advertising brought this amazing opportunity to have target consumers in a one to one kind of approach. But then it got really messy with all of these different kinds of data approaches. Whether it was cookie based targeting or tracking people when they didn't know it. What I find absolutely just motivating about this era is this data from each of these different data owners is authenticated, it's consented, they're exchanging the value with the consumer on the other end, getting the agreement and then they're really, you know, making advertising better for the consumer in the end. So Mike's examples, he just used Uber, jpmc, United, really delivering better experiences with advertising with high quality data and then the measurement. Michael, I promise you we can get you good incrementality tests. I'm not sure who's not getting them for you, but, but, but good increment tests built right into that structure. Transparency is there, the data is there. It's just a really exciting time to be in this industry. I think we all kind of shuddered at the thought of cookies going away and privacy really impacting our industry in a negative way five years ago. But I really think it's actually created a much brighter day of better data for all of us to be using and more transparency across the board.
- [15:57]
Sky Canavis
And I think now adding to the complexity is that some of the biggest growth opportunities for commerce media networks are those off site and upper funnel opportunities as we move more into social and streaming and in store and then measuring campaign success looks different from lower in the funnel where you'd have a different focus, for example on conversion based metrics. So I want to ask how some of the metrics you use to evaluate campaign performance vary across these channels. The ones that are more geared towards awareness or consideration.
- [16:34]
Marcus
Yeah, I can jump in on that.
- [16:36]
Christine Gramier
Perfect.
- [16:37]
Marcus
Yeah, we're kind of in the thick of this now. We somehave launched on Chewy in in January and Chewy's develop you know developing their advertising solutions for for vendors and there's on site advertising which is really similar to what you'd expect on you know Amazon was sponsored search sponsored brands we were bidding on keywords but for us to as a small smaller brand like how can we use Chewy to help build awareness of the brand and so we do utilize their off site advertising capabilities. It's primarily through through meta and we think about the way we measure different steps of this campaign as we're trying to build awareness of the brand. So therefore what's the right metric? It's probably not the cost per new subscriber on Chewy. We're looking a lot more at the cost per click metrics engagement metrics around the ads and then we're using Chewy's their zero P data to target against people who we think would be open to mave mid funnel. That's where we start to look more at cost per order information new to brand buyer information. So what was the cost to actually get someone who buys other better for better for your dog products to purchase MA and how does that compare to other other places we advertise to understand is this a place that we can continue to use to grow awareness of the brand even beyond just purchasing on Chewy as a whole but just awareness of the brand and then on the bottom funnel tactics it's really simple. It is you know this, this is someone who's considered and shopped the brand on Chewy. What is the, what is the cost to convert that customer?
- [18:33]
Mike Glasser
Michael I can, I can build and maybe give a the larger cpg you know sort of multi retailer multi brand perspective. Sky the first way I think about it is is sort of the measurement KPI really has to be aligned to the brief and so whatever the program is if it's on site off site we'll bring in our measurement science team. We're making sure that brief is very clear and the KPI is aligned back to the brief. If it's a more of a brand awareness and equity campaign, you know inclusive retail media we'll look at reach and resonance and short term ROI as measured by our internal MMM models. If it's more sales driven then we probably like many many cpgs are pushing hard to get to an incrementality understanding and from incremental roas to get to understanding how lifetime value changes over time. Right. Are we bringing people into the portfolio, are they wanting to come back and everything? And so Michael, the reason why I add that is I think in your DTC world we're taking inspiration by a lot of the way classic DTC brands operate, looking at cost of acquisition and lifetime value and saying how can we replicate some of that in the retail media world? When admittedly we might not have all the same data sets. But I think importantly with certain partners, with certain clean room solutions, we are getting there. We can start to build that understanding of understanding if they're a new shopper in our portfolio, are they coming back? Are they buying more of the brand? And I think that's where some of the really exciting work is happening right.
- [19:59]
Sky Canavis
Now and in this space. What's the state of collaboration now to support measurement or how do you see that evolving? Christine, what's your take?
- [20:11]
Christine Gramier
I think a perfect setup from those examples Michael and Mike just used in the last question. We see publishers, suppliers and retailers coming together in really, really collaborative ways. We offer a clean room architecture that makes this easy for them all to come together in a very neutral way. A really simple application that hits on this is there's a public case study that's widely available, the Albertsons Media collective along with Pinterest for Mondelez, the brand you all probably know, Triscuit crackers. And so think about this as off off site advertising. So Albertsons took their really rich data about people who've bought crackers in the category and then specific to, to the brand, they created a set of audiences that were exactly aligned to a Triscuit program. That program was all about recipes and, and really that program, they were able to measure a direct sales increase together. Where Pinterest allowed the ad exposure data shared into a clean room, Albertsons allowed the audience data and the conversion events allowed into the clean room. And they were able to measure that sales lift with an incrementality structure as well as look at some of those longer term lifetime value metrics because we were able to, to hold the data in the clean room architecture. So these processes are becoming more and more standardized, more and more regular. They don't just have to be a special case study. They're really starting to happen 365 days a year. And that's like the right now I think what's maturing, one of the things I'm really excited about on the horizon is really this concept of retailers sharing their conversion data, their sales data in allowing the suppliers to measure full funnel their, their, their brand marketing. So maybe what they spend on Metta what they spend on their, their programmatic and really understanding how all of that different upper funnel brand spend along with the retail media networks brand spend then drive sales for that brand in that retailer. So, so really excited about the where we are now and getting, getting the wheels greased. But as well looking at what's around the corner, the safety of the clean room architecture and the transparency is really allowing people to do things I never really thought I would see in my career to be transparent.
- [22:40]
Mike Glasser
Hey Christine, I'll build on that. Just from the brand perspective. I think we're equally excited for a long period. What we'd often find ourselves in is a situation we build our MMM models in house. It is very much the yardstick against which all of our brand media and retail media gets evaluated. It's incredibly important. But we also know that our retail media partners are building their own in house measurement solution and at times we'll come to the room with our best data and our best marketing science folks and it might tell two different stories and to a degree each of us is grading our own homework. And so I think what you're articulating Christine, is perhaps that sort of third solution in a more neutral space where if we all provision our data and look objectively we can get to maybe a more true or more accurate understanding of what's happening. Again with the spirit of we have the same goal. We want to improve the shopping experience, we want to grow the business in the most sort of efficient way. And I think what you're articulating Christine is A where we're going as an industry but B allows us to start to answer those questions 100%.
- [23:47]
Sky Canavis
And in terms of looking ahead or what we're seeing so far with AI, what are some of the top use cases you're seeing there to support or enhance measurement capabilities?
- [23:59]
Christine Gramier
Who wants to jump on that one? I'll go first. I'll go first. So two places that we are really excited about AI helping brands who really suppliers and retailers who've brought their data together. One is particularly for smaller companies like Michael talked about the mid to long tail suppliers. They don't have time to be in there spending a bunch of time, they just need it to work. So two of the ways we've really seen that help one is data normalization, data categorization. Nobody wants to mess around with a different taxonomy from Pinterest than from Meta than from you name it. So normalization in just the data cleaning and data organization is a big spot that we've already made investments and seen it really help. And then two is a fun one just beginning to happen around Signals for Optimization. So if you can really create a good measurement architecture, you can then use the AI to then identify which people you might want to Target on Pinterest vs meta vs I keep using those examples. Amazon, Google, all of the different publishers that are out there. So Signals for Optimization is the other spot we really see AI starting to make some great impacts.
- [25:14]
Marcus
Yeah, I would say like we're not cutting edge on AI. There's too many other things we're starting to be cutting edge on. And when it comes to integrating AI, I think like it goes back to what I said earlier, which is we use AI enabled tools mostly because it's just a faster speed to insights. I think Christine gave a great example about how you can categorize information faster. And we work with a couple measurement partners that they, they use AI and as a way to just help us get to faster data analysis synthesis and then we can make decisions faster. We from a. And that's primarily in our mixed modeling and coming from mixed modeling 10 years ago, where you're looking at like the past six months of spend and there's a lot of channels that it doesn't always get a good read on. The speed iteration we can get now through AI allows us to get model updates a lot more frequently. We can make business decisions faster. We're talking about like spend optimizations a week out now versus spend optimizations that may not be impacted for the next three to four months.
- [26:25]
Mike Glasser
I'll give just quick just to round it out. I think Christine, like you said, we're using it to, you know, comb through myriad data to try and find, you know, patterns and insights that allow us to take what's working and apply it across the business. And then similarly we're finding a lot of great use cases when it comes to, you know, budget forecasting, bid strategies, all that kind of good stuff. So just two very top level ones. There's probably a ton more, but obviously an area that we're investing in, excited about and you know, leaning hard in.
- [26:56]
Sky Canavis
So we have time for one more question. And I think I'd be remiss not to talk a little about or ask about the largest player in retail media space, which is Amazon, which commands the lion's share of US retail media spending around 75%. And that's not something that we see changing much in the near future according to our forecasts. So Mike and Michael, I want to ask a bit about your experience with Amazon and how some of the advantages of working with the platform outweigh the drawbacks.
- [27:27]
Mike Glasser
Yeah, maybe I'll start and then Michael, I'll pass to you. I mean Amazon is a phenomenal partner of ours. It's an important part of our business. Again, it's where the consumer is shopping. So we want to make sure we're there. I think if I take a step back and I just look at the broader retail media space, I think Amazon is very much sort of innovating and setting a lot of the pace for this industry and that's exciting. It's important for us to watch and understand. So not only do they have a massive audience that we can go reach and engage, but they sorts of formats on site, on off site, they've got devices in your living room, they've got great content opportunities. And then I think perhaps most relevant if not important for this conversation is when you look at the power of Amazon marketing cloud and how we can leverage that tool to bring it all together. Our brand media investment, our commerce media investment. Understand how is the consumer shopping, how is she engaging with our products and how can we better deploy our resources? That is the opportunity. And so I would say probably sky that the pros far outweigh any if there are. And I think we often look to Amazon and our strategy strategies on Amazon to say where do we anticipate parts of the retail media industry to go and how can we prepare for that?
- [28:40]
Marcus
Yeah, I don't have too much to add. I think the one, one thing to call out is that, you know, like what Mike said, like the cons, where the consumer shops is where we want to make sure that we are. And it's clear with Amazon scale that's where consumers are shopping. I guess the drawbacks from earlier stage company is that there's less of a browsing behavior on Amazon and people tend to look for very specific things on Amazon. So for a brand that's in the raw, raw dog food space, you know, we, in order for us to break through, we're competing against keywords on Amazon that are, they're gigantic and they're. The entire pet industry is probably bidding on them. And so as an early stage brand it is very hard to break through that. That's one of the biggest drawbacks for an early stage company with Amazon is that it's hard to actually realize any sales that are not from an ad click. And so the cost of doing business for an early stage business is very high. But like I said, balanced with what I said earlier around the scale, and that's where people are shopping. And so there's a ton of consumers out there that you can reach through their tools as well.
- [29:58]
Sky Canavis
Thank you. That's all we have time for. Michael, Mike and Christine, I want to thank you again for this great conversation. I, and all of us in the audience, I think, really appreciated it.