Subscribe
Share
Share
Embed
In this episode of Marketing Un:Learned, Ian Jindal welcomes back Florian Clemens, who leads strategy, proposition and measurement for Tesco Media and Network. Following his appearance on RetailCraft, Florian returns to unpack four core assumptions that retail media practitioners need to challenge—and to share what he has learned by testing, experimenting, and building one of the UK's most sophisticated omnichannel retail media platforms on top of Tesco's 24 million Clubcard customer base.
About the Guest
Florian Clemens leads strategy, proposition, and measurement at Tesco Media and Network, where he drives a multi-year growth strategy, competitive positioning for the brand, and performance budgets, as well as client-facing measurement frameworks. He has been working in retail media since 2014, when he joined Amazon Ads and built the global accounts team. Before entering retail media, he served as a brand manager at P&G and Danone, providing him with a unique perspective from both sides of the purchase order.
Episode Outline & Key Topics
Does advertising make the retail experience worse?[03:35]
Florian challenges the assumption that advertising is an intrusion that shoppers must endure. He introduces Tesco's "win-win-win" framework—creating value for the shopper, the advertiser and the retailer simultaneously. Examples include Cadbury Christmas grottoes outside stores (physical, experiential activations that delight families while delivering extended brand engagement), Moretti sponsoring Italian recipes on Tesco Real Food (where surveys showed improved shopper experience and brand equity lift), and research showing two-thirds of shoppers find Disney+ ads at the snacking aisle completely relevant and normal.
Is advertising something that shoppers passively suffer?[13:52]
Moving beyond the idea that advertising is a cognitive overhead, Florian explains Tesco's Clubcard Challenges—a gamified loyalty programme where shoppers select up to 10 brands they've bought before, set personalised spend milestones over six weeks, and earn escalating Clubcard points. Shoppers are in control, advertisers only pay when customers hit spend thresholds, and the mechanic blends performance, loyalty and engagement in a transparent, shopper-first way.
Is relevance the most important factor?[21:35]
Florian's counterintuitive argument: "100% relevance equals 0% incrementality." While search results must be relevant (someone searching for Gillette razors expects Gillette results), pure relevance risks becoming a self-fulfilling loop with no room for inspiration, discovery or brand building. Tesco has introduced "conquesting" (bidding on competitor brand terms with visually separate banner ads), audience targeting based on lifetime behaviour rather than in-session intent, and research into when and where to inject inspiration without degrading experience—balancing relevance with incrementality on both an X and Y axis.
Is it all about monetisation?[31:16]
Florian introduces "Smart Stock Audiences"—a Unilever co-test where Tesco suppresses ads for categories shoppers have just purchased (e.g., if you bought mayonnaise last weekend, you won't see Hellmann's ads this week). This prioritises shopper experience and advertiser efficiency over maximum ad serving, using Dunnhumby's 30 years of predictive data science to create audiences around high-propensity trialists, future decliners and recently stocked households—all processed in aggregate to remain 100% privacy-safe.
Key Insights & Quotes
"If someone searches for Gillette shaving blades and gets an ad for Gillette shaving blades and buys the Gillette shaving blades, what have we done here? It's amazing ROAS, but not really ROI."
"Is relevance and inspiration not the opposing sides of a continuum, but actually an x-axis and a y-axis?"
"There might be some times when it's better not to serve an ad… if they've just bought a big jar of mayonnaise this weekend, this might not be the time to advertise more mayonnaise."
"It's called relevance, not relevancy. I don't know how some people keep calling it that."
Florian and Ian also explore the strategic planning process behind Tesco's omnichannel campaigns, how the team pitches for advertiser budgets alongside major UK media platforms, the role of A/B testing and shopper surveys in balancing monetisation with experience, and how Clubcard's scale and Dunnhumby's data science enable sophisticated audience strategies that work in aggregate without compromising individual privacy.
Resources & Links
- Tesco Media and Network - https://www.dunnhumby.com/tesco-media-insight-platform/
- RetailCraft podcast - Florian's previous episode
- Florian on Linkedin - https://www.linkedin.com/in/florianclemens/
- Ian Jindal on Linkedin - https://www.linkedin.com/in/ianjindal
- Epsilon on Linkedin - https://www.linkedin.com/company/epsilon/
————————
About Marketing Un:Learned
Marketing Un:Learned explores the challenges that leading-edge digital marketing poses to established and received wisdom. All new initiatives face scrutiny—the "what-abouts," statements of inertia, and deprioritising questions. In this series, we take those challenges head-on and learn how exemplars deliver persuasively, perhaps changing our thinking along the way. In partnership with Epsilon, we focus on innovation in retail media, digital advertising, CRM and personalisation, speaking with expert practitioners who have moved beyond optimised, well-known processes.
Creators and Guests
What is Marketing UnLearned?
Marketing UnLearned explores the challenges the leading-edge digital marketing poses to established and received wisdom.
All new initiatives, until proven, are subject to scrutiny and challenge: the ‘waddabouts’, the statements of inertia, the “why bother?”, the deprioritising questions. Within these challenges there is often a grain of truth, but in this series we’ll take the challenges head on and learn how the exemplars deliver persuasively - perhaps changing our thinking along the way.
In partnership with Epsilon our first series will focus on innovation in the areas of retail media, digital advertising, CRM, and personalisation. We’ll speak with 10 expert practitioners who have moved beyond the optimised and well-know digital marketing processes. More than a simple ‘always sunny at 30,000ft’ case study, we’ll put the challenges to our guests and hear how they were overcome, how their thinking developed and learn about the ‘new state of the art’. While we may UnLearn some pieces of accepted wisdom, we’ll replace them with new, effective learning. Everyone wins with Marketing UnLearned.
Ian Jindal (00:01.496)
hello and welcome back to the studio. We're in the marketing unlearned studio today and we're having a well. Let me just say it's the podcast Martin learned is our series where we explore the challenges at leading edge digital marketing and how things that we think are received wisdom are open to challenge. So what maxims need to be revisited?
What certainties do we need to reconsider? And of course, having unlearned something, we can't let you go without sharing maybe the improved, better and more relevant approach, the relearned marketing as it were. So in the studio today, we have a guest who was recently on the Retailcraft podcast and made such a splash with his zingers and quotes, especially his now famous win, win, win formulation.
that we asked him back today just to focus on the marketing bits and the lessons he's unlearned and then relearned with us. So it's a pleasure to welcome Florian back to the studio. How are you Florian? No, not at all. I think this counts as by popular request, but just in case there's anybody tuning in today who hasn't already heard your retail craft.
Florian (01:11.782)
Really well Ian, thank you for having me.
Ian Jindal (01:25.824)
attendance to a visual craft session. Why don't you tell everyone who you are, a little bit about your role and the background and then we'll roll into it.
Florian (01:35.274)
Absolutely. So I'm Florian. I lead strategy, proposition and measurement for Tesco Media. So my role is to lead our multi-year growth strategy, but then also drive our propositions. How do we compete for brand budgets? How do we compete for performance budgets? And then the client facing measurement, do clients actually see an ROI of their spend with us? So it's a really interesting chunk of what we are trying to figure out. And I've...
I probably have this role because I've been working in retail media since 2014 when I joined Amazon Ads. I built the global accounts team there. And then 100 years ago, I was a brand manager at P &G and at Danone.
Ian Jindal (02:16.02)
Right, so you've you've basically sat either side of the purchase order for some of Europe's biggest and best marketing brands. So we're particularly keen to know your thoughts. Now, in typical response, when I said what have you learned and unlearned, we came up with four things. So it's a bumper edition today that we're going to race through. So the four topics we're to look at.
The idea that advertising makes the retail experience worse, does it? Advertising is something that is thrown upon the shopper that she has to suffer, does she? We talk about relevance. We all say relevance is most important, but is it? And then lastly, it's all about making money, well, is it? So I mean, four massive topics. Let's dive into them one by one. So one of the things you've sort of unlearned or challenged
is that advertising makes the retail experience worse. Advertising is there to get in the way of our pleasure, happiness and efficiency as shoppers. tell us a bit more about what advertising in that case could add to life and how you came to this view.
Florian (03:35.551)
Yes, you already mentioned that I keep going on about the win-win-win so that advertising is actually a positive experience for the shopper, that it delivers for the advertiser, but also that it is positive for the retailer.
And as we are now building retail media platforms on top of the retail business, also with the retail side of the house, there's always the worry that putting more advertising into the store or into the app is going to make this experience worse. And part of that is true. For example, there is always a competition for space. And if you put in more advertising placements at the top than everything else moves around, that is 100 % the case.
But what we are increasingly seeing is that there are ways to actually identify that win-win-win. So for example now for Christmas, have catberry Christmas grottoes. we have, that are beautiful, that are just outside the stores.
Ian Jindal (04:37.324)
So how on which platform would I see this and what what I experienced as a customer. It's a physical grotto.
Florian (04:40.712)
You would see this the stores, you would see this physically at stores. It's a physical Cadbury Christmas Grottoes and the interesting bit with that is that it is a really nice Christmassy experience for the shoppers.
And then for the advertiser, you have people spending minutes and minutes in an extremely branded experience that is driving what they now want to communicate for Christmas. And then for us as the retailer, of course, we are giving up some space, but in the end, it is a very positive thing. It improves all of the shopping experience for the trip. But also, of course, the advertiser is
paying for that and it's not something that that we would have to stand up for ourselves. So that's one of these areas where we can say yes this is truly a win-win-win.
Ian Jindal (05:28.451)
Mm.
Florian (05:36.243)
Now the interesting bit with this is that we can also see this on other properties, for example, digital, where we have our real food recipe platform, Tesco Real Food, and we had Moretti sponsor all of the Italian recipes on there. And it's a similar situation where we have a very relevant advertiser there who wants to, of course, set themselves up as an Italian beer and want to drive that
connection. also, were then because we were doing this for the first time, we were actually doing surveys of the visitors there if they would if they found that this was an appropriate experience. And we were actually really positively surprised that not only were we able to show to the advertiser that all of their brand equity metrics were going up, so associations with that this is a true Italian beer, but also then people telling us as the retailer that this is a positive
experience on the real food on the real food website and that we were actually making the experience of that website better So it's a really nice space for us to figure these things out We just had at our upfronts. We released some research Data where we are now looking into how much further you can go with this So for example this the survey there was that we were asking people in a snack aisle
Ian Jindal (06:45.902)
Hmm
Florian (07:05.738)
Would you find it odd to see advertising for Disney Plus? Just as a, you know, one of the streaming platforms. If you are at a snacking aisle, would you find it strange that a streaming platform is advertising there? And we found that two thirds of our shoppers were telling us that they actually would find this completely normal because it is such a snacking location and there absolutely is a link and there absolutely is a relevance there.
Ian Jindal (07:13.695)
interesting.
Ian Jindal (07:31.438)
Mmm.
Florian (07:35.702)
So it is for us just an interesting space where we think we can find something where advertising is not making the experience worse, but where there is a win-win-win.
Ian Jindal (07:46.83)
So I'm liking this and I'm totally persuaded and then I get this little voice that says, actually how do I manage and broker that? So for example, let's look at Disney or Netflix or any streamer plus a snack. So in the old days, you might have had a co-branded pack.
you know, the Disney plus Doritos pack, or you might have had a buy 20 bags of Doritos, get a month free or some form of offer on shelf, et cetera, et cetera. So the tie up is clear, but I'm interested that you've covered a number of formats here now. So we use the shelf edge, we're to the on shelf promotion, the buy ones, the packaging, but you've mentioned to me in...
like a couple of sentences. What I'm imagining is an amazing grotto in Cadbury's purple with Santa at the end and smell of chocolate. And on the other hand, I've got a recipe site, you know, which is a living, breathing view of Italy. How do you work out when someone says, hey Florian, I've got a big pile of money I want to dump on you for a win, win, win. How do you work out what type of activation?
they should do because not everybody can have a grotto and not everybody can have a recipe site. So how do you work out who goes where so that there is this win-win-win?
Florian (09:17.768)
So it definitely starts with relevance. And I know that we want to talk about that in a bit. But it is also a question of what is the unique experience and the unique overlap that we can create here? Because also, these are budgets that we are pitching for.
And we need to be able to explain to the advertiser or possibly the media agency why this is a better relationship, why this is a better connection than something else that they could be doing with that money.
Ian Jindal (09:49.443)
Hmm.
Florian (09:49.595)
And we find that when we then talk about really customer insights, we start with what are actually your business opportunities? What are the audiences that you can reach here? And what, how can you move the needle with these audiences? That that is for us really the best way in that we start with a clear brief, a clear opportunity, and then start thinking about what are actually the surfaces. So we've done this also, if you want to stay in the non
Ian Jindal (10:17.134)
Hmm.
Florian (10:19.468)
pandemic part of things that we don't actually sell in store. example, with Vauxhall, we've done last year, we did a really big collaboration, which also then involved the ClubCup points that people were receiving, charging points outside of the stores that are then Vauxhall branded and where they were getting then specific discounts. So it is about how do you create a much more omni-channel
Ian Jindal (10:43.502)
Hmm.
Florian (10:49.529)
experience, where then a part of that is, are all of these individual puzzle pieces.
Ian Jindal (10:56.046)
Yeah. And are you the one coming up with this? Because I was interested in what you said about how it's competitive in your pitching. So this is a different view from the outside of the relationship. So I sort of imagine that everyone knocked on your door, asked if they could give you money for an ROI.
and it was all kind of, you know, automated. Whereas what you're describing is a much more creative coming up with concepts, pitching them. It's much more of a developmental approach. Is that a fair distinction or would you say that's actually true and that you do have more of a creative, you know, imagining role?
Florian (11:41.181)
It is both. So we have a strategic planning team that works with the bigger advertisers on the bigger briefs.
And where then, for example, because we work with big CPGs on multi-year category strategies, we have a pretty good view of what is being launched. And therefore, we increasingly have an opportunity for these big launches to then also talk about how do we make a big splash? And we are now increasingly involved in at least receiving briefs that go to the bigger UK media platforms.
Ian Jindal (12:09.454)
Hmm.
Ian Jindal (12:18.862)
Mmm.
Florian (12:19.84)
team of strategic planners that then answers that, where we are able to say this has worked in the past, this is what we would recommend, this is what we can, out of all of our omnichannel capabilities, put more than 20 ad products, this is what we would stitch together for the specific brand and the specific brief. And this is what we are doing in store with physical experiences or with screens. have coupons, have digitally, what do we want to do on the app? What are we doing with search? What are we doing with
display. All of that needs to get stitched together into a complete orchestration but then also an optimal aural.
Ian Jindal (12:59.65)
Fascinating. Okay, well, I'm persuaded then that well-placed advertising makes connections and takes us to the win-win-win. Although you had persuaded me last podcast, but turning then to the customer side, the idea of advertising is it's a cognitive overhead that the customer has to suffer.
in order to get towards buying what they want. And so one of the things you're challenging around that is that the advertising isn't passive, it isn't something that assaults you, it's something that you can get engaged with. So tell us a bit about that, that the age of enduring advertising is over.
Florian (13:52.435)
So let me start with one very specific example of the ClubCard challenges that we just launched. And that builds on our ClubCard CRM, where then our 24 million UK ClubCard customers can participate.
Ian Jindal (14:06.062)
And every time you say that I'm just imagining the value of that 24 million database. Anyway, carry on.
Florian (14:15.018)
So what this is, is you see a list of brands, all of which you've bought in the past, at least once. And they offer you, if you now in the next six weeks, you step up your spend and you hit certain spend milestones that are personalized to you, you get more more more club card points.
Ian Jindal (14:38.42)
like it's an accumulator that ratchets up, you know.
Florian (14:40.05)
It's an accumulator. yeah, exactly. then you as a customer, you can select out of a much longer pool of brands that you've bought in the past, you can select up to 10 for the next six weeks. So you are in charge, you can decide which ones of these you want to have a deeper relationship with. And then of our pitch to the advertiser is yes, as an advertiser, you would then need to fund these club cut points. But actually, you only pay if people have already
converted if they've actually hit the spend milestones and if they've bought your products. And I just think that that is a very transparent and super clear offer to both sides where then the customer, the shopper is in charge which of these brands they actually want to connect with. And as an advertiser you only pay for performance, you only pay if people actually buy and you can be in this pool without taking any risk.
Ian Jindal (15:39.594)
So I'm simultaneously loving and hating this at the same time. So I love the cunning. I love the structure. But let's look at this now as if it's my money. So if I'm an advertiser, then I think, that's great. I'm only paying on performance. And in fact, if they buy me but don't buy the other nine items, then I'm getting the performance for nothing because they don't get the points.
So it's a great way to spend money because I can only win. But do you then say to the the project owner, the advertiser,
In order to make yourself more attractive, you should also discount or offer something. Or is it totally vanilla and it's, the same old product that they bought once and have never bought again, trying to reactivate them? What's the conversation with the with the advertiser to help them make the most of it from their part?
Florian (16:40.202)
So the insights we typically start the conversation with is to then say people go to the yogurt aisle 18 times a year, but they only buy your yogurt four times a year on average.
And we've seen this works extremely well for categories where people have a pretty broad repertoire of what they then, what they buy, what they convert. But it actually, even in the supermarket, you typically have the situation that if it is sensitive toothpaste or if it is laundry detergent, you kind of have a couple of brands that you would be okay with. That way you go like, don't really care if it's A or B. And then offering
Ian Jindal (17:00.622)
Hmm.
Ian Jindal (17:17.411)
Yeah.
Florian (17:23.966)
at just that specific incentive, there is a certain gamification to this also because you then also report to people how they're hitting these milestones and that's all nicely animated, how you're hitting the milestones and so on. I think it is a really interesting space and at the same time of course it is also a monetary discount. Now it's being
Ian Jindal (17:36.987)
Ha
Ian Jindal (17:44.174)
But I like what you're saying though about certain categories. So as a customer, know, things might be in or out of stock. You know, if I'm passing my supermarket, I want some milk. They're out of the half litre, I'll buy a litre. You know, is it the end of my world? No, I'm still happy. It's just not what I'd expected to buy. So I'm liking that aspect of my job. From the customer perspective, do they think...
Yeah, I'm happy with this. I'm getting a good deal. Or are they thinking, hang on, I've gone with brand B because brand A was a bit expensive. They're pushing me to spend more. But yet at the same time, they're telling me that they're freezing prices. So I'm feeling a little bit gamed here. How do you make sure the customer gets all the fun of the game without feeling that they're being gamed?
Florian (18:38.516)
So first of all, they're vastly different clusters of different shoppers with different needs and there is a massive customer segmentation of some people being very price focused.
Ian Jindal (18:52.376)
Mm-hmm.
Florian (18:53.174)
and who are very much focused on the discounts they can get right now in store through the ClubCard prices right now. And there are some other people who actually feel like they should, they can save these ClubCard points and start building something up. And then we see people using these ClubCard points for all kinds of things, including holidays and we have ClubCard spend partners that might actually be
Ian Jindal (19:01.294)
Hmm.
Florian (19:23.108)
Express evening or something like that. So we see very different behaviors here. However, people want to prioritize that. What I think is people will have the feeling of spending, they're going to spend this money anyway. You select the brands, there's no one putting a gun to your head. You select the brands, you know it's going to be the next six weeks and you're going to get a feel for this. When you then say, well actually I'm not sure how much of brand X I'm actually going to buy in the next six weeks.
Ian Jindal (19:24.75)
Hmm.
Florian (19:52.961)
just bought the big laundry detergent or something maybe I'm just I definitely I'm going to need the Kleenex I'm going to need the Philadelphia or whatever it is.
Ian Jindal (19:55.918)
Yeah.
Ian Jindal (20:01.44)
Right, that's very interesting. So you must get a lot of data back as well in terms of customer propensity and attitude and intention based on what I put into my game. So some very interesting signaling there that you could probably mine as well.
Florian (20:19.818)
Probably yes. mean at the moment we are really just trying to, this is an alpha stage test. We are so happy to see that this really connects with shoppers and is actually also great for our advertisers who see this as just a really interesting way to drive loyalty. The connection of performance and pay-per-for performance and loyalty hasn't really been done that often before. So it's a super interesting space but it is really just
Ian Jindal (20:25.226)
You
Florian (20:49.802)
in symphony.
Ian Jindal (20:50.926)
Great. Well, look, I'm sure next time we have you on, you can tell me all about the data analysis behind it. But you've sort of led us on very organically to the third point that we're going to unlearn.
namely that relevance is the most important factor. Now as somebody who spends a quarter of his life screaming the word relevance at anyone who will listen, then I'm fainted what you're going to say to this because your contention is that while relevance is great, 100 % relevance equals 0 % incrementality. So
I sense a zinger there. I just need you to explain it to us.
Florian (21:35.428)
First of all, The first thing is it's called relevance and not relevancy. I don't know how some people keep calling it that. Yeah.
Ian Jindal (21:40.11)
I blame Americans. Actually, I must edit that out of the podcast. I don't blame them. I love Americans. It's a linguistic tick that everything has to be, you know, a CY at the end.
Florian (21:51.627)
It's like a double nounification. Basically, what we see is that there is a real push for relevance and I understand that that is the number one thing we need to figure out. And we see, for example, in, let's say in search, if you search for something, the search results need to be 100 % relevant to that term.
Ian Jindal (21:54.542)
You
Florian (22:16.126)
But then, and you see this on all e-commerce platforms, the question is, how do you ever bring in new inspiration? How do you talk? If someone searches for Gillette shaving blades and then gets an ad for Gillette shaving blades and buys the Gillette shaving blades, what have we done here? mean, it's amazing ROAS, but not really our eye.
Ian Jindal (22:26.627)
Hmm.
Florian (22:40.49)
And it's also a of a lost opportunity where you haven't been able to engage with someone on a more general male grooming. And both as an advertiser and as a retailer, that actually might have been a much more incremental moment.
Ian Jindal (22:54.584)
Okay, so let's pause you there because there are some obvious steps, which is if I'm buying razor blades, then obviously maybe some soap, shaving cream, whatever. You could then say, well, maybe I'll keep it relevant, but step it to broader male grooming or whatever. But then do you then jump to say, well, look, if you're shaving,
You might have a date, so I'll suggest flowers or you could be going on holiday or you know, where's the where's the what's the difference between relevance? 10 steps removed and something totally abstract or not obviously linked. So when you talk about inspiration or adding new things, what's the mechanism for adding it? And is it really just a phase of relevance or is it generally?
genuinely inspirational different.
Florian (23:55.773)
And also does it have to be relevant in this search right now or does it just have to be relevant to you as a person?
Ian Jindal (24:04.282)
Very good point. So that's a better question than mine. So answer that.
Florian (24:07.112)
Because there are two dimensions to that, think. One is, for example, if we think about we've just launched Conquesting so that if you search for Coca-Cola, you might actually get a Pepsi banner ad on top of your search results. And that was exactly this kind of a balance where Conquesting, because you can bid on your competitors' brand terms.
Ian Jindal (24:25.442)
Mm-hmm.
Ian Jindal (24:28.94)
And you call that conquesting. Nice. That's very medieval, I like it.
Florian (24:35.594)
Yes. But the point with that is we don't touch the search results. If you search for Coca-Cola, your search results are just going to be Coca-Cola and we don't inject Pepsi into that. But we have then a banner at the top that is visually separate from the search results that we are keeping clean.
Ian Jindal (24:47.054)
Hmm.
Ian Jindal (24:54.583)
Yes.
Florian (24:55.298)
and we allow them, for example, it can be very relevant for someone to know that Pepsi just now has club-cut prices, that it is on offer, and that you could actually get a really good deal there.
Ian Jindal (25:10.222)
Okay, let me connect back though to our snacks and Disney. So you could say that Florian, you're a habitual snack buyer and I know from my enormous wealth of data in the club card that you're probably a Netflix or Disney plus subscriber as well. I know where you live, blah, blah, blah. So I could.
literally just say if you're buying Doritos or something show a Netflix banner which is standard advertising. So I think what I'm going with this question is I understand your point about splitting a highly relevant answer to a focus question from the ads but given that you know so much
you know, segmentation, past history, etc. then you must be able to draw strands of relevance but just deeper, further, and in broader context and other people. is it really saying that relevance isn't important or you're just talking about a broader relevance?
Florian (26:19.772)
I just know that the point is that if you go to the extremes, think that if you go to the extremes, then the purest relevance is just not going to give you any incrementality or any inspiration.
And so it is about, so increasingly the question is, actually relevance and inspiration and incrementality not the opposing sides of a continuum, but actually an x-axis and a y-axis? And are we able to increasingly find areas where we think something is extremely relevant for someone, but on a personal level, on basically how display advertising, targeting...
Ian Jindal (26:50.542)
Hmm.
Florian (27:04.33)
has always worked, that yes, you might just now be building a basket and you're trying to do a quick evening shop here on the app. But then what is the right time to then bring in audience targeting and say, actually, we know that you're also into skincare?
Ian Jindal (27:28.408)
Mm-hmm.
Florian (27:29.13)
or we think we actually have a really good offer from a streaming service for you that we should show to you. And the blend of doing this, so for example then, is it for example at the end of your shopping trip, after you've checked out, on the thank you page, is that then the moment where you start a new journey?
Ian Jindal (27:48.33)
Right, so how are you measuring this then? Because again, I'm persuaded of the logic of the approach, but depending on how you measure things, it will either look like a success or look like a failure. So if you're charging on a per search or conversion per search, then the broader picture might seem like a distraction. Whereas if you're looking at annual customer profitability or growth in spend, these could be
very positive. So in coming to a view around relevance, what is your measurement framework that we should consider for your point?
Florian (28:32.97)
I see these as basically concentric circles, where at the core you have your fundamental add metrics, so things like click-through rate, where...
especially in the performance space, for example, when it comes to conquesting, if this ad for Pepsi when people search for Coca-Cola doesn't get any clicks or gets below average clicks, that is probably would be a cause for us to look at this more closely and say, do we need to govern relevance in this space? That's one. The second thing is we do a lot of testing, A-B testing together with our retail colleagues, and we then also monitor the overall
Ian Jindal (29:06.189)
Hmm.
Florian (29:16.3)
experience. For example, the number of items people add to cart, overall basket values and anything, any diagnostics that would tell us that there is something going wrong in this shopping experience. Are we putting speed bumps in here? Are we distracting? Are we causing people to somehow lose the thread? Which also would be really some real alarm bells for me to say that there is something, for example, in the relevance.
Ian Jindal (29:38.158)
Hmm.
Florian (29:44.428)
And then you come on the outer edges of this, you come towards does this actually have a positive payout for the advertiser?
And so we do a, we also, do their, if that is a B testing, if that is surveys and so on, where we are able to show that it is a positive experience for the shopper and it has either a brand building effect or it has a customer lifetime value effect, which we can then, we can then see over time. Because then look fundamentally, if it's not connecting with the shopper,
Ian Jindal (30:11.342)
you
Florian (30:24.646)
If it's not relevant, it's just not good advertising. I would absolutely say that. And that is either because the content is wrong or because the targeting is wrong or because the call to action is wrong, all of this. But somehow, if the message is not connecting with the recipient, then you are not going to see the ROI of the ad. And I don't think that's ever going to change.
Ian Jindal (30:46.509)
Right.
Good, well, let's wrap up then, because I think in a way it's all things have pointed to this, which is the last point about it not just being all out nonstop 100 % monetization, but rather looking at the customer's trajectory and then picking the adverts based on what they've actually done. So just tell us a bit about this Florian and the changes you've made there.
Florian (31:16.532)
So this is called Smart Stock Audiences, and that's something that we've tested together with Unilever, where fundamentally the idea is there might be some times when it's better not to serve an ad. And here the fundamentally, if someone has bought mayonnaise just last weekend, they might be in the absolute core demographic target group.
Ian Jindal (31:35.149)
Mm-hmm.
Florian (31:40.837)
audience for the Helmets mayonnaise. They might be lovers of it. They might be competitive users, whatever it is. But if they've just bought a big jar of mayonnaise just this weekend, this might not be the time to advertise more mayonnaise. That's the thinking. So they are called smart stock audiences. And we are using these to inject these into our overall planning for display advertising to say, how do we control
Ian Jindal (31:53.491)
Yes.
Yeah.
Florian (32:10.7)
for efficiency for the advertiser but also how do we control for the shopper experience for the consumer experience where you will always feel better when you are receiving ads for products that you haven't just bought.
Ian Jindal (32:26.51)
So we see this in the opposite way which is I'll go to my supermarket, I'll be building my basket and go, hey Ian, you haven't bought toilet paper, tomato sauce because typically you buy these every 32 days. And so it's 34 days, may we suggest. So I get that. I also get the other thing.
which is the suppression because the big items you buy washing machine is guaranteed for 25 years and the next Friday get an email saying would you like to buy another washing machine is like no guys but they only have 200 skews so it's quite easy to say he's bought the washing machine get him to buy the washing fluids and so on in your case you know a family check out might have 200 or more products in it
And so you have to juggle all of this data about what I did and didn't buy and tell the ad serving and optimize. That's a much bigger task than the normal of just promote everything every couple of weeks. So you're kind of taking on quite a big administrative logical challenge there.
Florian (33:46.971)
It is a massive task and is that have in Dunhamby has 20 plus years in this kind of data analysis. It's actually just now Club Cards 30th birthday. But what we are but what we know.
Ian Jindal (34:00.395)
god!
Florian (34:03.594)
What we are really trying to do is then for the campaigns that are running to actually drive these kinds of predictive audiences. And that could either be that you are a high propensity trialist for a certain brand, that we try to predict that, or that you're a future decliner, that we think that your consumption of this brand is going down, or for example, smart stock that you already have too much of this category at home. And...
Ian Jindal (34:28.419)
Mm-hmm.
Florian (34:30.918)
Those are, of course, we don't do this for all of the customers and also, you it's always very much in aggregate, but we create these audiences for the specific campaigns that are being.
Ian Jindal (34:43.818)
Right, so it's not on an individual basis. Yeah.
Florian (34:45.93)
Exactly. We keep any kind of data processing on an individual level. At the absolute minimum, it's always about big aggregate audiences that we create and therefore we want to make really sure that we are always 100 % privacy safe.
Ian Jindal (34:57.762)
Hmm
Ian Jindal (35:04.494)
Well listen, that was a very good moment to end on today. I think we said pause because I feel with all of these there's an ongoing conversation that we need to keep returning to. So I will say pause because it would be lovely to get you back in the studio in a couple of months to hear how all these ideas have developed. But you've given us, I think, more than one lesson to be unlearned.
rather a process, I think, of challenging so many bits of marketing around, you know, relevance, experience, what the ROI is, but also, I'm sad to say, going back to your phrase of the win-win-win, which is balancing the needs of the advertiser, the retailer, and the customer, which is, to me, sounding like a lesson well learned. Florian, it's been lovely chatting to you again.
Thank you so much for joining us in the studio.
Florian (36:02.325)
Thank you for having me.