Contact Us
< Back to the Blog

Making Data Passback Work in Retail Media: How Retailers are Demanding More from Platforms

Jonathan Elliott

Michelle Gitter is currently serving as the Beauty Commerce & Measurement Insights CMI Senior Manager at Unilever in New Jersey and brings over six years of experience in media management and analytics. Previously, she was the Performance Media Manager for Northern Europe, based in Rotterdam, where she managed performance media strategies across the region. Her role at Unilever also included responsibilities as Global eCommerce and Data & Analytics Assistant Media Manager, where she significantly contributed to eCommerce initiatives in the EU and North America. Prior to Unilever, she gained valuable experience at Tesla and VICE Media, where she worked on localization, digital content, and cross-market campaign management.


  • Retail media is a very old advertising concept that has found a dramatic upsurge in its digital incarnation
  • The pandemic accelerated its growth and, after sales, the chief attraction for retailers in FMCG/CPG is data
  • Valuing that data is now a central discussion between retailers and platforms
  • Cookieless could well distribute power away from retail platforms to media houses
  • The interaction of data passback, retail media, and the cookieless world is hard to predict, but it might be a bumpy one for a few years!

Jonathan Elliott: Retail Media has surged in the last two years. What do you think was driving it?

Michelle Gitter: It's a mixture of things. Of course, during COVID, as a retailer, like everyone, we saw people shopping online more. Retailers then had to compete and evolve and make their apps and web pages a lot stronger. And that also drove traffic. So, by having this traffic, they were also able to connect with a new set of people who wouldn't usually look at their ads, which led to a recognition of the amount of investment, the number of advertising opportunities available, and retailers seeing the value in it. So, retailers saw that these high margins could come their way. That accelerated the process.

JE: Let's talk about data passback about data generally. And then what datapassback is.

MG: It's important to distinguish between audiences, targeting options, and data passback, the data we get from the retailer that helps us drive campaigns better, especially in the future.

As we continue to optimize, the idea of being able to hyper-target a person based on their previous shopping behaviors is compelling, and it is not just through the retailer site but the entire World Wide Web. So, it doesn't have to restrict itself to just the specific retailer website. Still, we also need to tread with extreme caution about where we draw the line between helping somebody buy something better versus being intrusive and being too personal.

Although somebody clicked 'yes' on a cookie page, maybe they did it just because they could access the web page and not because they're willing to give their data. We need to tread very, very carefully there. And, as Unilever, we need to be example-setters in ensuring that whatever data we use is very clean and thoroughly consented. And we do that with a lot of care.

So, when it comes to audiences, it's very powerful - you can target people and not waste a lot of precious inventory. What used to be just extreme mass communications, like on the radio, became more targeted when we could – for example, on TV – select only to have the ad on at a specific time of night or after a particular TV program. But now we also have this added data layer, where we know this is what people purchased. This is not just where people are in their viewing habits or what content people watch and consume – whether they like 'Love Island' versus 'David Attenborough.'

For the longest time, as long as we've been working with any media partner, like with TV, we got back GRPs - gross rating points. Eventually, with social media partners, that became - how many impressions, how many clicks did this generate?

We still want to know how many impressions we got, how many clicks we had, and how much time we spent on the page we got - on the product detail page, for example.

But now we also want to understand how many sales it drove and what our return on advertising spend is. How many of these customers have not purchased our brand before? It is a layer beyond what we get from our social media partners.  

Those metrics are the only justification the retailers can give: "We have better inventory, so you should pay a premium with us." Because we do pay a premium on a CPC (cost per click) or CPM (cost per mille/thousand), so we do spend more on our retailer than we do on a social media site. We pay more than we do for any other display advertising online. To understand how they justify it, we need to understand two things.

1. It's where people are already with a shopping mindset. So if people are on their web page, if people are on their apps, and people are in their environments, they clearly want to buy something. If I'm on Amazon, I'm likely willing to purchase something - I'm not just browsing. I know that many people say, "I kind of just browse on Amazon," and things like that, but most people who visit Amazon are thinking of, at some point, shopping there, and it is one or two clicks away from a potential sale.

That's one way to justify the premium price. But it is not the only one.

2. It's the data that you can get back. And without knowing what data I get back, for example, how many sales I drove, how many new customers I was able to get through my advertising, I don't even know if the first part - the being-in-the-shopper-mindset part - is actually that much better than just having the regular old advertising platforms like TV, radio or magazines.

So that's why it's essential to have this data passback. And then connect it also to the three critical elements in assessing data – is it optimized, measurable, attributable - so that we know the answers to questions like "How do we make things better? How do I continue optimizing my campaigns? How do I continue making sure that this action happened because of this ad? And how do I create my media mix? Do I just want to have search and sponsored products? Or do I want to go beyond that? And then: do I want to add a layer of Amazon live, so live stream commerce, and then having a video being displayed because it's a new product and there is a new application of it and so, just seeing an image of it is not enough so we should have some more video - kind of advertorial advertisement." Data influences all these decisions.

That's why data passback is so essential because we need to understand how we can continue improving and how we can continue understanding this environment and the value that it brings.

JE: You must be thinking ahead about where data passback is going. What are the real hot buttons?

MG: There's one trend that is connected to Multi-Touch Attribution (MTA) – as we acquire more data, we'll make it a lot more accessible - and Marketing Mix Modelling (MMM) as well. These tools will help us gauge how we should be investing across our full entire media ecosystems, also including more traditional channels, more established digital channels, and then retail media, and having this information and having a closer grasp on retail media data allows us to be able to say this is where it plays a role within the entire ecosystem.

In the future, there will be a fourth wave, a fifth wave of advertising. Maybe it's going to take a little bit of time to get to this fourth or fifth wave. But we're already seeing that this is the third wave of marketing. And we need to jump on it; otherwise, we'll just be left behind. And the safest, smartest way to do it is by getting this data passback and then running studies to understand what works well, and what doesn't. It's by understanding how it plays a role within our full plan and within our media mix that we'll make the most of its future potential.

And then it is taking advantage of all of this put together and making really, really strong plans through partnerships with retailers, just like we have with existing social sites, like Meta and Google, just like we do with existing TV channels and traditional media providers.

So it's through a partnership with this kind of next-generation of media houses that we'll be confident and comfortable having this kind of complete plan with retail channels.

JE: That is an opportune moment to talk about 2024 and the cookieless future. How do you think that will impact retail media and data passback?

MG: Cookieless is looming, and we're all very conscious of it. With Cookieless, the off-retail partners, the Meta so Facebook, Instagram, the Googles – (Google or YouTube), Twitter or X, are now really struggling because they cannot tell us: "It is because of this ad on our websites, or on our apps that this happened because we cannot track this connection." So that puts a barrier between off-retail and on-retail.

This also empowers the retailers to say, "We have the strongest first-party data," or for them to choose and say, "These are the external partners that we wish to connect with."

And this goes beyond social sites like Google and Metas. For example, here in the Netherlands, we have a media platform called DPG, which has many publications. So they have the weather apps, they have a rain app, they have news sites, they have female-skewed magazines, things like that, that can, from there, be taken to a retail platform.

They even have a microsite, which is a mixture of all the different kinds of supermarket discount magazines together in one app owned by this media house. That then brings it all together, and you can advertise there.

So now, the retailers can work with these media powerhouses or media conglomerates and select whomever they want. Then, they could sell their media and act as a media agency. And that's when it gets slightly dangerous because we already have a media agency; we don't need retailers to sell us media. But they are doing it because they have the cookies data, the cookieless world keeps from us. So, I don't think the world is making it easier for us, and although we never expected that, I also don't believe that it's that much kinder to the consumers whom cookies were supposed to help.

In a way, it is a transfer of power. But then it also allows us to be more cautious and conscious of how we do everything. But that's where it gets a bit more complicated, more convoluted, and that, especially in the first two or three years, will just be a bit of a mess. The good thing is that we've been preparing for it for about five years now, and now we're finally at the stage where it's like, "No, it's going to happen." But that's where it will take a lot of work to navigate and figure out what's right and what isn't in the first few years. Cookieless will sort itself out, but it will be a bumpy ride getting there.

No items found.
Brand Masters Talks
No items found.

Step into the future of online ads with us!

Stay up-to-date on the latest industry insights and trends by subscribing to our blog post email notifications.

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.