Latest / IAB Europe podcasts / AI in Retail & Commerce Media: From Hype to Hands‑On

Transcript

AI in Retail & Commerce Media: From Hype to Hands‑On · IAB Europe podcasts

0:05
SPEAKER_02: Welcome to the Retail Media Roundtable, the podcast of MyVER that dives into the dynamic world of retail media advertising.
0:13
SPEAKER_02: I'm Marie Katy, the Industry Development and Insights Director at Hive Europe, and in each episode I invite two expert guests from across the industry to share their insights, exchange ideas, and explore the latest trends and challenges shaping the retail media landscape in Europe.
0:29
SPEAKER_02: So whether you're here for practical tips, expert perspectives, or just a deeper understanding of this rapidly evolving space, the Retail Media Roundtable is your go-to for engaging and thought-provoking conversations.
0:41
SPEAKER_02: So grab a seat at the table and let's dive in.
0:45
SPEAKER_02: Welcome everyone to this week's episode of the Retail Media Roundtable podcast from IB Europe.
0:50
SPEAKER_02: My name is Marie Claire.
0:52
SPEAKER_02: I am the Senior Director of Industry Development and Marketing.
0:55
SPEAKER_02: And I'm very happy to be joined today by James Taylor and Cedric Epiard.
1:01
SPEAKER_02: James, would you like to kick off introducing yourself, please?
1:04
SPEAKER_02: And also I'd love to know what your peak of the week has been.
1:10
SPEAKER_01: Peak of the week.
1:13
SPEAKER_02: The pinnacle of the week.
1:15
SPEAKER_01: Gosh, I'll think about that while I'm doing the intro.
1:18
SPEAKER_01: My name's James.
1:20
SPEAKER_01: I'm the founder and CEO of Particular Audience.
1:23
SPEAKER_01: Particular Audience was started back in 2017 to effectively democratize the technology on the Amazon.com front end for other retailers.
1:32
SPEAKER_01: We specialize in uh fairly advanced uh hyperpersonalization tools, recommendation systems.
1:40
SPEAKER_01: We have a transformer-based search engine, which is significantly uh more advanced than legacy keyword search engines.
1:49
SPEAKER_01: Um, and we have a full retail media suite uh built into all of that.
1:53
SPEAKER_01: And we we believe that similar decisioning that powers organic search and discovery should power um retail media as well, uh, introduces better customer relevance, more automation for the advertisers, and better yield for the retailers.
2:08
SPEAKER_01: Um the the tripartite win.
2:10
SPEAKER_01: Um peak of the week.
2:12
SPEAKER_01: I had I've just moved house and uh I don't, you know, England is uh bathed in sunshine this week, and I had a coffee with my uh soon-to-be three-year-old son uh this week.
2:26
SPEAKER_01: So that would be my peak of the week.
2:29
SPEAKER_02: It's a great peak of the week.
2:30
SPEAKER_02: You can't beat coffee with the kids.
2:32
SPEAKER_02: Um Cedric.
2:34
SPEAKER_01: I didn't give him any.
2:35
SPEAKER_02: Oh, okay.
2:36
SPEAKER_02: That's that's also okay.
2:39
SPEAKER_02: Cedric, over to you for intros and your peak of the week.
2:43
SPEAKER_00: Um I will present myself before.
2:46
SPEAKER_00: Um Cedric, I'm the VP product of Miracle Ads, and I worked uh since 17 years on retail, retail media, polymatics.
2:54
SPEAKER_00: I previously worked for C Discontreen, which is a French big retailer in France.
2:59
SPEAKER_00: And um my peak of the week will be uh less important than uh James one.
3:04
SPEAKER_00: Um it was uh two days with my team uh on a training, on a skills uh communication training on the on the beginning of the week.
3:15
SPEAKER_00: And it made me think that um the skill that we we need to have, it's not AI, it's to be able to communicate uh between ourselves and more and more with AI, we will need to have this kind of skills.
3:27
SPEAKER_00: So it makes me reflect that uh we need to stand by sometime to to reflect ourselves to understand what is what is matter, what are our main objectives.
3:37
SPEAKER_00: So it makes me think um beside my uh daily agenda of uh meeting and product, etc.
3:44
SPEAKER_00: So it was uh really important for me this week.
3:47
SPEAKER_02: It's interesting that you say that.
3:49
SPEAKER_02: As I've literally just come off of a two two-day team offsite, and we also had training on communication.
3:54
SPEAKER_02: So maybe, like you say, it's more and more of a need for actual interpersonal and communication skills in an increasing world of of AI.
4:03
SPEAKER_02: So that's uh it's interesting.
4:05
SPEAKER_01: I think a big the big problem is we're all sat behind keyboards a lot of the time, right?
4:09
SPEAKER_01: Like we're all in Slack and Teams, and it's amazing what you can just shortcut if you get on a call.
4:15
SPEAKER_01: 100% talk through.
4:17
SPEAKER_02: Yeah, yeah.
4:18
SPEAKER_02: But with that in mind, we are going to focus on the topic of AI today.
4:23
SPEAKER_02: Um, because there are obviously lots of lots of great stuff.
4:26
SPEAKER_02: Uh there's a lot of great stuff happening.
4:28
SPEAKER_02: Um, and I really want to get um into kind of the hype of AI, um, but also the hands-on practical applications of AI today.
4:38
SPEAKER_02: So I'm gonna start with my first question, uh, which is we hear a lot about AI transforming retail and commerce media.
4:47
SPEAKER_02: But when you strip away that hype, um, if it is a hype, what would you say right now?
4:54
SPEAKER_02: Let's have a reality check.
4:55
SPEAKER_02: What's the most practical um ways that organizations could start using AI today?
5:03
SPEAKER_02: Um, Cedric, I don't know if you want to kick us off, please.
5:07
SPEAKER_00: Yeah, it's a complex subject because yeah, you we have a lot of uh buzz around the AI, and um the AI is evolved so fast that it's uh I I love to use this uh this sentence that AI never sleep.
5:21
SPEAKER_00: At every morning you will have news about something that didn't work yesterday or work now on AI.
5:28
SPEAKER_00: So my my advice on what I'm currently doing is to at each task that I have in my day, I'm trying to, and you need time, you need to invest time on that on that.
5:38
SPEAKER_00: If you are not investing time, it will be uh difficult.
5:41
SPEAKER_00: You need to check if it's AI, if at AI could help you to improve the quality, the speed of your every task is and at the end, maybe you will find uh uh an autonomous workflow that can work alone, but almost of the time it's more about find the task that AI can help you.
6:00
SPEAKER_00: For example, I will give you give you an example, a practical example.
6:04
SPEAKER_00: No, I'm able to have an agent that is my chief of staff.
6:09
SPEAKER_00: It's uh it was a long run to do that because on the first stage I need to collect all the information of my day, because uh a shift of task is it's like a shadow of you.
6:20
SPEAKER_00: Uh so it was a part to just to say that I need to have granula for transcript of all of my meetings.
6:26
SPEAKER_00: If the AI have all my context, it's able to get it, and I I will give feedback with that after that.
6:34
SPEAKER_00: So it's a daily step by daily step.
6:36
SPEAKER_00: Um, you find the task that it's consumes a lot of time for you.
6:40
SPEAKER_00: You try if AI is able to uh help you on a day-to-day basis, and if not, you need to repeat it maybe two months after, three months after, because as I said, AI evolved a lot.
6:52
SPEAKER_00: So we need to be to have in mind that the the the skills that AI have here is not the skill that it will have tomorrow.
7:01
SPEAKER_00: So it's my personal plan to do that.
7:03
SPEAKER_00: And sometimes my chief of staff doesn't work at all uh because uh cloud relaunched a new version that broke everything, and uh we need to to work with OI, and it's uh a bit disturbing when you have these kind of days.
7:18
SPEAKER_02: I I love that.
7:19
SPEAKER_02: I love um I'm I'm already thinking about how to create uh a like a PA or a cheaper staff with with AI.
7:25
SPEAKER_02: So um my brain is worrying.
7:27
SPEAKER_00: James, I can give you advice how to do that and how to make automate it, but it's it's easy.
7:33
SPEAKER_00: I really know it's easy with cloud code on cloud code work, it's really easy to do that.
7:37
SPEAKER_00: So if you need advice, I will give you after that.
7:40
SPEAKER_00: I'm sure that you will give me give me back uh other feedback on other advice in your side.
7:44
SPEAKER_00: Perfect.
7:45
SPEAKER_01: Cool.
7:46
SPEAKER_01: So Drake, how how do you how do you find the uh accessibility of this agency chief of staff to all of the systems that you use and databases and SaaS tools and things that you need a you would need a real chief of staff to access.
8:02
SPEAKER_01: How do you handle that?
8:03
SPEAKER_00: Yeah, yeah, it's um it was my fear because I was previously working for a company which was really close the bank of stuff.
8:10
SPEAKER_00: I think we will discuss in uh in the next couple of questions.
8:14
SPEAKER_00: Uh, but uh uh at Miracle, we are very open uh to open the tools to be able to catch the information or to work on, because the main difficulty is to just to have an LLM that answering your question, but have no capabilities to do anything.
8:29
SPEAKER_00: So you you are very happy to have a big brain, but you can do nothing with that.
8:33
SPEAKER_00: So at Miracle, um founder uh Philip asks us to use AI at everyday basis.
8:40
SPEAKER_00: And if we need something, which means unlock a tool or to buy a license, there is no problem at all.
8:47
SPEAKER_00: So we have all the conditions that help us to try, test, learn on that subject.
8:52
SPEAKER_00: So it's really easy for me because I didn't have to ask something.
8:57
SPEAKER_00: Yeah.
8:58
SPEAKER_00: James, I am sorry, go on.
9:01
SPEAKER_01: Sorry, there you go.
9:02
SPEAKER_01: I'm sick.
9:02
SPEAKER_02: I was just gonna say, does this resonate with you and how are you like what would be your kind of most practical tips to start using AI, maybe with some retail media thoughts as well?
9:15
SPEAKER_01: Yeah, yeah, and and so I think also important to caveat when we're talking about the transformation that retail media may or may not see, there is the consumer uh and their AI journey.
9:28
SPEAKER_01: There is obviously the the advertiser uh and and their application of AI, and then there is the retailer operationally.
9:36
SPEAKER_01: Um, but I think that everybody in that, and also for agencies and basically anyone that's a professional, to Cedric's point, you know, how are they using AI?
9:48
SPEAKER_01: For me, the biggest barriers are it being able to authenticate and access and execute across the platforms that I would expect a new hire to work on and work with.
10:00
SPEAKER_01: And um I have been I'm hoping to have a pretty wild Saturday night uh this weekend.
10:10
SPEAKER_01: And actually, I'm a bit late to the party on OpenClaw because I sort of wanted to be a follower rather than a leader in this and see who else, you know, jeopardized their system security or deleted all of their data or whatever it may be that might have gone wrong.
10:28
SPEAKER_01: I feel like it's had some time to mature to a stage where I'm going to um I've I've actually got a device here which is gonna be my open claw MacBook Pro, and it has already got email accounts, Slack accounts, CRM logins.
10:47
SPEAKER_01: Um let's see.
10:49
SPEAKER_01: So ask me next time how it went.
10:52
SPEAKER_01: Um because like you, I would love to have a always-on virtual chief of staff, EA support rep type type uh agent.
11:04
SPEAKER_01: Um how that corresponds to to retail media.
11:08
SPEAKER_01: So I think firstly you have currently manual workflows when you're booking advertisement.
11:23
SPEAKER_01: Um this is not limited to commerce media.
11:26
SPEAKER_01: And there is uh a vision that if you could have agents that are able to accurately interpret reason and effect change, that you will have a more efficient or less hands-on and arguably more efficient from a metrics and performance uh standpoint, you'll make more money from your advertising campaigns.
11:50
SPEAKER_01: Now, I you know, there's been some sort of counter discourse around, okay, but if I have an agent and an agent and they're both optimizing on two different sides of the same metric, you know, where the where the head's butt.
12:03
SPEAKER_01: But I don't necessarily think that's too much different to a buyer and a seller in any marketplace ever in the history of human humanity.
12:10
SPEAKER_01: So I think that there are lots to do.
12:14
SPEAKER_01: I think there's a trust issue still.
12:16
SPEAKER_01: Um, I don't fully trust AI.
12:18
SPEAKER_01: Um, I've had bad tax advice from AI, I've had bad plumbing advice from AI, both of which have cost me lots of money.
12:27
SPEAKER_01: Um, I wouldn't necessarily want an autonomous agent to go and cost me lots of money accidentally uh on the advertising side of things.
12:38
SPEAKER_01: Um, so I think that the systems that enable this for advertisers need to be a ask for permission rather than forgiveness first policy.
12:51
SPEAKER_01: And then perhaps over time, when you start to just be approving absolutely everything in the list all the time, you might switch off approvals.
12:58
SPEAKER_01: And I think that the dexterity of which approvals you can switch on and off uh in a platform is going to matter more and more.
13:06
SPEAKER_01: Um none of that is to stop you having an agent that simply monitors and communicates with you, right?
13:14
SPEAKER_01: Which I think is something that is possible today and with the right APIs.
13:19
SPEAKER_01: In fact, we've actually just uh relaunched a whole bunch of reporting APIs because we want to make sure that they are accessible via MCP or directly through agentic workflows.
13:31
SPEAKER_01: Um, stuff that didn't need to be a reporting API previously, but fueled something in our dashboard.
13:37
SPEAKER_01: I think increasingly people won't necessarily need the dashboard if they can be talking to the AI.
13:44
SPEAKER_01: From a retailer standpoint, same thing operationally, there's a bunch of gains to be had.
13:49
SPEAKER_01: Um, same thing with you know yield management and performance management, um, monitoring and potentially future change.
13:56
SPEAKER_01: But I think really where it's, and you know, dare I say this is not always agentic.
14:02
SPEAKER_01: In fact, it's uh you know seldom agentic, but with the predictive AI and how they can use the search and recommendation systems to automate relevant and high performance targeting and remove a lot of the manual configuration uh flows for their advertisers.
14:18
SPEAKER_01: I think that's been possible for years, but the majority of the you know incumbent legacy commerce media technology platforms are still doing manual keyword targeting and manual placement targeting and manual segment targeting.
14:32
SPEAKER_01: And it that's just not the future, right?
14:34
SPEAKER_01: And I probably speak on behalf of Miracle Ads with this as well.
14:37
SPEAKER_01: Um, you know, we we see a very much a different world there.
14:43
SPEAKER_01: Um and then the consumer.
14:45
SPEAKER_01: Well, shall we speculate, Cedric?
14:47
SPEAKER_01: What do you think?
14:48
SPEAKER_01: Uh is the consumer gonna relinquish all of their shopping behavior to an agent or not?
14:54
SPEAKER_00: It depends on the customer, the type, the kind of customer.
14:58
SPEAKER_00: Um, if it's uh an AI enthusiast like me, yes, it could be, but at the end, uh what we saw uh with the open AI uh tentative about a checkout on this system is that they come back to reality.
15:12
SPEAKER_00: It's really complex to manage a retailer side.
15:16
SPEAKER_00: You need to uh to manage the logistics, you need to manage your support, uh, you need to manage a lot of things, and after that, if your uh your retailers have all the information about you and understand also uh what is uh your last order, etc.
15:32
SPEAKER_00: It it could also give you the right advice on on the right product.
15:37
SPEAKER_00: So I'm sure that on the customer side we will have more and more checkout directly on the on the on the llm for sure, but I think that it will be more up into the llm, for example, uh stack art.
15:51
SPEAKER_00: You have the the the app on uh on openai, you you will be able to ask uh openai to uh have information on to make an order, but in reality, it's an app into openai that that is making the job.
16:04
SPEAKER_00: So I think we will be more in that subject in tomorrow, and it will be for uh tools uh for objects that you don't care, not you don't care, but you you don't want to think a lot uh if it's a good product or not.
16:20
SPEAKER_00: You will have all the discovery in ZLM, I'm sure of that, because on the discovery will be more the intent.
16:26
SPEAKER_00: I want to uh uh to to make hiking with my family.
16:30
SPEAKER_00: I need to find the the the right product to do that.
16:33
SPEAKER_00: I did I did not know nothing about hiking, so uh let me understand what I need to do uh and what I need to buy.
16:41
SPEAKER_00: Um after that, you will spend time on a lot of different retailers.
16:46
SPEAKER_00: We are not thinking that LLM will will take all.
16:49
SPEAKER_01: It's not uh yes, I think this is exactly hiking's the best example, right?
16:55
SPEAKER_01: You go into an area that you know nothing about or you want to know more about.
16:59
SPEAKER_01: I I shopped for niche for mooths last night, right, with the help of AI, right?
17:05
SPEAKER_01: Yeah, I love Negroni and I like I wanna so it really like helps helps you through that process.
17:11
SPEAKER_01: I was I was at I spoke on the panel at the retail innovation forum that Dun Humbey put on this week up in London, and one of the questions was what's AI good for today in the shopping journey?
17:22
SPEAKER_01: And my answer was filtering.
17:24
SPEAKER_01: Yeah, right.
17:25
SPEAKER_01: It is it's just another way to filter through lots and lots and lots of information, right?
17:31
SPEAKER_01: And the same way that recommendation search systems have always been methods of information filtering.
17:36
SPEAKER_01: Filtering is a great use case for us as consumers.
17:41
SPEAKER_01: Saves us time, right?
17:42
SPEAKER_02: It's a more effective way of yeah, it's time saving, which I don't know about you guys, but in today's it, you know, when everyone's so busy and uh it just it's invaluable.
17:52
SPEAKER_02: I I think I just want to move like beyond just the shopping and the consumer point.
17:57
SPEAKER_02: I think it's a really important point.
17:58
SPEAKER_02: And you know, maybe we maybe we leave plumbing advice for plumbers and not AI just now, and tax advice for accountants.
18:06
SPEAKER_01: Well, and like actually any advice, you sort of everyone's got advice to give.
18:11
SPEAKER_01: You don't have to take it, and I think there's no difference with AI.
18:14
SPEAKER_01: It's fine to read it now, but you know, I I didn't buy every vermouth it suggested, actually.
18:19
SPEAKER_02: Yeah.
18:20
SPEAKER_00: Uh what is the major challenge on that is that uh sorry, uh sorry, Marie.
18:25
SPEAKER_02: No, no, carry on.
18:26
SPEAKER_00: Uh um, but it's uh the major challenge is that you you will be able to have an expert of each subject with an eye.
18:32
SPEAKER_00: Uh, if you are speaking to your friend, you will have an advice, but maybe not an expert advice.
18:37
SPEAKER_02: Yeah, yeah.
18:38
SPEAKER_00: Um it's um I think it's uh the CEO of uh Anthropic said that you have a room with all doctorants and uh uh very experts uh in each um in each kind of knowledge you will have in your pocket now.
18:53
SPEAKER_00: So it helps to have um good advice, I think.
18:59
SPEAKER_02: Just just other use cases.
19:01
SPEAKER_02: So beyond kind of and and James, you touched on a little bit like yield management for retailers, etc.
19:08
SPEAKER_02: From what are you actually seeing, or where are you seeing the most meaningful AI use cases actually working today?
19:16
SPEAKER_02: We've talked a lot about kind of things that might work tomorrow, but today, what do you think is working and working well?
19:23
SPEAKER_02: And maybe there's something that's working but working not too well.
19:26
SPEAKER_01: Yeah.
19:28
SPEAKER_01: So um I think this conversation is gonna be biased to on-site applications.
19:36
SPEAKER_01: Um, it's where Cedric and I probably know a lot, and where it comes to in-store and omni, maybe maybe we know slightly less.
19:44
SPEAKER_01: Um, although I don't want to speak on behalf of Cedric, I'm sure he knows a lot more than me.
19:49
SPEAKER_01: Um I think first of all, what is what is yield, right?
19:57
SPEAKER_01: Yield is the the most possible value from each individual pixel in an on-site context.
20:08
SPEAKER_01: Um now to a retailer, it's not just retail media, it's also the sale of the item.
20:15
SPEAKER_01: So you have this um, you know, let's just simplify a formula here.
20:21
SPEAKER_01: You know, you have something about conversion rate to impression, right?
20:24
SPEAKER_01: And then you have is there an opportunity to add advertising revenue within that without degrading the ultimate sale, right?
20:33
SPEAKER_01: So ultimately we're looking to maximize that profit per pixel.
20:37
SPEAKER_01: Um now if you break it down, how does AI help us help us to do that?
20:45
SPEAKER_01: Um now it's definitely not just PCTR, it's definitely not just predicted CTR, right?
20:53
SPEAKER_01: This is it's not a complete solution.
20:55
SPEAKER_01: Uh it that that is an ad tech company saying that they have smart, intelligent approaches to solve uh for ad yield, but ad yield is not the full goal of the retailer, right?
21:09
SPEAKER_01: So yield is is the whole thing.
21:11
SPEAKER_01: Now, then there are other things like we talk about pixel conversion and potentially ad profit as well.
21:19
SPEAKER_01: Um, but even that's not the full picture, right?
21:22
SPEAKER_01: The top line sales of an item um are not the profit, the bottom line to the retailer.
21:29
SPEAKER_01: Um, it gets really complex.
21:31
SPEAKER_01: You know, you have private label that has higher margin.
21:33
SPEAKER_01: Okay, you have third-party brands.
21:36
SPEAKER_01: You might have certain trade agreements where if you hit a certain sales volume for a certain item, you're gonna get a better margin on it because they're gonna give you more free units on the next order or uh a discount off your next order or some sort of bonus or repayment or credit back.
21:50
SPEAKER_01: You know, that there's this sort of nebulous fabric of incentives that retailers can get through certain items.
21:57
SPEAKER_01: And then on the flip side, what about like Recovering lost money.
22:01
SPEAKER_01: Like I bought stock and inventory and it's just not selling.
22:05
SPEAKER_01: So my cash is stuck in this mountain of t-shirts in this warehouse that I can't ship.
22:10
SPEAKER_01: So I need to reduce the price on that.
22:12
SPEAKER_01: Do I?
22:13
SPEAKER_01: Or can I increase sell-through on that?
22:14
SPEAKER_01: And even worse, in grocery, you might have spoiling inventory that's about to go off.
22:19
SPEAKER_01: Like you need to move that stuff out the doors quickly because you don't have the luxury of time.
22:23
SPEAKER_01: So when we come to yield management, I'm thinking like the entire value chain there.
22:30
SPEAKER_01: Ultimately, what is it going to mean to a customer, to a retailer, to have more cash on hand and have moved stock through quicker at the highest profit?
22:40
SPEAKER_01: Right.
22:40
SPEAKER_01: And AI can literally help us in every step of that journey.
22:44
SPEAKER_01: To give you a feel for that, you know, where our recommendation systems sit, there's a cascade of algorithms.
22:51
SPEAKER_01: And you can A-B test an algorithm in the slot that can tell us the difference between a highly merchandised filter set sort of rich algorithm versus the bare-naked algorithm.
23:04
SPEAKER_01: What happens with the click-through and the click conversion rate of those?
23:08
SPEAKER_01: And how does that correspond to an overall advertising engagement if you're doing a CPC campaign?
23:14
SPEAKER_01: You may also be doing a CPM campaign or a CPA campaign or even a fixed tenancy campaign.
23:20
SPEAKER_01: So there's just all of these like multi-dimension problems all the way down each of the goals that a retailer wants to have.
23:28
SPEAKER_01: Is it helpful to have AI to reason through that?
23:30
SPEAKER_01: Absolutely, yes.
23:32
SPEAKER_01: You need the measurement of each of those tiers in the first place.
24:08
SPEAKER_00: Should I promote with a coupon?
24:11
SPEAKER_00: Should I uh below my price, decrease my price, or should I uh make advertising?
24:18
SPEAKER_00: Or maybe it's a make of that.
24:20
SPEAKER_00: With AI, this very difficult problem could be solved.
24:25
SPEAKER_00: And you said um the retailer can have a goal to uh uh to clear his inventory because he needs to have more space in the warehouse.
24:35
SPEAKER_00: So uh it's the same for the sellers and the same for another advertiser.
24:40
SPEAKER_00: So you you will fix the goal, but you have you fix your rules, and after that, you fix uh the eye will be able to handle the the complex equation, which is which is also a real-time equation.
24:53
SPEAKER_00: It's not possible to do it manually, because at each hour your inventory stock level is different, and your the competition is also moving at real time.
25:06
SPEAKER_00: That's why AI and algorithm are so important these days, because the pacing will be more and more important, the decisioning will be more important, and at the end, the main the main things that we need to have in mind is what are our main goals?
25:25
SPEAKER_00: Are we clear with that or not?
25:27
SPEAKER_00: And we need to spend time on these subjects.
25:29
SPEAKER_00: What is the vision?
25:31
SPEAKER_00: What we need to achieve, and uh the the AI will be will be do all the the stuff with the the hard stuff for us.
25:40
SPEAKER_01: It's interesting, even there, like and then it's we talk about pacing, right?
25:45
SPEAKER_01: Pacing is this like legacy thing of advertising where for some reason we want to stretch a budget over a campaign period.
25:53
SPEAKER_01: Um why pace?
25:56
SPEAKER_01: Um ultimately, you probably should be pacing in line with your ability to ship product, right?
26:02
SPEAKER_01: Like pacing should almost be related to uh how much you can physically get out the door.
26:08
SPEAKER_01: And if you advertise and it leads to sales, you know, that that's almost how pacing, or it's one argument of how pacing should work.
26:17
SPEAKER_01: Cedric, how do you think about pacing?
26:18
SPEAKER_01: Because it shouldn't be this arbitrary thing of just spend this equally over the next 30 days, right?
26:24
SPEAKER_00: Yeah.
26:25
SPEAKER_00: Um, I mean between it, it will depend on your objective.
26:29
SPEAKER_00: For pacing, we are we are working on a pacing feature on our product, and um why?
26:34
SPEAKER_00: Because we are seeing that sellers and advertisers have uh a small amount of uh bullet daily budget, and they're just delivering uh advertising on uh zero o'clock to nine o'clock the morning.
26:47
SPEAKER_00: So you are not fitting the right hours that are your that the CVR is uh is a high peak.
26:54
SPEAKER_00: So it's it's not worth for them to do that.
26:56
SPEAKER_00: So we are imagining pacing for that, and we also saw uh, and I also saw in my previous experience that uh on the each day of the month you will not have the same CVR.
27:07
SPEAKER_00: On the beginning of the month, people have money, of course, but at the end of the month it's more difficult to spend money.
27:13
SPEAKER_00: So you need to invest more on the beginning of the month.
27:16
SPEAKER_00: So I think yes, I'm in between with this thing because we can provide the tools, but at the end, it needs to be understood by the by by uh by the sellers, by the the brands, uh, etc.
27:28
SPEAKER_00: Also by the retailers themselves.
27:30
SPEAKER_01: Yeah, I I think that's a really, really good point, right?
27:32
SPEAKER_01: Which is let's say that payday happens at the end of the month, right?
27:36
SPEAKER_01: So people are flush with cash at the start of the month, right?
27:39
SPEAKER_01: Now, a sufficiently accurate AI solution uh may be able to sufficiently forecast when the best time it is to run your limited ad budgets at a time where you're likely to get the best performance from those things.
27:55
SPEAKER_03: Yes.
27:56
SPEAKER_01: Um so like this is a really good application of how the legacy pacing for pacing's sake, I think we like need to just completely get out of our minds when it comes to retail, because we're so close to that point of conversion.
28:08
SPEAKER_01: So retail media shouldn't be pacing for pacing's sake, they should be pacing because it's going to be a more efficient way to spend the money.
28:16
SPEAKER_01: Yeah.
28:16
SPEAKER_00: Yeah.
28:18
SPEAKER_00: And we need to, and as a retailer or technology provided for the retailer, we need to open the box to let the AI work with this kind of data.
28:29
SPEAKER_00: We we we we need to be transparent with that.
28:32
SPEAKER_02: So that leads me nicely onto my next question.
28:34
SPEAKER_02: I've really I found no, no, it's fine.
28:36
SPEAKER_02: I found that that conversation about kind of the AI and the algorithm and the pacing, like the the kind of potential there.
28:45
SPEAKER_02: I feel like we're probably just on the surface of that right now.
28:48
SPEAKER_02: It feels like.
28:50
SPEAKER_02: Um, I don't feel that's kind of operating at any scale.
28:53
SPEAKER_01: I could be wrong, but AI has actually been pretty bad at math.
28:57
SPEAKER_01: Uh, generative AI has been pretty bad at math so far.
29:01
SPEAKER_01: So I think that even then, you know, we verge back to statistical approaches.
29:07
SPEAKER_01: And it's like even in a forecasting, let's say that you're you're taking a uh just to put it simply, right?
29:13
SPEAKER_01: Let's say we're taking like a long-term trend data set and we're layering over that a moving average of a more recent period, right?
29:21
SPEAKER_01: And you're looking to have come up with a prediction that takes into account both that short-term and that long-term prediction set.
29:28
SPEAKER_01: You know, we get into an area called temporal or ordinal uh models, which is all about prediction, right?
29:34
SPEAKER_01: It's it's it's forecasting based on long-term, short-term.
29:37
SPEAKER_01: In fact, there's literally an algorithm called that.
29:40
SPEAKER_01: And um is that generative AI?
29:44
SPEAKER_01: No, it's not.
29:45
SPEAKER_01: Right?
29:45
SPEAKER_01: It it's could even be statistical, moving back into sort of forward into machine learning.
29:52
SPEAKER_01: Um, I don't know if generative AI has an absolute solution on it yet, apart from the ability to be able to look into the results of the forecast, look at the weightings and reason through those.
30:05
SPEAKER_01: Um and that's where you're having like like what is the type of AI that we're specifically talking about for each of these functions because it is a multi-layer statistical through machine learning, and then generate on predictive AI, it's generative AI.
30:19
SPEAKER_01: Like, what are we calling what?
30:21
SPEAKER_01: But it's usually a combination of technology.
30:24
SPEAKER_00: Totally agree.
30:24
SPEAKER_00: And at the end, you will have a mix of technology.
30:28
SPEAKER_00: Uh, you will have the the LLM will be orchestrating all the goals, and after that, you are using algorithms to achieve, really achieve at a cost which is acceptable because AI it's really costly technology.
30:44
SPEAKER_02: So that that the kind of leading me into my next uh question, which is retailers, retail media no ex brands often struggling with kind of legacy systems, legacy data um setups, fragmented data, silos organizationally and you know, and internally, they're obviously going to be blockers, I would have thought, to AI adoption.
31:08
SPEAKER_02: How does that how does that kind of pan out in reality?
31:12
SPEAKER_02: Uh and what what kind of what's the biggest blocker right now you you see um in adoption of of AI um tools?
31:22
SPEAKER_00: Human.
31:24
SPEAKER_00: The humans, people, yeah, people okay.
31:28
SPEAKER_00: People will be the main the first main blocker because um I saw more and more people that are enthusiastic with AI or doomers, and um some people are in between, but uh we need to to convince people in our organization in the retailer side, on one side is the same uh at the end that AI is not a danger, it's uh uh a tool that helps to accelerate, to uh elevate yourself to do other tasks and to to think about uh other subjects.
31:59
SPEAKER_00: So you you you need to have all people thinking in the right direction.
32:05
SPEAKER_00: The second it's security and scaling because you need to be sure that and security it's uh it's a two phases.
32:14
SPEAKER_00: The first is to be sure that the wrong data will not spread over the internet, of course, and the second is security because there is always uh we have already a lot of securities in our companies that is used to block the access of some information.
32:34
SPEAKER_00: And if you want to have, as we said on previously on the discussion, if you want to unlock uh AI as its first potential, you need to give it access to the to the information.
32:46
SPEAKER_00: I think the the two main topic that we need to handle first uh is security and human uh uh behavior.
32:55
SPEAKER_02: James, would you agree?
32:57
SPEAKER_02: Is there anything else?
33:00
SPEAKER_01: Uh definitely agree.
33:02
SPEAKER_01: Um I think that it could be maybe easiest to think about it within a framework.
33:10
SPEAKER_01: Um and it's sort of how I I my mind tends to try to work.
33:17
SPEAKER_01: Um so um has anyone seen this news about X uh XAI, SpaceX?
33:26
SPEAKER_01: Yeah uh basically buying an option to buy Cursor, the the development agent.
33:35
SPEAKER_01: Right?
33:35
SPEAKER_01: Like I think it's like is it 10 billion now with an option to acquire the 60 billion within a year, right?
33:47
SPEAKER_01: Now um Cursor is a it's an IDE, it's what engineers can use to put code together and develop systems.
33:57
SPEAKER_01: Um, and it basically gives you the option of which uh AI model you want to support that that coding and development work.
34:06
SPEAKER_01: Um Cursor is rolled out to 100% of my engineering team.
34:12
SPEAKER_01: Um I think that uh AI adoption has been the most uh most forthcoming really with the the engineers, and it it's a really good application of of AI because it's um you know code is language, right?
34:31
SPEAKER_01: Code is language, and code is typically available on computers, uh, and therefore it's very accessible and readable um by AI.
34:41
SPEAKER_01: Um so AI has become a phenomenal coder.
34:45
SPEAKER_01: Um you know, I maybe 70% of the code we write today is is not human written.
34:55
SPEAKER_01: Um, it still is human reviewer, and and remember that because that puts us at level three potentially of the framework I'm about to share.
35:04
SPEAKER_01: Um so uh there's a four-tier framework.
35:12
SPEAKER_01: Tier one is where you might speak to a chatbot, right?
35:18
SPEAKER_01: You know, it's a separate system.
35:21
SPEAKER_01: You uh you may have uh we'll call it assistive, right?
35:26
SPEAKER_01: You you may have a conversation, you may or you may not take that, you may copy and paste things across.
35:33
SPEAKER_01: Um level two.
35:34
SPEAKER_01: So um this is introducing agentic workflows.
35:40
SPEAKER_01: Um so this might be very, very, very narrow, right?
35:42
SPEAKER_01: Like you might literally have some automation um around how you're using it.
35:47
SPEAKER_01: Um ideally, the agent has some context though.
35:51
SPEAKER_01: It's not just like a sort of blind Zapier-based automation, it's it's something that is at least interpreting an inquiry, um, but is is very narrow.
36:00
SPEAKER_01: Um, level three is where you know every person has an agent with them, right?
36:06
SPEAKER_01: So all my developers have cursor, they have they have uh they're they're using AI.
36:11
SPEAKER_01: Um, but it but again it's still relatively narrow, but you've scaled it to a point that is roughly the same uh scale as your human workforce.
36:22
SPEAKER_01: Um now level four is a multi-agent autonomous workforce.
36:29
SPEAKER_01: Um so is particular audience doing this yet?
36:36
SPEAKER_01: No, we are not, right?
36:38
SPEAKER_01: Um, do we want to?
36:41
SPEAKER_01: Possibly we want to do this.
36:43
SPEAKER_01: Um, how would we tackle it?
36:45
SPEAKER_01: Um, the the best analogy to this is if you hired a graduate or someone straight out of school, so young, not especially professionally experienced, um, but they do have access to all the knowledge in the world, right?
37:02
SPEAKER_01: What would you let them do?
37:04
SPEAKER_01: Um, and I think that we have to think on this like multi-agent autonomous workforce in the sense of if we hired one, 10, 100 graduates tomorrow, what would we let them do?
37:20
SPEAKER_01: They probably start quite narrow individually.
37:23
SPEAKER_01: They probably have their own roles, they probably have limited access and permissions.
37:26
SPEAKER_01: And you'd have to orchestrate that workforce in the way that you would do with hiring a whole bunch of humans, right?
37:33
SPEAKER_01: There is no, I've just turned on AI and it runs my business now, um, unfortunately.
37:39
SPEAKER_01: Uh so I think that this, and by the way, level four has only been possible in the last three months, right?
37:47
SPEAKER_01: It's that new.
37:49
SPEAKER_01: Um, so it's something that we are knocking off things bit by bit.
37:54
SPEAKER_01: Well, can we do this?
37:55
SPEAKER_01: How can we do this?
37:56
SPEAKER_01: You know, I think it has to be a very sequenced, uh carefully done thing, and absolutely back to Cedric's point.
38:04
SPEAKER_01: Therefore, the human is still the bottleneck.
38:07
SPEAKER_01: Um, yeah, I'll pause that.
38:15
SPEAKER_02: Yeah, that framework is a very kind of practical, real way of thinking about AI today.
38:22
SPEAKER_02: Um I've I've just frantically made notes, James.
38:26
SPEAKER_01: Unfortunately, everyone should be in level three, and I think that if a if a listener to this podcast is is in a business and they're not at level three, um there's not really too much of an excuse as to why not, um, apart from big, sluggish, old businesses not having the transformation agility that they should have, perhaps.
38:53
SPEAKER_01: Lots of rules.
38:55
SPEAKER_01: But I would warrant that even in those uh large companies, individuals are still using AI at on the personal, on the personal level.
39:03
SPEAKER_02: So I think it's yeah, it's fascinating.
39:05
SPEAKER_02: And I'm thinking about, you know, kind of my own role and my own team and and how we how we make our way through the framework and where we sit.
39:13
SPEAKER_02: So it's a very practical way of thinking about it, and hopefully everyone will listen to this and take that away and take it back uh internally and um and do a kind of internal assessment, I guess, on the framework.
39:24
SPEAKER_02: Um we are we are kind of coming coming to the end, but just uh we always like to finish with um very quick fire round.
39:32
SPEAKER_02: Cedric, you kind of touched on this to be fair at the beginning, but the first question is how are you currently using AI?
39:38
SPEAKER_02: So, how do you use AI today?
39:40
SPEAKER_02: And maybe you know you could tell us where you are in this framework.
39:43
SPEAKER_02: Um, and what would you like AI to do in the future?
39:46
SPEAKER_02: Is it to get to level four here or is it to do to do more?
39:51
SPEAKER_02: Um, Cedric, do you wanna do you want to kick us off?
39:54
SPEAKER_00: Yeah, thank you.
39:55
SPEAKER_00: Um as I said, um I want to have my chief of staff or professional usage, but also for my personal usage.
40:04
SPEAKER_00: And um, James, I have also OpenClaw installed in my computer.
40:09
SPEAKER_00: So yes, uh, I want to practice to have the level four uh on my personal side.
40:14
SPEAKER_00: On the company side at Miracle, we are working actually on the level four.
40:19
SPEAKER_00: Um it's difficult because we need to link the dot uh because at each point some someone is doing automatically something, another something, but you you need to link, you need to evaluate if it's okay or not, and it's if it's scalable or not.
40:33
SPEAKER_00: So um, my ultimate goal is um with AI, uh, because I'm a AI enthusiast, as you see on my speech, is that I will not I don't want to spend time to uh due tes, but I want to spend time about my main objective, my main goals.
40:50
SPEAKER_00: And if uh ZI is able to uh help me to not spend time to uh writing an email and uh to uh to resume a discussion, I will win time to spend on the the the key things or on our life at the personal stage and also on the professional stage.
41:10
SPEAKER_02: Yeah, it's about that time efficiency again, isn't it?
41:13
SPEAKER_00: Which is yes, always because time is accelerating with AI, so we need to be more and more on the the what we want, what we need.
41:23
SPEAKER_02: Yeah, yeah, okay.
41:26
SPEAKER_01: Yeah, I think um yeah, I think just to say like almost exactly the same thing, but I guess to position it in like the opposite way of like, well, what am I not doing with AI?
41:40
SPEAKER_01: Um which which maybe is increasingly gonna be the the question.
41:45
SPEAKER_01: Um you know uh sales, right?
41:54
SPEAKER_01: Sales is is human.
41:57
SPEAKER_01: Um people buy from people.
42:01
SPEAKER_01: I used to have a very large team of business development representatives that would do prospecting and cold calling, and they would book uh six meetings a month per BD rep, and we would convert 12.5% of those meetings within six months for an average contract value of you know five, six figure uh contracts, right?
42:25
SPEAKER_01: It was a really nice predictive model.
42:27
SPEAKER_01: Um there was a huge proliferation of BD reps at tech companies with all the VC funding that happened through COVID, and that became a very congested channel.
42:39
SPEAKER_01: Um, much harder to get cut through.
42:42
SPEAKER_01: Everyone was getting cold called from 100 BDs a week, right?
42:46
SPEAKER_01: Now AI has almost made that problem uh worse because there have been companies that have spun up sales agents, and you know, I I get the worst uh outbounds that oh, I see you follow X venture capital company.
43:06
SPEAKER_01: I do too.
43:07
SPEAKER_01: Should we talk about business process transformation?
43:09
SPEAKER_01: It's like what are you, what are you like, what are you doing?
43:12
SPEAKER_01: It's and and it's so bad.
43:14
SPEAKER_01: And I think that that's only gonna push uh go to market.
43:19
SPEAKER_01: And this doesn't just apply to me as a tech vendor, I think it applies to retail media networks that they sell to their suppliers as well.
43:24
SPEAKER_01: You know, I really would shy away from any actual client, B2B client touch points with AI.
43:32
SPEAKER_01: You know, I think that is more and more human.
43:34
SPEAKER_01: We're investing more in events and being physically present and participating in uh you know industry bodies like IAB than we ever have done before because the human relationship is just so, so, so important for that.
43:50
SPEAKER_01: Um, I I mean it when I say I am not looking at firing any of my workforce because of AI either, right?
44:02
SPEAKER_01: I truly believe they are positively uh augmented.
44:06
SPEAKER_01: They are, they are, it's a compound win, right?
44:09
SPEAKER_01: They are that much more valuable in their roles with AI.
44:14
SPEAKER_01: And I'm not alone in this.
44:16
SPEAKER_01: You know, I think some of the big tech companies that have let people go have sort of almost used it as a market signaling and we're we're so forward and actually they've overhired in the past and unfortunately they needed to reduce their headcount.
44:29
SPEAKER_01: I really like uh lovable the company out of Sweden.
44:32
SPEAKER_01: They were asked they're valued more than Klana now um AI uh basically website developer that you can work with it's really good you should try it out.
44:43
SPEAKER_01: Um they were asked in the press after raising all of this money um how come you if you're an AI company why's your head count correct and the founders' response was was very simple it was well the more people I have using AI the more productive we're gonna be as a business right and that's that's perfectly correlated.
45:03
SPEAKER_01: So I hope that that compound um multiplier starts to increase more and more with AI in a way that's very safe right much like Miracle particular audience is powering many many many large retail websites we cannot afford for something to break in production right so there are very real constraints on where we can apply probabilistic technology like AI as well so I understand there are constraints I love the human to human thing and I don't think that's going to go away um and I think that we can use AI to make us all uh probably more time rich take Cedric's um conclusion I I think that is you've both summed it up extremely well so I've taken you know time is accelerating with AI you're right it is I kind of took away a formula James more people multiplied by more AI equals more productivity and I think that that is something we should all all um bear in mind and I think the other thing you know listeners should take away and think about is what am I or what is my business or what is my team not currently doing with AI um bearing in mind the need for human interaction that in actual fact the increased need for human interaction and communication you know Cedric and I have both been on training courses workshops about communication this week how can we use AI to give us more time to make those human connections and that communication process better in actual fact so I think there's there's so much opportunity both with using AI and the and the time and the efficiencies that gives us to do other important aspects of of our roles um we've talked a bit about you know the use cases within retail and commerce media yield management using AI to you know uh better predictability in terms of shifting stock um and all of those things so I think we've had a bit of a reality reality check as well as to where AI is currently kind of um helping us um some of the blockers it's been a really insightful discussion particularly for me I've learned an awful lot you're both experts in this so thank you so much for taking the time to be on our podcast and and speak to me and and our listeners and and um yeah hopefully see you both again soon so thank you very much.
47:35
SPEAKER_02: Thanks for having us thank you thank you for tuning in to the Retail Media Roundtable we'll be back soon with another episode so make sure to subscribe to stay updated.
47:49
SPEAKER_02: In the meantime you can visit ivyp.eu to explore the latest retail media resources and insights to help you navigate this dynamic space