Unlocking Insights with the Amazon Marketing Cloud with Robert McGovern

[00:00:00] [SPEAKER_02]: Hello and welcome to eCom Insights, the podcast where we give you data-led tips to optimize your profits online.

[00:00:11] [SPEAKER_01]: Hello, welcome to eCom Insights listeners and I'm Matt and joining me is Rial, my co-host. Hey, Matt, good to be back.

[00:00:21] [SPEAKER_01]: And we've got a special guest I'm super pumped about today is Robert McGovern.

[00:00:26] [SPEAKER_01]: He's our head of paid and director of advertising and we're going to be talking about the exciting world of

[00:00:32] [SPEAKER_01]: the Amazon Marketing Cloud and also Amazon advertising, no doubt. How's it going, Rob?

[00:00:37] [SPEAKER_00]: Yeah, very good. Thank you. Happy to be here.

[00:00:40] [SPEAKER_01]: Well, I think just following the format of the previous episode,

[00:00:43] [SPEAKER_01]: Rial and I sort of introduced ourselves to the listeners. I think it's just useful to go over a

[00:00:48] [SPEAKER_01]: bit of your background. So let's start off with how long have you been in the world of

[00:00:51] [SPEAKER_00]: the wonderful world of Amazon advertising? Well, the world of advertising in general,

[00:00:56] [SPEAKER_00]: I've been, I actually started in university. So when I was in uni, I was president of

[00:01:04] [SPEAKER_00]: and ran a comedy society. And from that, I actually got a job with Avalon, which are like

[00:01:10] [SPEAKER_00]: a big comedy production company. And I flired for their Edinburgh Fringe. So I went and

[00:01:17] [SPEAKER_00]: lived in Edinburgh for a month, was out every day, flaring for like all these big

[00:01:26] [SPEAKER_00]: people like Russell Brand and Al Murray, you know, really big names. So I was out

[00:01:31] [SPEAKER_00]: flaring for all of those, did really well and was invited back the next year and sort of led

[00:01:36] [SPEAKER_00]: and managed their whole flaring and marketing team for the next year. So yeah, been in

[00:01:42] [SPEAKER_00]: advertising since during university, which is possibly a bit of an early start.

[00:01:47] [SPEAKER_02]: I was wondering how you were going to connect advertising to comedy clubs, but nicely done

[00:01:52] [SPEAKER_02]: there.

[00:01:53] [SPEAKER_02]: Oh, yeah, absolutely.

[00:01:57] [SPEAKER_00]: I also ran events and marketed the event. So, you know, there's multiple things as well.

[00:02:03] [SPEAKER_02]: So Rob, I know you come from pretty quantitative background, pretty much as

[00:02:07] [SPEAKER_02]: quantitative as they get in terms of physics. So it'd be interesting to hear, you know,

[00:02:12] [SPEAKER_02]: mixing in the physics component and how that's helped you develop, I guess, on the

[00:02:17] [SPEAKER_02]: advertising side and what's been useful there, if anything.

[00:02:20] [SPEAKER_00]: Yeah, so I mean, I think a lot of people think of marketing as, you know, it's kind

[00:02:25] [SPEAKER_00]: of moved on from that kind of mad men, John Ham days of like, I'm gonna sell cook, like,

[00:02:31] [SPEAKER_00]: you know, this is very different. You know, now digital marketing is all very sort of heavily

[00:02:38] [SPEAKER_00]: statistics focused, you know, it's about reaching the right person at the right time with

[00:02:41] [SPEAKER_00]: the right message. And there's just reams and reams of data available. And I feel like if

[00:02:48] [SPEAKER_00]: you don't have a strong sort of mathematical or scientific backing to that, it's quite

[00:02:54] [SPEAKER_00]: easy to get lost in this, you know, unless you're really focused on what you're doing.

[00:02:58] [SPEAKER_00]: So yeah, so I have a Masters in Physics and that gave me a really strong grounding

[00:03:04] [SPEAKER_00]: in sort of the scientific method and analysis. And, you know, as I mentioned,

[00:03:10] [SPEAKER_00]: I did comedy, so I wanted some form of creativity in my life. And

[00:03:14] [SPEAKER_00]: yeah, digital marketing seemed to be like a perfect blend of that.

[00:03:18] [SPEAKER_01]: But I guess what was the first thing that drew you to eCommerce advertising then? What was that

[00:03:25] [SPEAKER_00]: light bulb moment for you? I was at a previous agency which did digital marketing more

[00:03:31] [SPEAKER_00]: generally, so, you know, Google and Facebook. And just after that had grown and I developed

[00:03:37] [SPEAKER_00]: quite a bit, essentially just wanted the new challenge. When I found Amazon and yeah,

[00:03:45] [SPEAKER_00]: it's been great. It's, you know, it's completely there's elements of the sort

[00:03:52] [SPEAKER_00]: of wider cross channel and a lot of the same strategies can be applied across, but it's also

[00:03:57] [SPEAKER_00]: very different in terms of the levers that you have available to pull and also the

[00:04:02] [SPEAKER_00]: direct relationship that Amazon has with its customers can often lead to very different sets

[00:04:08] [SPEAKER_02]: of data being available. Yeah, and I think the evolution of Amazon stayed relatively the

[00:04:14] [SPEAKER_02]: same for a while meaning sponsored product heavy. But really, you know, an explosion in the last

[00:04:19] [SPEAKER_02]: two years or so maybe a little bit longer, really, really investing in different ad products.

[00:04:24] [SPEAKER_02]: I mean, clearly they've got the infrastructure to do it plus the data sets to do it. And

[00:04:28] [SPEAKER_02]: that's always a good recipe for in theory, at least, you know, good advertising products. So

[00:04:33] [SPEAKER_02]: it's good to ride the wave. So I guess moving on to the Amazon marketing cloud.

[00:04:41] [SPEAKER_02]: First things first, you know, what is the AMC, the Amazon marketing cloud, given my

[00:04:47] [SPEAKER_02]: discussions with customers, it's the awareness just doesn't seem to be high. Right. And I'm

[00:04:51] [SPEAKER_02]: talking pretty sophisticated customers that have been on Amazon five, seven years running pretty

[00:04:57] [SPEAKER_02]: large Amazon businesses, you know, 20 30 million north of that even. And, you know,

[00:05:03] [SPEAKER_02]: sort of the thing that people say I've kind of heard a little bit about but don't really

[00:05:05] [SPEAKER_02]: understand. So Rob, could you help everybody please understand what the AMC actually is?

[00:05:11] [SPEAKER_00]: The AMC, as I said, the Amazon marketing cloud is essentially a tool for data measurement. It's

[00:05:20] [SPEAKER_00]: a way of pulling in signals from your, you know, your seller vendor account, your

[00:05:27] [SPEAKER_00]: advertising account and your DSP, which is demand side platform advertising account as

[00:05:35] [SPEAKER_00]: well. It's a way for you to understand what kind of customers that you're having, how they kind of

[00:05:47] [SPEAKER_00]: interlink and the paths that they're taking through your kind of customer buying journey.

[00:05:53] [SPEAKER_00]: And it's really just, you know, it's a tool, it's a data clean room that you can use to

[00:06:01] [SPEAKER_00]: to, you know, parse through these insights and pull out some actionable insights from what your

[00:06:09] [SPEAKER_00]: customers are doing, which can help instruct your marketing strategy or, you know, even other

[00:06:14] [SPEAKER_00]: things like your product strategy or, you know, I guess it is still marketing but like your

[00:06:18] [SPEAKER_00]: customer communication strategy, things like that, how you want to position yourself. You know,

[00:06:23] [SPEAKER_00]: a great way to do that with access to data that Amazon doesn't typically give out as easily with,

[00:06:32] [SPEAKER_00]: you know, direct sort of customer insights. There is a few caveats, you know, for example,

[00:06:39] [SPEAKER_00]: there are still data privacy laws that protect customers. So it does need to be like large

[00:06:44] [SPEAKER_00]: enough groups you can drill down on one individual and follow them around Amazon.

[00:06:50] [SPEAKER_00]: I think it needs a minimum of 100,000 depending on the query.

[00:06:54] [SPEAKER_01]: You mentioned data privacy as a point. I think, you know, that is potentially a game changer for

[00:07:02] [SPEAKER_01]: Amazon, Amazon marketing cloud. Could you just give this some context as to, you know,

[00:07:08] [SPEAKER_01]: mention two points. One about the paths to purchase which, you know, if I'm selling on

[00:07:14] [SPEAKER_01]: Shopify with the Google Analytics account and the various tools you'll actually have

[00:07:18] [SPEAKER_01]: had access to that data for quite a while now. But in contrast to Amazon, what didn't we have

[00:07:24] [SPEAKER_01]: before? How has it changed the game on those two points really? Like a lot of what it does

[00:07:30] [SPEAKER_00]: give you now is, you know, things like path to purchase, things like where, you know,

[00:07:36] [SPEAKER_00]: what different signals that customers wear, what different, you know, touch points they

[00:07:42] [SPEAKER_00]: get across your advertising journey. You know, it's something that, you know,

[00:07:49] [SPEAKER_00]: compared to what was available or what is available still on things like the AMC,

[00:07:54] [SPEAKER_00]: the AMS rather, there's only just this advertising campaign or keyword spent this amount

[00:08:01] [SPEAKER_00]: and got this return on a very limited sort of view where you have to trust the attribution

[00:08:07] [SPEAKER_00]: that Amazon's giving you. Whereas the AMC allows you to look at far more things, you know,

[00:08:14] [SPEAKER_00]: for a couple of examples I can pull out. Did someone touch multiple advertising touch points

[00:08:21] [SPEAKER_00]: in their journey? AMC allows you to look at that. Did they look at what was the order

[00:08:26] [SPEAKER_00]: in which they looked at them? Were they looking at sponsored products, then a sponsored

[00:08:29] [SPEAKER_00]: brand, then a sponsored display, back to sponsored products? You can look at that.

[00:08:34] [SPEAKER_00]: Another one is, you know, say that you have a very large purchase. You know,

[00:08:42] [SPEAKER_00]: you're a purchase which is a couple of hundred pounds. That's not a decision

[00:08:45] [SPEAKER_00]: you make quite quickly. You know, that's a decision that takes a while. In the AMS,

[00:08:51] [SPEAKER_00]: in the traditional advertising on Amazon, you can only look at what has influenced them

[00:08:57] [SPEAKER_00]: in the past 14 days. In the AMC, you can define these look back windows. So if you know

[00:09:03] [SPEAKER_00]: it's going to take a lot longer, that someone's going to start having their decision influence,

[00:09:09] [SPEAKER_00]: you know, someone's buying a 200, 300, 400 pound item on Amazon. They're not making

[00:09:15] [SPEAKER_00]: that decision in two weeks or, you know, not to speak for everyone, I'm certainly not

[00:09:18] [SPEAKER_00]: making that decision in two weeks. You know, that's a decision that I take, you know,

[00:09:24] [SPEAKER_00]: a couple of months out to go, is this the right object for me? If I'm buying like a

[00:09:27] [SPEAKER_00]: high end laptop or something, you know, I'm not making that decision quickly. I'm browsing,

[00:09:32] [SPEAKER_00]: I'm checking and comparing. And I want to know if I'm advertising that high end item,

[00:09:38] [SPEAKER_00]: I want to know what are people looking at? Yes, I think we'll come back to that a little

[00:09:43] [SPEAKER_02]: bit later. I think maybe just taking a bit of a step back, how do you access the Amazon

[00:09:47] [SPEAKER_02]: marketing cloud? Is this open to anyone who's advertising on Amazon? What sort of conditions

[00:09:51] [SPEAKER_00]: need to be met? So at the moment, you do need to have a DSP account. So something that

[00:09:58] [SPEAKER_00]: needs to be running DSP, which is demand side platform, which is programmatically buying ads.

[00:10:08] [SPEAKER_00]: And as long as you are running some activity on there, you should be able to reach out

[00:10:13] [SPEAKER_00]: to an Amazon representative who will set up a separate AMC instance for you where you can

[00:10:22] [SPEAKER_00]: connect multiple accounts. So this also can stretch over multiple regions, for example. You

[00:10:30] [SPEAKER_00]: can have a single instance that represents your entire EU. For when we set up personally,

[00:10:36] [SPEAKER_00]: I make sure that we set up our campaigns with country identifiers as well, even when they are

[00:10:42] [SPEAKER_00]: in different accounts, simply so that we can also break out those insights and restrict

[00:10:47] [SPEAKER_00]: them to individual countries, as well as looking at pan-EU countries as well.

[00:10:51] [SPEAKER_02]: So I think a related question, who's using the Amazon marketing cloud? Is it a daily tool? Is

[00:10:56] [SPEAKER_02]: it a monthly tool? Are you yourself using it or is it the data team? Because one of the things

[00:11:03] [SPEAKER_02]: that I was maybe a little bit confused by with the go-to markets, this is maybe 18 months ago,

[00:11:07] [SPEAKER_02]: two years ago, with the AMC. It was pretty much built for people who knew how to do

[00:11:13] [SPEAKER_02]: elementary coding. In order to do anything, you basically had to write SQL and ask very

[00:11:20] [SPEAKER_02]: specific questions. They would basically give you the data. It's kind of an infrastructure go-to

[00:11:24] [SPEAKER_02]: market, which is here's a bunch of data. Here's what the definition is of the data,

[00:11:28] [SPEAKER_02]: but it's on you to... You go figure it out.

[00:11:32] [SPEAKER_02]: Construct the query, right? To ask the right question to get the result. That's just very

[00:11:39] [SPEAKER_02]: limited, right? How many sort of organizations and agencies or whoever it is are in a

[00:11:43] [SPEAKER_02]: position ready to do that. It's changed a lot, right? I think it's become a lot more

[00:11:48] [SPEAKER_02]: user-friendly. Is it the data team? Is it interesting to hear who's using it and

[00:11:54] [SPEAKER_00]: I guess how frequently? I would say that whilst it has become a bit friendlier,

[00:11:59] [SPEAKER_00]: I still think there is need for some kind of data team resource for it. Unless you have

[00:12:07] [SPEAKER_00]: a partner who is able to run those queries for you and put them into some kind of

[00:12:12] [SPEAKER_00]: visualization for you, then you do still need... This would be like a bit management

[00:12:15] [SPEAKER_00]: platform or something like that. Yeah, exactly. Something like that. Unless you have that

[00:12:22] [SPEAKER_00]: linked into your AMC instance, in order for you to write any queries, you will need to have

[00:12:31] [SPEAKER_00]: at least some understanding of the principles of coding. I guess to kind of be transparent,

[00:12:38] [SPEAKER_00]: I'm not a coder. That's not my background, but as we mentioned at the start, I have done

[00:12:45] [SPEAKER_00]: physics and part of that I did some very rudimentary sort of C++ and... Or short.

[00:12:54] [SPEAKER_00]: Long time ago, over 10 years, can't remember. But I did some kind of visualization tools

[00:12:59] [SPEAKER_00]: and I had to run the code so just understand how the code ran rather than being able to

[00:13:04] [SPEAKER_00]: build it myself. So whilst I've got that kind of background, I can use a lot of the... They

[00:13:11] [SPEAKER_00]: example queries. If you want to run a query which pulls this data, then here's what the

[00:13:18] [SPEAKER_00]: full query is already built for you and it sort of gives you comments that tell you

[00:13:23] [SPEAKER_00]: here's where you want to change this bit if you want to look back further or here's

[00:13:27] [SPEAKER_00]: where you put in the campaigns that you want to look at, so on and so forth. So

[00:13:30] [SPEAKER_00]: for someone like me who understands how code works but couldn't write it,

[00:13:34] [SPEAKER_00]: that's enough for me to be able to build out a few instructional queries and be able

[00:13:40] [SPEAKER_00]: to pull things. But I do think you fundamentally do need someone with SQL experience who can

[00:13:47] [SPEAKER_00]: in SQL experience who can come in and troubleshoot for you. I don't think it's something that

[00:13:52] [SPEAKER_00]: someone with zero coding experience would be able to just do straight off.

[00:13:56] [SPEAKER_02]: Yeah, so it sounds like for like 90% or even more, 90-80% of people who want to use this,

[00:14:01] [SPEAKER_02]: they would have to rely on a partner to effectively do this and properly productize it.

[00:14:05] [SPEAKER_01]: It seems to be a low, we mentioned the low adoption rate and certainly at the last

[00:14:11] [SPEAKER_01]: unbox in London with Amazon, AMC was everyone was talking about it but you sort of scratched

[00:14:18] [SPEAKER_01]: the surface of actually how well are you adopting it and what to extend. It was a bit

[00:14:24] [SPEAKER_01]: opaque for these reasons that you need as a brand and I'm thinking from like a brand's

[00:14:28] [SPEAKER_01]: OK, I need a DSPC or a DSP partner. I need a team that's got some rudimentary coding capability

[00:14:36] [SPEAKER_01]: and then I need a team, the data team to make understand the vast amount of data that's in

[00:14:42] [SPEAKER_01]: Amazon and make sense of it. There's a lot of barriers to adoption but it seems very

[00:14:47] [SPEAKER_01]: valuable. Could you give us some sort of sense about maybe some insights or

[00:14:53] [SPEAKER_01]: the value that AMC can surface? Let's talk use cases.

[00:14:59] [SPEAKER_00]: Yeah, absolutely. I've been raring to chat about this one as you both know me, you both know

[00:15:06] [SPEAKER_00]: my favorite use case that I do not stop talking about which is custom attribution is I think

[00:15:15] [SPEAKER_00]: the biggest one for me. It's really where I think like there's a lot of the future where

[00:15:22] [SPEAKER_00]: advertising needs to go to be really actionable. The analogy that I like to use on this is

[00:15:36] [SPEAKER_00]: to explain why what's existing in the platform which is last touch where all of the sales get

[00:15:42] [SPEAKER_00]: attributed to the last advertising touch point within those last 14 days gets all of

[00:15:49] [SPEAKER_00]: the credit for the sale. The analogy that I like to use for this is if you go out for a night,

[00:15:56] [SPEAKER_00]: you start off, you have a glass of wine, you have a couple of glasses of, you have a couple

[00:16:00] [SPEAKER_00]: of pints of beer or cider which every one and then before you go to bed you think

[00:16:05] [SPEAKER_00]: I'll be sensible, I'll have a large glass of water, that will help me. In the morning you

[00:16:11] [SPEAKER_00]: wake up, you've got a hangover. Which drink caused the hangover?

[00:16:17] [SPEAKER_00]: Amazon's default attribution says the glass of water. That's why you have a hangover because

[00:16:21] [SPEAKER_00]: you drank that glass of water because that was the last thing you did before you got your

[00:16:26] [SPEAKER_00]: hangover. We all know that that's not it. It's a mixture of all the things you did

[00:16:33] [SPEAKER_01]: across the full night. It could be real getting the gin and tonics in at one point I think.

[00:16:39] [SPEAKER_00]: Exactly. We struggle to remember. But that's what AMC allows you to do. There's custom

[00:16:51] [SPEAKER_00]: attribution models where you can look at every single touch point across a journey or you can

[00:16:57] [SPEAKER_00]: weight it to what was the first thing they did and what was the last thing they did.

[00:17:01] [SPEAKER_00]: You can weight it according to different options and look at these different things.

[00:17:08] [SPEAKER_00]: So that allows you to understand a bit more about what is actually causing these things,

[00:17:21] [SPEAKER_00]: what's actually causing your customers to buy your products.

[00:17:27] [SPEAKER_00]: We had a client before in the garden tool space selling 15 million to 20 million a year.

[00:17:40] [SPEAKER_00]: When we did an attribution, they had very, very little DSP spent so there wasn't really

[00:17:48] [SPEAKER_00]: enough for it to skew anything. But looking at sponsored products, sponsored brands,

[00:17:52] [SPEAKER_00]: sponsored display on the AMS platform, we found that when we compared it to some of

[00:17:58] [SPEAKER_00]: the custom attribution models we made, the platform over reported sales to sponsored products

[00:18:04] [SPEAKER_00]: by 20 to 30%. It's a big jump. So we were able to use that to shift the budgets around.

[00:18:14] [SPEAKER_00]: When we looked at the return that we were getting in some of these sponsored product

[00:18:20] [SPEAKER_00]: campaigns to shuffle that around and understand that the return was not as great. In fact,

[00:18:25] [SPEAKER_00]: there were some very upper funnel activity sponsored display campaigns that we had

[00:18:30] [SPEAKER_00]: which were reporting a zero return on investment in the platform. But when we went into AMC and

[00:18:37] [SPEAKER_00]: we looked at it over a longer period and looked at it over a custom attribution,

[00:18:41] [SPEAKER_00]: they were running profitably. So this was something that you would look at, you would say,

[00:18:47] [SPEAKER_00]: no, that's not delivering anything, let's switch that off. But when you look at the impact it's

[00:18:52] [SPEAKER_00]: had across multiple other campaigns that are just claiming it for themselves, it's actually

[00:18:59] [SPEAKER_00]: delivering and it's something that should be running. I think it's important for the

[00:19:04] [SPEAKER_01]: listeners to just reiterate how the marketing funnel works and AMC allows you to surface these

[00:19:11] [SPEAKER_01]: insights. So you can imagine you're running a significant investment in the top of funnel

[00:19:16] [SPEAKER_01]: campaign awareness using DSP, for example. And then I've had an instance with a large

[00:19:23] [SPEAKER_01]: pet supplement brand that they were looking at the full funnel again when actually

[00:19:28] [SPEAKER_01]: the return on the awareness top of the funnel of DSP is not looking great, ones and twos.

[00:19:35] [SPEAKER_01]: But look at how well sponsored products are doing guys, look at all these sales.

[00:19:40] [SPEAKER_01]: So we worked out that DSP had increased the brand research by 155% because obviously

[00:19:47] [SPEAKER_01]: people are seeing the brand, they're seeing video with the brand, they go,

[00:19:51] [SPEAKER_01]: oh, I'll search for X brand, you know, or similar brand basically. So, you know,

[00:19:56] [SPEAKER_01]: your brand research will increase because your awareness has increased. So that's why

[00:20:00] [SPEAKER_01]: AMC is really important to show that because clients will always go to, well,

[00:20:04] [SPEAKER_01]: that's the biggest rise number, let's put all our budget there sort of thing. But that's not

[00:20:08] [SPEAKER_02]: the marketing funnel. I guess I'm curious, Rob, like how does the discussion with the client go,

[00:20:12] [SPEAKER_02]: which is, hey, all these things you've been reporting for, I don't know, the last six months,

[00:20:17] [SPEAKER_02]: we're now using a tool that gives us better insights and turns out it's quite a different

[00:20:21] [SPEAKER_02]: picture here. That's, I presume, quite a difficult conversation or not? Well, I mean,

[00:20:27] [SPEAKER_00]: from my perspective, we've been preceding that conversation from the start, you know,

[00:20:31] [SPEAKER_00]: like, you know, shout out to the absolutely fantastic team that we have here. And they,

[00:20:38] [SPEAKER_00]: you know, they live and breathe advertising. Compliments will get you everywhere,

[00:20:41] [SPEAKER_00]: won't they, where we are? But they, you know, they live and breathe, you know,

[00:20:48] [SPEAKER_00]: especially in the advertising team, they completely live and breathe advertising. So they

[00:20:52] [SPEAKER_00]: understand everything that you were saying there, Matt, about like the full funnel

[00:20:56] [SPEAKER_00]: and understanding. So, you know, for example, that campaign that I mentioned in the previous

[00:21:01] [SPEAKER_00]: example that had said zero, we had been running it because we had already seen less

[00:21:08] [SPEAKER_00]: definitive but still impacts there of what it was providing for, you know, for the kind

[00:21:13] [SPEAKER_00]: of awareness and those other pieces and, you know, that it was delivering a great number

[00:21:17] [SPEAKER_00]: of impressions and clicks and people coming through. So, you know, the team are aware of

[00:21:25] [SPEAKER_00]: these things and they communicate and they want to be looking at things about, you know,

[00:21:30] [SPEAKER_00]: they are saying, yes, we understand these ROAS is very high. However, like when you look at it

[00:21:36] [SPEAKER_00]: holistically, when you look at what all these things are doing. And then by doing that,

[00:21:43] [SPEAKER_00]: we're able to get to the point where we can run these tests with a greater sort of more

[00:21:50] [SPEAKER_00]: impressive data set through something like the ANC. And it's not a case of, oh, well,

[00:21:55] [SPEAKER_00]: everything that we were doing before is wrong, but a case of, oh, everything you've been saying

[00:22:00] [SPEAKER_00]: was right. And that becomes to be that's a much nicer conversation.

[00:22:06] [SPEAKER_02]: Although I can envisage cases where that's probably not the case. But yeah. So if I

[00:22:11] [SPEAKER_02]: understand correctly here, like we're saying or you're saying that the using the Amazon

[00:22:17] [SPEAKER_02]: marketing cloud can actually change the overall ROAS of, you know, all the aggregated campaigns

[00:22:23] [SPEAKER_02]: or of the accounts, but also will change the individual ones. Let's say associated with

[00:22:29] [SPEAKER_02]: whatever the product is, sponsor brand, sponsor display, et cetera, et cetera. So it's,

[00:22:33] [SPEAKER_02]: it's, yeah, it's a pretty fundamental difference here. This is a key, obviously metric,

[00:22:36] [SPEAKER_02]: obviously a key metric around how you gauge efficiency and all those sort of things.

[00:22:42] [SPEAKER_02]: And this is like a headline number that's actually can materially change

[00:22:46] [SPEAKER_02]: via the Amazon marketing cloud insights. So this is a big deal.

[00:22:49] [SPEAKER_02]: Yeah, absolutely. Yeah.

[00:22:51] [SPEAKER_01]: Interesting. I think we've covered one or two use cases for AMC already about being able to

[00:22:58] [SPEAKER_01]: attribute activity all the way through the funnel. Last touch attribution, is that correct?

[00:23:05] [SPEAKER_01]: Or, you know, where, where, where should I put my money? But is there any other use cases

[00:23:09] [SPEAKER_01]: you know, available right now that brands should look at on Amazon marketing cloud?

[00:23:16] [SPEAKER_00]: To continue on the vein that I'd already been mentioning about the,

[00:23:23] [SPEAKER_00]: I mean the custom attribution and then we'll move on to some other ones so that I'm not

[00:23:26] [SPEAKER_00]: just banging the same drum.

[00:23:29] [SPEAKER_00]: To be fair, it's a big deal, right? What you're saying is a big deal.

[00:23:34] [SPEAKER_01]: Yes, definitely.

[00:23:36] [SPEAKER_00]: But there's some other ones as well. So, you know, looking at things like

[00:23:42] [SPEAKER_00]: path to conversion. So not just sort of attributing it across a full thing, but

[00:23:49] [SPEAKER_00]: specifically what groups of campaigns are my customers moving through? Which can give you

[00:23:56] [SPEAKER_00]: some ideas of, you know, what, depending on how you've segmented your campaigns, you know,

[00:24:02] [SPEAKER_00]: it can either give you an idea of, you know, what campaigns are working really well together.

[00:24:06] [SPEAKER_00]: Again, it can give you this idea of how do I shuffle the budget around? But it can also

[00:24:10] [SPEAKER_00]: give you an idea of like, what communication is working for these customers? If a customer

[00:24:18] [SPEAKER_00]: is coming in and looking at one product, what do they actually end up going and buying at

[00:24:24] [SPEAKER_00]: the end of that journey? Like, what is a customer searching for? If I communicate

[00:24:28] [SPEAKER_00]: one of my USPs to this group of customers, how do they then track that along? Are they looking

[00:24:35] [SPEAKER_00]: at the same, you know, are my customers very neatly defined and that the ads they click on

[00:24:40] [SPEAKER_00]: all have a similar messaging or are there overlaps where I'm messaging with one message

[00:24:47] [SPEAKER_00]: to one group of customers and then move on and come into another. So it's also a good

[00:24:53] [SPEAKER_01]: thing about customer profiles. Yeah, absolutely. So I think when you mentioned about customer

[00:24:57] [SPEAKER_01]: profiles, I'm just saying that I've put my CMO on Chief Marketing Officer because now they've

[00:25:02] [SPEAKER_01]: lent into the podcast. Hang on a minute. You can actually isolate what types of customers

[00:25:07] [SPEAKER_01]: are interacting with which creative and products? Yeah, yeah. Is that correct?

[00:25:13] [SPEAKER_00]: Yeah. So you can, if you've set up your campaigns to do that, if you've set up

[00:25:18] [SPEAKER_00]: your campaigns in a way that you are, you know, messaging, having specific messaging,

[00:25:25] [SPEAKER_00]: then yeah, there would be a way to look at what are people clustered into specific messages or

[00:25:34] [SPEAKER_00]: are they more sort of broadly going across or are there like specific overlaps? Yeah,

[00:25:41] [SPEAKER_00]: you need to make sure that you're setting up your campaigns in a way to support this as well

[00:25:45] [SPEAKER_00]: in the EMC as well. But absolutely, that's the best case that I can see for it.

[00:25:50] [SPEAKER_01]: That's super powerful for me. I think because, you know, the listeners can imagine that you

[00:25:56] [SPEAKER_01]: have a billboard that you can change the message on but you can just show to specific demographics

[00:26:02] [SPEAKER_01]: or people at a certain time to really work out which message is responding best to

[00:26:07] [SPEAKER_01]: mail or email, for example, or older demographic, younger demographic. Yeah, it's definitely one to,

[00:26:13] [SPEAKER_00]: if you're wanting to do something like that, it's definitely one that I'd be doing with

[00:26:16] [SPEAKER_00]: a strong DSP campaign where it will be even easier to set up these things. We have a lot of

[00:26:23] [SPEAKER_00]: testing flows that we've developed to run sort of messaging and customer profile tests

[00:26:28] [SPEAKER_00]: and things like that. But yeah, EMC allows you to sort of supercharge that and look at it.

[00:26:35] [SPEAKER_00]: You know, you can run your main test in the DSP and then look into how is that pulling

[00:26:40] [SPEAKER_00]: through to the EMS. What's the path after we've already created this customer profile in the DSP?

[00:26:47] [SPEAKER_00]: What products do they come in and look at when they're on Amazon directly?

[00:26:52] [SPEAKER_01]: Yeah, I think from a previous podcast I talked about some of my marketing background and it

[00:26:57] [SPEAKER_01]: just chimes with me that the amount of meetings I had from my marketing campaigns

[00:27:01] [SPEAKER_01]: and you'd have the brand manager or the CMO go, well let's see, yes our target audience

[00:27:05] [SPEAKER_01]: is 25 to 45 year old women and they like high heels and fashion and driving cars and

[00:27:15] [SPEAKER_01]: might have you in the go. And then you realise actually from the data, it's some completely

[00:27:19] [SPEAKER_01]: different data set. So I had something similar that that homeware brand that I mentioned,

[00:27:24] [SPEAKER_01]: the target demographic was actually 56 year old women, not 40, 35 to 45, you know.

[00:27:30] [SPEAKER_01]: You'll actually be able to play that data back.

[00:27:33] [SPEAKER_00]: The number of times I've gone, yeah here's who you think your customer is and here's

[00:27:38] [SPEAKER_00]: who's actually buying, you know, that level of...

[00:27:41] [SPEAKER_00]: So that's super powerful.

[00:27:43] [SPEAKER_00]: Yeah, there's some other ones as well. So overlap analysis is a really strong one.

[00:27:50] [SPEAKER_00]: So again more specifically looking at what's the efficiency if they look at one ad

[00:27:56] [SPEAKER_00]: versus another ad versus both of those ads over their journey. You know, we had one client

[00:28:06] [SPEAKER_00]: who in the sort of hair and beauty sector and they saw 500% increase in efficiency

[00:28:16] [SPEAKER_00]: with DSP and AMS working combined compared to AMS working by itself. So we saw some absolutely

[00:28:25] [SPEAKER_00]: astonishing increase in efficiency just by having that awareness off site. So when they got onto

[00:28:31] [SPEAKER_00]: Amazon, they were already warmed and brand.

[00:28:36] [SPEAKER_02]: So just to make sure I understand this, you're saying someone who was exposed to,

[00:28:39] [SPEAKER_02]: let's say, just a particular ad type, one ad type, let's call it sponsored products for

[00:28:43] [SPEAKER_02]: instance, versus somebody who was exposed to both DSP and the sponsored product,

[00:28:49] [SPEAKER_02]: that sort of almost incrementality, is that the 500% that you're referring to?

[00:28:54] [SPEAKER_00]: So it was that they were five times more likely to convert. So their conversion rate went up by

[00:28:59] [SPEAKER_00]: five or by six, rather 500%.

[00:29:03] [SPEAKER_01]: Really important marketing principles, the more a consumer sees a brand, the more likely they are

[00:29:08] [SPEAKER_01]: to buy. I think there's various stats out there like Forester is saying it's like 27...

[00:29:13] [SPEAKER_01]: a consumer has to see a brand 27 times before they consider buying it, whether that's

[00:29:17] [SPEAKER_01]: TV, DSP, whatever.

[00:29:20] [SPEAKER_00]: The lower end I've heard is seven. That's the very lower end, so I think that might be one of them.

[00:29:28] [SPEAKER_01]: That's the standard one, the seven times. But yeah, I mean, the modern consumers,

[00:29:32] [SPEAKER_01]: it's in the 20s. I don't know that much.

[00:29:34] [SPEAKER_02]: It's one of those things as an individual consumer you would just deny, right?

[00:29:39] [SPEAKER_02]: Yes.

[00:29:40] [SPEAKER_02]: Come on, that's not true. I'm not that, I don't know, affected by these things.

[00:29:45] [SPEAKER_00]: So there's a psychological principle, it's called the mere exposure effect. And

[00:29:49] [SPEAKER_00]: that's it, is just the more times you see something, yeah. And I think the mere

[00:29:54] [SPEAKER_00]: exposure effect is the one where the seven times comes from, which is like,

[00:29:58] [SPEAKER_00]: that you've seen it seven times and at that point you can start measuring an

[00:30:01] [SPEAKER_00]: increase in familiarity. So it might not be enough to make them to buy, which might

[00:30:05] [SPEAKER_00]: be where the 27 comes in, but it's enough that seven times is when they've started

[00:30:09] [SPEAKER_00]: elevating above something they haven't seen seven times, with no reason other than,

[00:30:14] [SPEAKER_00]: I've seen it.

[00:30:16] [SPEAKER_00]: Like, that's it.

[00:30:18] [SPEAKER_00]: Yeah, our brains are very strange. There's one more one that I just wanted to call out

[00:30:25] [SPEAKER_00]: specifically because I think it's a very important one. We kind of touched on it

[00:30:28] [SPEAKER_00]: earlier as well. But one of the ones I think is important is time to conversion.

[00:30:34] [SPEAKER_00]: So again, as I was chatting before about this idea of high-end products and that you

[00:30:43] [SPEAKER_00]: wouldn't buy a high-end laptop or high-end product in two weeks time, you can actually

[00:30:50] [SPEAKER_00]: look at, you can use the AMC to look at what kind of gaps in time people are taking to,

[00:30:56] [SPEAKER_00]: from their first exposure to any of your advertising, how long does it take for them

[00:31:00] [SPEAKER_00]: then to come on and purchase. And the real advantage of that is that it does allow you

[00:31:08] [SPEAKER_00]: start thinking about things like when I've made a new, when I've made a shift in strategy,

[00:31:14] [SPEAKER_00]: how long should I be looking at before I need to call these results? Because if I'm

[00:31:20] [SPEAKER_00]: something where like 80, 90% of my purchases come within a week, then I need to know that

[00:31:25] [SPEAKER_00]: I need to be moving fast. I make a change. If I haven't seen something happen in a week,

[00:31:29] [SPEAKER_00]: then have I made the right change? Probably not. Whereas if I'm something that is taking

[00:31:38] [SPEAKER_00]: if 80% is in two months time, and I make a big change now, and I wait two weeks,

[00:31:46] [SPEAKER_00]: see nothing's happened and go, oh, that's failed. Let's kill the test.

[00:31:51] [SPEAKER_00]: I've just spent two weeks on nothing basically. So I think time to conversion

[00:31:56] [SPEAKER_00]: is a really good thing to help you plan out your tests and experiments.

[00:32:00] [SPEAKER_02]: I love that. I think that's such an important point. I think so many people,

[00:32:03] [SPEAKER_02]: I mean, people will use a benchmark of like two weeks or something. But I think it's probably,

[00:32:09] [SPEAKER_02]: you know, it's tied to the 14 day default attribution period, which is pretty arbitrary.

[00:32:14] [SPEAKER_02]: And, you know, the advertising industry way before Amazon even sort of offered ads

[00:32:19] [SPEAKER_02]: has been very much like that numbers just been batted around since, you know,

[00:32:23] [SPEAKER_02]: Google started, you will almost right. And it's just very arbitrary. It's probably right for

[00:32:27] [SPEAKER_02]: certain large categories, but very wrong for other ones. Right? I think this is such an

[00:32:30] [SPEAKER_02]: important point. And that's the ties into the, you know, the attribution custom attribution stuff

[00:32:35] [SPEAKER_02]: as well. Right? Like if you're talking how long should that window be and how should we attribute

[00:32:39] [SPEAKER_02]: all the different ads that you're seeing? I think you get very, very different pictures of

[00:32:45] [SPEAKER_02]: campaign performance and how you should be splitting a budget.

[00:32:48] [SPEAKER_01]: I'm just going to throw in another curve ball into that as well. It's like the number of

[00:32:52] [SPEAKER_01]: screens that it's the time to attribution, but the number of screens that consumer may be

[00:32:57] [SPEAKER_01]: the product on. I checked my stats here, the average number of screens,

[00:33:04] [SPEAKER_01]: screen devices in an American household is over 10. It's 10.37 for those people who are

[00:33:09] [SPEAKER_01]: interested. But anyway, my point is it's like smartwatches, mobiles, laptop, et cetera.

[00:33:15] [SPEAKER_01]: You might be looking at your, you know, your fancy high value products a few times going

[00:33:19] [SPEAKER_01]: should I, shouldn't I, shouldn't I? And you'd be over multi-screen. So AMC should be able to

[00:33:25] [SPEAKER_01]: attribute that journey as well. So not only just the time to purchase, but

[00:33:29] [SPEAKER_01]: the past the purchase, I guess, is my point. Yeah. Yeah, absolutely.

[00:33:35] [SPEAKER_02]: Bringing it back to reporting. I mean, it sounds like almost the AMC will tell you

[00:33:41] [SPEAKER_02]: very different story to what the standard reporting would be right within the DSP. Are

[00:33:46] [SPEAKER_02]: you ignoring the DSP reports or how is it that you, I guess, work with both of them or

[00:33:51] [SPEAKER_00]: consult the two different stories? So no, I think is a sort of fundamental

[00:33:57] [SPEAKER_00]: answer. So even with Last Touch and even with the problems that we've discussed with Last Touch,

[00:34:03] [SPEAKER_00]: it still is a useful indicator. And I think the speed at which you need to be moving,

[00:34:10] [SPEAKER_00]: almost take the opposite angle of the last point I made about running tests for a long

[00:34:14] [SPEAKER_00]: enough time to make sure they're running. At the same time, a lot of the time you do

[00:34:18] [SPEAKER_00]: be moving fast and responding to things when you see things change. And I think that

[00:34:25] [SPEAKER_00]: certainly at the moment, I don't think it's sort of feasible to be running your AMC

[00:34:33] [SPEAKER_00]: templates and dashboards and reports on the same timeframes that you'd be looking at for

[00:34:40] [SPEAKER_00]: the AMS and the more direct reporting that you get. I think what it is, is that

[00:34:47] [SPEAKER_00]: to me, you use your ANC reports to give you a more detailed look at sort of conversion factors

[00:34:54] [SPEAKER_00]: and things like that. So it's things like when I see a specific campaign has a ROAS,

[00:35:03] [SPEAKER_00]: I know that in my head that actually some of that's being diverted elsewhere. So I can't

[00:35:08] [SPEAKER_00]: get quite as much budget as it would imply from what I'm immediately seeing. So there's

[00:35:16] [SPEAKER_00]: has to be slightly dialed up compared to what I'm seeing, and this number needs to be slightly

[00:35:20] [SPEAKER_00]: dialed down compared to what I'm seeing. But it still allows you to move quite quickly. So

[00:35:25] [SPEAKER_00]: it's one of those things where we give different targets depending on what the

[00:35:30] [SPEAKER_00]: activity is supposed to do or even looking at different metrics. So if you're looking

[00:35:34] [SPEAKER_00]: at something that's completely up a funnel, again, to bring it back to that campaign I

[00:35:37] [SPEAKER_00]: saw that wasn't delivering any sales, then it's a case of, okay, that's fine, but we're

[00:35:41] [SPEAKER_00]: looking at CPMs, we're looking at CPCs. How many people is this reaching for that budget?

[00:35:45] [SPEAKER_00]: Because we know that it's delivering later down the line, even if we can't see it here. So it's

[00:35:49] [SPEAKER_00]: about making sure that you understand the purpose of your ads and the ANC gives you a tool

[00:35:55] [SPEAKER_00]: to better hone in those purposes. And you mentioned a number of metrics.

[00:36:01] [SPEAKER_01]: And I guess for some listeners, it's a bit like metric bingo this podcast, isn't it?

[00:36:05] [SPEAKER_01]: Because there's a lot of metrics, there's a lot of different metrics and acronyms.

[00:36:09] [SPEAKER_01]: And I've got the latest one we've heard about is recently,

[00:36:14] [SPEAKER_01]: IROAS has been thrown around. What is this and why should brands be interested in it?

[00:36:20] [SPEAKER_00]: So this is brilliant. And it kind of comes back to a lot of the things that we're talking

[00:36:28] [SPEAKER_00]: about attribution about like, and these conversion factors in terms of how much

[00:36:35] [SPEAKER_00]: does this campaign say it's giving versus how much is it really giving?

[00:36:43] [SPEAKER_00]: So IROAS is incremental return on ad spend. So it's not just about what sales has this ad

[00:36:51] [SPEAKER_00]: delivered, it's what sales has this ad delivered that you wouldn't have got already.

[00:36:57] [SPEAKER_01]: That's interesting.

[00:36:58] [SPEAKER_00]: If you'll permit me, I've got another analogy that I like to use for this one as well.

[00:37:03] [SPEAKER_02]: It's better than writing out a formula, Rob. So go for it.

[00:37:05] [SPEAKER_02]: You ready for this, listeners? Here we go.

[00:37:09] [SPEAKER_00]: So imagine that you run a pizza parlor, you're on a pizza shop, right?

[00:37:14] [SPEAKER_00]: I also want to know, I think I've read this on a blog somewhere.

[00:37:18] [SPEAKER_00]: It's just a really great analogy. So apologies that I can't credit the blog where I read this.

[00:37:24] [SPEAKER_00]: So congratulations to them. Apologies for using your analogy.

[00:37:28] [SPEAKER_00]: But you decide that you want to run some flyer in for your pizza parlor.

[00:37:33] [SPEAKER_00]: So you hire two flyers, flyer one, flyer two, and you give them each a stack of flyers and

[00:37:39] [SPEAKER_00]: you decide I want to be able to track how efficient these flyers have been.

[00:37:44] [SPEAKER_00]: So I'm going to give discount code one to the first one and discount code two

[00:37:48] [SPEAKER_00]: to the second one. And then whichever pizzas I see come in with this discount code,

[00:37:54] [SPEAKER_00]: I know that that flyer has been doing their job and delivering me lots and lots of value.

[00:37:58] [SPEAKER_00]: And you wait about two weeks, that magical number for testing.

[00:38:03] [SPEAKER_00]: And over the two weeks, you see that flyer one has delivered maybe 90% of your pizza sales

[00:38:08] [SPEAKER_00]: now have flyer code one on them. Flyer code two, about 5%, and 5% just haven't bothered with

[00:38:16] [SPEAKER_00]: the discount code at all. So you think super clearly, flyer one is doing all the work.

[00:38:23] [SPEAKER_00]: Absolutely. Let's bin flyer two. They're not doing anything.

[00:38:27] [SPEAKER_00]: Let's keep flyer one. You wait another two weeks and suddenly all your sales have tanked.

[00:38:34] [SPEAKER_00]: What's happened? So you go out and you decide this time I'm going to look and see what flyer

[00:38:40] [SPEAKER_00]: one is actually doing. Flyer one, standing at the door to your pizza parlor and handing a

[00:38:44] [SPEAKER_00]: flyer to every single person who walks in. So they've delivered nothing, right? That's

[00:38:56] [SPEAKER_00]: funny. They're absolutely branded search of very niche core terms.

[00:39:01] [SPEAKER_00]: As your branded terms. Yeah, your absolute branded

[00:39:03] [SPEAKER_00]: terms. It's your complete lower funnel for someone who's already going to buy.

[00:39:10] [SPEAKER_00]: Whereas flyer two, they're running around different streets. They're delivering. They're

[00:39:14] [SPEAKER_00]: reaching new people and bringing them in. So in terms of which one had a better incremental

[00:39:19] [SPEAKER_00]: value. It's two, right? They were the one who was delivering additional value.

[00:39:25] [SPEAKER_00]: As I said, incremental value can be a lot lower on things like branded terms.

[00:39:33] [SPEAKER_00]: For example, we did a test recently where we switched off a branded campaign and we did see

[00:39:40] [SPEAKER_00]: sales drop but not by the full amount that the branded campaign was delivering. So we actually

[00:39:45] [SPEAKER_00]: saw a bit of a conversion factor on that one to see that for our branded campaigns, if they're

[00:39:52] [SPEAKER_00]: not delivering three times the ROAS of a generic campaign, they're probably spending too much on

[00:39:56] [SPEAKER_00]: your branded campaign. Obviously this will change depending on industry and the individual

[00:40:02] [SPEAKER_00]: campaign and their own branded terms. What I would say is that with the incremental ROAS,

[00:40:07] [SPEAKER_00]: it's not an invitation to switch off all of your branded campaigns just before people go

[00:40:11] [SPEAKER_00]: panic and go, ah, right. Let's switch off that. People are going to buy anyway.

[00:40:16] [SPEAKER_00]: Amazon has so many ads on its page. If you're not running a branded campaign,

[00:40:25] [SPEAKER_00]: other than in some circumstances, you won't be seen when someone loads up the page.

[00:40:29] [SPEAKER_00]: It's all just ads until someone scrolls down. But it's worth testing. It's worth looking at

[00:40:34] [SPEAKER_00]: it and saying how much is it actually delivering and therefore coming up with

[00:40:39] [SPEAKER_00]: here's the ROAS that it needs to be above everything else I'm running because I would

[00:40:44] [SPEAKER_02]: have got a lot of those customers. Is this ROAS something that Amazon Marketing Cloud actually

[00:40:49] [SPEAKER_02]: gives you or is this like using something from the AMC to calculate this? It's not something

[00:40:55] [SPEAKER_00]: that I'm aware of AMC having any specific tools. It's more of a case of understanding your

[00:41:00] [SPEAKER_00]: industry and your market and your advertising and what it's doing to be able to measure

[00:41:06] [SPEAKER_00]: some of these things. There are lots of different ways that you can measure these

[00:41:12] [SPEAKER_00]: types of things. There's things like charity tests on DSP that you can run to look at,

[00:41:17] [SPEAKER_00]: IROAS. It's not the best test, but as I mentioned, you can switch things on and off.

[00:41:24] [SPEAKER_00]: Obviously there's impacts of if you switch something off in Europe,

[00:41:29] [SPEAKER_00]: stock disappears the next day, you can't really claim that that was what caused it.

[00:41:35] [SPEAKER_00]: So bear in mind you're not falling into some of these pitfalls, but there's definitely

[00:41:38] [SPEAKER_00]: ways that you can run some tests to get ideas for this kind of thing.

[00:41:44] [SPEAKER_01]: An incremental revenues, very close to all of our hearts, has been able to track this across

[00:41:50] [SPEAKER_01]: not only Amazon, but I guess you've had BlinkOut campaigns as well. Could you

[00:41:56] [SPEAKER_01]: benchmark incremental ROAS on a consumer site versus Amazon for a brand?

[00:42:00] [SPEAKER_00]: Yeah, I always think that this is a thing that brands need to look at in terms of

[00:42:08] [SPEAKER_00]: what they want their strategy to be. Amazon gives you such a wide pool of customers and

[00:42:16] [SPEAKER_00]: I know that a lot of brands do have a lot of fees that they have to pay on Amazon,

[00:42:21] [SPEAKER_00]: which they obviously don't need to pay on their own DTC sites. But there is that question of

[00:42:29] [SPEAKER_00]: a smaller slice of a bigger pie, basically.

[00:42:31] [SPEAKER_01]: Yeah, I think we all have it, don't we? With brands, when we start advertising with them

[00:42:37] [SPEAKER_01]: and they always raise it as, I don't want to do anything that disturbs my DTC customers,

[00:42:44] [SPEAKER_01]: I don't want to take them away to Amazon. You can't go anywhere near them, no pixel,

[00:42:50] [SPEAKER_01]: no remarketing, that's it basically. It's usually a conversation, isn't it? But

[00:42:54] [SPEAKER_00]: that's a really important point.

[00:42:55] [SPEAKER_00]: Which is always funny because you can use the pixel and the tag on the website

[00:43:00] [SPEAKER_00]: specifically to exclude them, it's a way to keep that away. But that's not a conversation.

[00:43:08] [SPEAKER_01]: But I guess my point is, if you're showing incremental ROAS on

[00:43:12] [SPEAKER_01]: East Hire on the DTC site versus Amazon, then that's going to inform

[00:43:15] [SPEAKER_01]: where you want to put your budget and your tactics, isn't it?

[00:43:18] [SPEAKER_00]: Yeah, absolutely. So I do think it's worth looking at that to say,

[00:43:22] [SPEAKER_00]: are you reaching people easier? Does Amazon's trust and prime shipping and things like that

[00:43:29] [SPEAKER_00]: help that conversion rate and help you get, as I said, it's a smaller slice of a bigger pie

[00:43:34] [SPEAKER_00]: and that's better for your incremental revenue in that way.

[00:43:40] [SPEAKER_01]: Yep, absolutely. Super interesting.

[00:43:42] [SPEAKER_02]: Moving on to, I get more about users or stakeholders. Is the AMC, it's clearly

[00:43:48] [SPEAKER_02]: helpful for people like yourself to help you make some of those bigger decisions around

[00:43:51] [SPEAKER_02]: where to split the budgets and incrementality and things like that. But are there

[00:43:59] [SPEAKER_02]: use cases or uses for, let's say, brand owners or CEOs or C-Suites as well? Or is it more,

[00:44:06] [SPEAKER_02]: I guess, someone who's looking after the ad budget or the day-to-day

[00:44:10] [SPEAKER_02]: sort of ad operations side of things?

[00:44:12] [SPEAKER_00]: Yeah, I guess we kind of touched on this a little bit earlier as well. But I guess just

[00:44:16] [SPEAKER_00]: to say that from my perspective, I'm obviously going to come at this from an

[00:44:20] [SPEAKER_00]: advertising perspective. That's the lens for which I'm going to be seeing this.

[00:44:25] [SPEAKER_00]: But I do definitely think that there is a lot of different people who can use this in different

[00:44:33] [SPEAKER_00]: ways. As we discussed before about that idea of what are these paths of conversions,

[00:44:38] [SPEAKER_00]: what messaging are people reacting to and coming through? That's something that a brand

[00:44:43] [SPEAKER_00]: owner can understand to sort of understand what is the customer profile that I'm actually

[00:44:49] [SPEAKER_00]: going for. So that's something that your brand owner and your brand marketer is going to be

[00:44:54] [SPEAKER_00]: interested in. Things like lifetime value is another big thing which will understand,

[00:45:02] [SPEAKER_00]: do we want to change pack sizes? Do we want to go more into the subscription model? Do

[00:45:06] [SPEAKER_00]: we want to look at changing the... Do we want to go in the other direction and offer smaller

[00:45:12] [SPEAKER_00]: introductory products and things like that? So yeah, I think it's going to be heavily

[00:45:18] [SPEAKER_00]: weighted towards advertising because it's going to... Obviously you'll intersect with

[00:45:23] [SPEAKER_00]: advertising very directly, but advertising is a way to push the business and it's a

[00:45:29] [SPEAKER_00]: great way to get fast data. I think everyone in the business from the top to bottom is going

[00:45:33] [SPEAKER_00]: to be interested in some form of this in terms of that. I think yeah, just marketing

[00:45:38] [SPEAKER_01]: out on again, what really is the gold nuggets within the world of AMC and why it is

[00:45:47] [SPEAKER_01]: so important to C-suite is actually Amazon has billions of data points about the consumers

[00:45:53] [SPEAKER_01]: in the UK and literally billions. You think about the whole universe across web, Twitch,

[00:46:00] [SPEAKER_01]: IMDB, Alexa, there's a very, very rich dataset that can be accessed. And from my point of

[00:46:07] [SPEAKER_01]: view, my journey was through Facebook and traditional digital advertising. Facebook

[00:46:14] [SPEAKER_01]: thinks it knows what you like and the things that you follow. Amazon knows what you like

[00:46:21] [SPEAKER_01]: and knows what you buy basically. And I think that's the best way to describe it.

[00:46:25] [SPEAKER_00]: It reminds me of a story of a lot of surveys went out for different products. And if you ask,

[00:46:35] [SPEAKER_00]: maybe in America, but I think most places, but if you ask most people,

[00:46:39] [SPEAKER_00]: what kind of coffee do you like to drink? They'll say, oh, a rich dark roast. That's

[00:46:45] [SPEAKER_00]: what I like to drink. If you look at buying data, that's mild, it's mild all the time.

[00:46:51] [SPEAKER_00]: That's what people buy. That's what we want to drink. But you ask them,

[00:46:54] [SPEAKER_00]: that's not what they'll say. So yeah, I think Facebook is what people say. Amazon

[00:46:59] [SPEAKER_01]: is what people do. Yeah. So I think that just to bring it into clarity, that's the way you

[00:47:04] [SPEAKER_01]: look at it. What data set do I trust? My one is on Amazon because it knows exactly what you

[00:47:09] [SPEAKER_01]: bought basically, rather than what you portraying yourself as in the world of social is very

[00:47:14] [SPEAKER_01]: different. We are different human beings on a podcast, for example, versus actually what we

[00:47:18] [SPEAKER_01]: do. But my point is that real data about your consumers is super powerful and the whole C

[00:47:25] [SPEAKER_01]: suite should at least lean in and have access to that because it's such a valuable data set.

[00:47:32] [SPEAKER_02]: I think it's pretty important. Rob, just to start closing this out here,

[00:47:36] [SPEAKER_02]: we always like to talk about things that have piqued your curiosity lately. So it could be

[00:47:42] [SPEAKER_02]: AMC or non-AMC related, more of a much general comment there what's piqued your curiosity.

[00:47:47] [SPEAKER_00]: So I've been getting into philosophy recently. So I've been watching a lot of philosophy

[00:47:52] [SPEAKER_00]: videos on YouTube. As you do?

[00:47:55] [SPEAKER_00]: As you do. No, it's good because I was a little bit ill recently, so I haven't been able

[00:48:01] [SPEAKER_00]: to go to the gym as much as I'd like to. So I go for a walk and I just listen to philosophy

[00:48:06] [SPEAKER_00]: videos on YouTube as I walk, which is a very relaxing way to spend the evening.

[00:48:12] [SPEAKER_02]: Nice. You're not going to write a

[00:48:15] [SPEAKER_02]: LinkedIn post linking these two things, right? Advertising and philosophy.

[00:48:17] [SPEAKER_02]: No, no, I don't think I can. I'm not sure how that works.

[00:48:23] [SPEAKER_01]: Comedy, flyers, all the mind boggles.

[00:48:26] [SPEAKER_00]: Yeah, I feel like I should try and get something advertising

[00:48:30] [SPEAKER_00]: focused in as something that's got the curiosity.

[00:48:33] [SPEAKER_01]: Well, just for listeners, if you've got anything that piques your curiosity,

[00:48:38] [SPEAKER_01]: then let us know on the hashtag, youcominsights or find us out on LinkedIn. Rob McGovern's

[00:48:45] [SPEAKER_01]: also on LinkedIn as well. So any questions far over is Oracle of Knowledge, philosophy,

[00:48:52] [SPEAKER_01]: standup comedy, all sorts of things. Ask me philosophy and standup comedy questions

[00:48:57] [SPEAKER_00]: and then a few advertising after that. Lots of breadth and lots of depth there.

[00:49:03] [SPEAKER_02]: Please drop us an email as well, ecom.insights at optimisent.co.uk.

[00:49:07] [SPEAKER_02]: Any feedback, suggestions? The good, the bad and the ugly we want to hear it all.

[00:49:12] [SPEAKER_01]: Yeah, and don't forget to hit that subscribe button,

[00:49:14] [SPEAKER_01]: listeners and finders out on all the usual channels, Spotify, etc. to subscribe to

[00:49:18] [SPEAKER_01]: the future podcasts. And thanks so much, Rob, for joining us today. I have a few news that

[00:49:23] [SPEAKER_02]: maybe the last time. Thanks for having me. Great. So until next time, keep thriving,

[00:49:29] [SPEAKER_02]: stay curious and we'll see you soon.