Stay up to date on the digital shelf.



    We'll keep you up to date!


    How to Access and Measure Retail Media Networks in the Wild West Era, with Andreas Reiffen, CEO & Co-Founder, Crealytics

    More and more retailers are bringing their retail media networks online in 2024, which is both a management and measurement challenge for brands, but also a significant potential early advantage for those who figure out how to manage it.  Andreas Reiffen, CEO & Co-Founder of retail media supply side platform provider Crealytics, joined the podcast to share how brands can tap into solutions and providers that can ease access and management, and brings insights into how, as ROAS dies, more blunt but more accurate forms of measurement can be deployed. 


    Our transcripts are generated by AI. Please excuse any typos and if you have any specific questions please email info@digitalshelfinstitute.org.

    Welcome to Unpacking the Digital Shelf, where we explore brand manufacturing in the digital age.

    Hey, everyone, Peter Crosby here from the Digital Shelf Institute.

    More and more retailers are bringing their retail media networks online in 2024, which is both a management and measurement challenge for brands, but it's also a significant potential early advantage for those who figure out how to manage it.

    Andreas Reiffen, CEO and co-founder of retail media supply side platform provider, Crealytics, joined Lauren Livak-Gilbert and me to share how brands can tap into solutions and providers that can ease access and management, and brings insights into how, as ROAS dies, more blunt but more accurate forms of measurement can be deployed.

    Andreas, welcome to the DSI podcast.

    We are so excited to have you here.

    Retail media is always an interesting topic, so we're delighted to have you on.

    Thanks so much for having me.

    Hey, Peter.

    Hi, Lauren.

    So that whole space continues to grow rapidly, and it seems like there's always new ways to buy media, new partners to maybe work with, and then a lot of different ways to measure success, and a lot of competition that comes out of that.

    And we know Amazon owns around 75% of retail media spend, but there are hundreds of other retailers are requesting retail media investment.

    And I don't know, to me, it kind of feels a bit like the Wild West, as we would say in the States.

    But there's a lot of money at stake, a lot of investment, and a lot of traffic to web pages that drive consumers to buy.

    So it's a big deal.

    I know from your perch at Crealytics, you've thought a lot about how some of this madness can be tamed to guide retail media buyers towards more productive investments.

    So we'd love for you to just lay out some of those thoughts you've been having from your view there.


    So what you mentioned is that we have 75% that is owned by Amazon.

    25% is shared among a multitude, dozens of different retail media networks.

    From a brand perspective now, in order to target those 25%, you have to go through so many different walled gardens.

    And one point you already mentioned is standardization of measurement.

    That's key to success because if you have different tracking and attribution windows, different tracking mechanisms, those numbers are just not comparable at all.

    So standardization on this side really makes sense.

    A second thing is we need to make it easier for brands to target all those retail media networks.

    And one way is to democratize the access to the retailer-side inventory.

    And we need a structure in the market that is more similar and more comparable to what we see in the broader open web display space, where we have a separation between DSPs and SSPs.

    So what we want is that as a brand, you could put your money into one DSP, let's say Citrus ad, or could be a PackView or Flywheel or whatsoever.

    And from there, you can target all different retail media networks without having to go through Citrus to target inventory on retailer A and through Gridio to target retailer B and like a multitude of different tools, which makes it even more complex.

    And that's essentially one of those pieces we have developed at Crealytics, which is a supply side platform which allows any DSP out there to plug into our API and show ads on the retailer side that then get optimized on the yield side on those retailer websites.

    So then that's mutually beneficial for the retailer and the brand, right?

    Because the retailer has to work with one platform and then the brand has to work with one platform and they're getting access to every retailer that they're interested in.

    Do I have that right?


    So it simulates a little bit what we see on within the Amazon ecosystem.

    Amazon owns the supply side piece, which we now offer to retailers so that they can do the ad serving, run an auction, optimize the website with banners and product ads, essentially.

    But then they can also plug in any DSP.

    They could work with a multitude of different partners, which means it's easier to get demand in.

    I don't think that this game will go fully back into a network mode.

    There will always be this one on one selling.

    But we also see that most retailers, they really struggle because they are simply too small to access national agency budgets.

    So agencies don't want to talk to dozens of smaller retailers.

    They would like to put their money into some DSP, maybe even something homemade, and target a multitude of different retailers.

    And this is like private markets, one on one selling and back filled with something that is network demand.

    But today, that's not really an efficient functional ecosystem.

    It's like there's a lot of sand in the engine that needs to be taken out.

    And there's a reason why, while I think Amazon has a US market share of 38% roundabout, but they own 75% in retail media.

    We have to ask ourselves, where is this huge gap coming from?

    And if you dissect the Amazon ad revenues, you will see that a 90% is coming from on site monetization.

    So clearly, with all that on site inventory, something is wrong.

    So it's a very informed claim that this isn't going as well as it should, but it's something that can be solved going forward if we make a few tweaks to the different actors in the ecosystem.

    And one thing we definitely know is that the only thing we can be confident of is that this market will continue to change at an incredibly rapid pace.

    I mean, we're just a couple of days as we record this from a Walmart buying Vizio.

    And that's essentially, I mean, from my read of it, I'm probably oversimplifying it, but essentially gets them a huge ad network to add to their.

    So, you know, that's $2.3 billion.

    So that's telling you what the opportunities are here.

    And also, this game is going to continue to change.

    And I would imagine the more open you can get this and the more the more expertise can be gathered by these DSPs to be able to make sure that those products get to market to their customers as quickly as possible through this open approach.

    I would imagine that that's part of the value.


    If you I think, you know, it's going to be a bit of a more controlled approach compared to the open web because we are dealing mostly with endemic brands.

    This won't be opened up for all brands that will now show inventory on on the retailers websites.

    But for any endemic brand that is selling through those retailers, there can be a much, much easier access than what you have today.

    Another limiting factor, one of the reasons why retailers beyond Amazon haven't managed to really catch up is, and it's very simple, if you compare how much ad inventory Amazon has opened up versus other retailers, you will see a huge gap.

    On a quarterly basis, we release a sponsored products report, and we see that every fifth product on Amazon, so 20%, is already sponsored.

    If you compare this to many others, you will notice that it's a fraction of the aggressiveness we see on the Amazon side.

    So why is that?

    And it's interesting, I just got fresh numbers in, and we compared for one retailer, we analyzed the performance of sponsored products versus organic products, and the click-through rate of the sponsored products is round about one-third compared to the organic products.

    Now a product that has a click-through rate that is only one-third also converts at a lower rate.

    So it's inventory, which you could fill with natural organic products that would sell through.

    So if that stuff doesn't really isn't fully optimized, you just distract people from what they are supposed to do in a retail environment, which is find products fast and buy them.

    So one thing in one solution to the problem is somehow you have to leverage AI and personalization to make sponsored equally good as organic.

    And once you get there, then you can open this up.

    If you have this performance gap, relevancy gap, you have to be very cautious.

    And I feel there are quite a few that just don't make any money through sponsored products, because what they bring in gets at the same time destroyed on the retail side by dragging down the conversion rate.

    And that's a key prerequisite to actually be able to scale up and get to where is Amazon today?

    Around about 9% of their Ecom revenue comes in through sponsored products.

    If you take a look at the broader retail market, that's maybe 1%, half a percent.

    So there's a lot of potential still in there.

    I feel like a lot of the challenge between the organic and paid dynamic is that they sit on separate teams.

    So like the retail media team is focusing on paid ads, and then the organic search is usually like the digital e-commerce team, and they're not connected.

    And so they didn't know they were running a sponsored ad.

    At the same time, they were trying a new search strategy.

    Do you find that when you're working with the retailers and brands?


    I think it's in the first place a huge organizational problem.

    And it's corporate politics that are the biggest obstacle in this game.

    And we are seeing this almost everywhere.

    It's the advertising retail media.

    They, of course, ask for more inventory.

    But the other side is usually incentivized by something like website conversion rate or conversion rate of the search results.

    And they see now, well, your sponsor products have a CTR click-through rate that is only one third.

    So people don't like them.

    Now you ask me to give you even more space for something that isn't working.

    And so the problem is, how do you make this decision?

    But there is a very simple formula and a very straightforward way on how you can tackle this problem.

    Because ultimately, every business is about numbers.

    And in this case, you want to bring margin in.

    There's so much pressure today in retail.

    So retail margin is very thin.

    Ad margin is, let's say, 80%.

    Retail margin is put on the 10-15% side.

    Now, as a retailer, the game you now play is you get two currencies or two revenue streams.

    One is from retail, your 10-15% margin you get on that revenue, and the second one comes from advertising.

    So you can simply make changes, test things, and run A-B tests.

    What you measure is retail margin plus ad margin.

    How does this become better?

    And the simple way is, well, both increase, then it's a clear, obvious recommendation, just make the change.

    But it could be that your retail margin suffers a little bit because relevancy goes down, but you overcompensate this through additional ad revenues that come in.

    And then this should be a conscious decision to just move forward.

    But it's just now that this more, let's say, holistic approach is being seen in the market, and we see some of the strategy consultancies approaching retailers and recommending them such an approach.

    So this siloed thinking doesn't help retail media to grow, but it's also dangerous because if you do too much of retail media while it's unoptimized, you cause more damage than you do good.

    So you have to see this in conjunction with each other.

    So Andreas, what you've been talking about here is really kind of the retailer side of what is the balance of ad opportunities, etc.

    that make for the best mix for my overall business, right?

    Let's look at it from the brand side for a moment, who are struggling to know that what they're investing in is paying off.

    And part of that is because it seems like every retailer network has a different kind of measurements that they share, how reliable are they, etc.

    And so the IAB and MRC are organizations that have been trying to work on a more standardized set of measurements across platforms.

    Love your sort of update on that, your point of view on that.

    But as this evolves, what is your advice to brands as they try to navigate how to make sure they're spending their money wisely in this Wild West environment?

    So first of all, the IAB has done an important piece of work.

    We said it's several dozens of retail media networks.

    And imagine everyone has a different way to attribute sales.

    Everyone has a different cookie duration.

    Everyone does something different, whether it's a one day post view and 14 days post click or extend this, compress it.

    It's very hard.

    Then some de-duplicate things, others don't do it.

    It's impossible to compare your performance on retailer A versus B versus C.

    And with a certain level of standardization, you make this, of course, a little bit more comparable.

    What this does not solve for is it's still gonna be wrong, and it's still gonna be fundamentally wrong.

    It's more consistent, but it does not mean that this will be accurate.


    For the simple fact that this is what I would call post-exposure measurement.

    People click on an ad, and they purchase.

    And we assume that the purchase happened because somebody clicked.

    But this is not true.

    And I can give you a simple example.

    We have in many instances today some post-view mechanism.

    This means if you go to some retailer and you search for something, a product shows up somewhere on the page.

    You are not even interested in this product.

    And then you buy accidentally from that same brand some other product.

    This is credited to the ad because it's the same brand, and it generated a view.

    So in this case, the sale would have happened without you buying this.

    So the whole ad industry also with Meta and Google was designed that we show ads to people who are predisposed to buy.

    So we spend a lot of money in areas where people would buy anyway.

    And this is something that gets completely disrupted at the moment outside retail media for the simple reason that all the privacy limitations have killed granular data.

    And we are losing more and more data points, so strategies like incrementality testing or media mix modeling are on the rise because they are privacy proof.

    You don't need the granular data, and it's also a more causal instrument of measuring success.

    On the retail media side, to your question, what can be done?

    The bigger brands, the more sophisticated brands, they have built measurement capabilities in-house.

    And I think you need to own this.

    No marketer, whether it's outside retail media or within retail media, should take numbers for granted.

    The rules of the game will always be that a publisher is selling ROI, and the advertiser who has to pay the bills needs to double check this and get a sense for where he or she stands.

    And so means generally run experiments and find out what you truly get.

    That's the advice for the more sophisticated players.

    And do you find to what we were talking about in the beginning of the podcast about brands being able to work with multiple different retailers, that the test and that enables a test and learn approach where they can try something, one type of ad on a specific retailer, see if it works and then pivot and then ask for the data back and understand what the consumer is looking for for that specific ad.

    Is that what you would encourage brands to do, especially as they're trying to prioritize where they spend their money?

    It's going to be very, very hard to do this today because those retail media networks are by far not as sophisticated as Google or Metta, where you can easily run or turn off certain geolocations.

    You can experiment much more easily.

    But what you can still always do is scale up, scale down, and observe what happens with your overall retail revenues.

    I mean, if you see 10X in sales on your ad campaigns, if you scale up, but you see no lift on the retail side, you should...

    you have an informed doubt that this might not work out as well.

    And so from the...

    oh, go ahead, Peter.

    No, I was just going to say the law of informed doubt.

    That's what we're stuck with at the moment, is that what we're saying?

    That's rough.

    But I know it's where we're at right now.

    It is early days for all of this, and building these systems and these processes are not easy.

    Sorry, Lauren, that was just my brain going, wow, that's rough.

    It's a great point.

    And I think, Andreas, you make a valid argument where the brands need to ask questions.

    They need to ask where the data is coming from, what it is, what they're being given from the retailer, and also look at the holistic picture, which I think is the key there.

    What is that omni-channel strategy of what your sales look like versus your retail media ad spend?

    So I love that approach.

    And staying on the brand theme.

    So in terms of what brands should be thinking about with all these changes in retail media, especially because we know the cookie is finally, I'm air quoting, going away in 2024, we think.

    So what should brands be doing to prepare themselves?

    How should they be thinking about these changes this year?

    Yeah, I think for those brands that also have a DTC presence and advertise for their DTC business on Google, on Meta, on the surface, this looks like bad news because things get more difficult now and you can't track as granularly as before.

    The reality is it's very good news because we used poor methodology simply because it was convenient and it delivered nice-looking results that nobody wanted to really question.

    Sometimes for those who tried, it was even hard because if you wanted to turn off retargeting campaigns that rarely ever delivered reasonable results, then somebody said, oh, but now I'm missing out on my 5 million in sales that were never actually there.

    So it's much easier now to have a proper testing approach, incrementality testing, for example, on prospecting for Facebook ads looks already better than what you can track.

    So I think the world got somewhat better through those privacy restrictions.

    I think we can spend our money more wisely.

    In retail media, as mentioned, I think my number one advice here is it feels oftentimes that it's a new discipline for many brands.

    They have very little expertise.

    Now they have some flavor of other brands that were heavy in DTC advertising.

    And many of those brands work with either agencies that bring the skills or they have built in-house teams that can actually deal with a level of complexity.

    If you have just one rep who sells to the retailer and puts money into some retail media network, you probably, I mean, if this is a considerable amount of money, you have quite little control.

    So on the brand side, you probably need to build expertise.

    And I do believe, especially on the measurement side, if you spend, I don't know, like a few million in retail media, you can either pay a tax from those retail media networks because you're being sold an amazing ROI, or you pay a tax for some measurement capability you need to build up.

    So, Andreas, it does sound, you know, the cookie being dead and all that, like the ROAS numbers are just not reliable any longer, and they're way overstating the impact that brands are getting from that.

    And that's really dangerous when you're making millions of dollars worth of decisions.

    So in a retail media network context for brands, what kinds of metrics should they be pushing for?

    What can they rely on in this time between the beginnings of this thing and when it will actually mature to the next brand?

    What is your advice there?

    Outside retail media, in the meta ecosystem or Google ecosystem, you can run geo-based incrementality tests.

    You just stop advertising in some geo locations, and you keep advertising in others, and then you compare the revenue differential, which gives you an indication of to what extent do your ads actually really impact your overall sales.

    In retail media, this isn't yet feasible because most of the retail media networks don't bring those capabilities.

    So what can you now do if you're a brand and you want to perform better than your competitors?

    So it can be as simple as either option one, scale up your ad spend, and see if your numbers, your overall sales numbers with the same retailer are also scaling up.

    So do you see an impact on your overall sales or you temporarily bring your spend down and you see, do you lose sales?

    And if you do this once in a while, you at least directionally understand the value of what you get.

    And if you then see that certain things you do do have an impact, you can continue investing in those specific activities.

    Yeah, like so often in these early era of things, all you have are blunt instruments.

    And so it seems like we're in the era of blunt instruments for finding your way to data that will help you make the right decisions.

    And maybe that's just, not maybe, it seems like that's the reality that retail media investors, investors at brands need to make.

    Does that sort of capture the essence of what you were saying?

    Yeah, I would agree with this.

    So folks, Andreas, first of all, I just want to thank you so much for coming on here and really laying out kind of the stark realities and present opportunities and some of the ways that are happening in the industry of making it more possible to be out there in the retail media network without the burden being on brands to try and figure out each one of these opportunities themselves.

    That's a super important advancement in itself, I think, particularly as a lot of these are coming online.

    And then the second is we're just going to have to kind of bear with the way the Wild West is working right now in terms of measurement and make sure you're not relying on bad old metrics, but just trying to push for the metrics you can have that are actually giving you the signals you need.

    It's actually also an opportunity if you think about it from a different angle, because if everything is perfect, everyone will be equally good.

    In this more difficult environment, there is an opportunity to set yourself apart if you just play this more complex game better.

    So in Google Ads, you just today almost press a button and you let a machine optimize everything.

    So everyone is kind of on the same level, and there's a very high level of efficiency in that game.

    Here you have just much more leeway, more possibilities to set yourself apart.

    I love that as a golden lining at the end of this podcast.

    Andreas, thank you so much for joining us and filling our listeners in on these opportunities.

    We appreciate it.

    Thanks so much for having me.

    Thanks again to Andreas for the sage advice and the golden lining.

    In these exciting times, you must gather with us at the Digital Shelf Summit in Nashville in April.

    So much to share and learn.

    Digital Shelf summit.com for more info.

    Thanks for being part of our community.