Inside media measurement at Google: Balancing brand, performance and AI

Key Takeaways

1. Balancing brand and performance marketing

Google emphasizes the importance of balancing brand and performance marketing. Siim highlights that while brand marketing creates long-term value by building recognition and emotional connections, performance marketing delivers short-term results. Both approaches are critical but operate on different timescales.

2. The role of creative testing

Creative testing is evolving with new tools and methodologies. Siim explains how real-time analysis of ad reactions—such as identifying moments of peak emotional engagement—can improve the effectiveness of creative assets. This allows marketers to optimize their campaigns and enhance performance.

3. Using proxy metrics for brand measurement

For smaller businesses or campaigns with limited budgets, Siim suggests leveraging proxy metrics like direct website traffic or share of search to gauge brand performance. These indicators can provide valuable insights without requiring large-scale measurement tools.

4. Experimentation is key to measuring effectiveness

Siim emphasizes the importance of running controlled experiments, such as exposed vs. control group studies, to determine the true impact of media campaigns. These experiments help businesses understand whether their marketing efforts are incremental and profitable.

5. AI enhances efficiency and learning

Siim discusses the practical applications of AI in marketing, including generating summaries, aiding in coding tasks, and simplifying complex data analysis. AI also serves as a valuable learning tool for marketers, helping them refine their strategies and approaches.

6. Attention metrics drive media effectiveness

Attention is a crucial factor in media measurement. Siim explains how tools that assess real-time reactions, such as facial expressions or viewer behavior, provide actionable insights into ad performance. These metrics ensure campaigns resonate with audiences.

7. Creative testing and always-on marketing

Siim highlights a survival-of-the-fittest approach to creative testing, where marketers continuously test multiple assets within live campaigns. This approach ensures that only the most effective creatives are prioritized, maximizing campaign success.

Read full transcript below:

Edward:

Hi, I'm Edward from Supermetrics and this is the Marketing Intelligence Show, the podcast that empowers marketing leaders to work better with their data and make sure every marketing dollar counts. Now let's get into today's episode.

Evan Kaeding:

Hey everyone, and welcome to another episode of the Marketing Intelligence Show by Supermetrics. I'm Evan Kaeding, director of Solutions Engineering here at Supermetrics, and I'll be your host today.

I'm really excited about this episode because we have Siim Säinas joining us from Google, whose title is Social Measurement and insights manager at Google for time at Google. Siim spent years in the agency world working across diverse brands and industries, which gives him a unique perspective on measuring media effectiveness and social impact. Today we're diving into his insights on balancing brand and performance marketing, the tools and metrics that matter, and some real world tips that any business can put into action. Let's jump in.

Evan Kaeding:

Welcome to the Marketing Intelligence Siim. Would you like to introduce yourself to the audience?

Siim Säinas:

Hi, glad to be here. My name is Siim, I'm a marketing analyst in Google. I look after specifically how social media performs for us and focus on consumer marketing team. So looking forward to this discussion.

Evan Kaeding:

So si I'm sure that many of our listeners will be interested in hearing about your day-to-day in Google and maybe a bit about your career story and how you got there. Could you enlighten us?

Siim Säinas:

Yeah, thank you Evan. I started in agency purely focused on at the time when I joined website analytics and it was founded by two Finnish guys, extremely brilliant and powerhouses on their own. I did that for seven or eight years. Then there was a startup in between and then now six or something seminar years, it's been in Google and in Google what I do is I focus on consumer marketing and a couple of different products that are more important in the APAC region where I'm based, where Singapore is hq, so primarily Google search, Gemini, pixel, Android, those are sort of the usual suspects along with the overall sort of advancements of making our social media marketing more effective and more efficient. So that's sort of my day-to-day job here. So it sort of crosses between data and marketing effectiveness and a bit of creative here and there and try to stand to just be helpful and useful to people.

Evan Kaeding:

Got to love working for finished co-founders, certainly something that I've found beneficial in my time here at Supermetrics.

Siim Säinas:

Yeah, those jokes about finished people mostly are true whatever jokes. But yeah, it was quite an interesting period of time. I felt like I sort of saw the tail end of some really major shifts in the webinar analytics world, partially driven by Google, but that also those changes also pushed me to the social media side of things. So what happened there was that there was a new product that was launched as part of that company that I early joined and it happened so that one of the clients was asking for a social media monitoring tool and the team at that time, which was absolutely amazing, built it literally on McDonald's in two weeks. So it was very boggy but the early versions of it. But it did some incredible things that later, even Avinash Kaushik who became, I would like to say my mentor presented even in a couple of presentations saying this is sort of the future.

But having said all of that, that was an exciting period of time. It taught me a lot about different types of brands and obviously being on the agency side, I really treasure that time as well because some people sort of look down on it, but I always find it super fascinating that on a single day you could be switching from selling car lubricants to diet Coke to software and something else in between. So I found that very fascinating. You learn very quickly. So I think if you're trying to get into the industry or marketing is sort of your field, I think start from an agency, I think those are the best breeding grounds. That's what I personally think.

Evan Kaeding:

Couldn't agree more. That's one of the big milestones in my career certainly was getting the opportunity to work at a large agency and work on large multinational brands, swivel chairing from selling anything from fried chicken to online video games to deodorant in stores around the us. So it was a great formative time for me as well. Great. Alright, awesome. Thanks Siim. And one of the several unique things that I think you can bring a perspective to for our audience is just how the machination works for the measurement of media at a giant Google. I know that you work quite a bit on media effectiveness, you work within the Android line of products. Can you tell us a little bit about how a large company like Google approaches things like media effectiveness and social measurement?

Siim Säinas:

Yeah, happy to. I think the first quick disclaimer is that all opinions are mine needs to be said. So I can definitely represent my working experience from working with many multinationals in the past, but I think it's broadly speaking sort of a mix of many different specialist agencies these days and largely speaking, splitting ourselves between brand and performance and trying at the end of the day to figure out what this whole budget is doing, how much of it is effective and how much of it is incremental and are we moving towards profitable growth or not, which is ultimately for all the businesses sort of the north star.

Evan Kaeding:

So you talked a little bit about the difference between brand and performance marketing. Could you maybe give us a view on how you view the dichotomy if you choose to accept it even?

Siim Säinas:

Yeah, totally. I mean I think we all sort of agree that in an ideal scenario a lot of things are just synergized beautifully and integrated and studies after studies have shown that more channels generally speaking are better for effectiveness. Question is just like how many can we manage well enough so that it actually has a real world impact. That's always been the challenge, but even budgets that are let's say in the hundreds of thousands that are not necessarily really large, they will still benefit from many channels rather than just focusing on one. So that's one, but I mean zooming back a little bit more even I think there's some really fantastic and insightful research coming out of the uk, some institutions also in Australia and the US that are 60 40 brand versus performance. I think everybody sort of knows that and I think broadly speaking, everybody who's on the marketing effectiveness side is vouching for a big yes.

I mean it tends to be true. I think it varies a little bit market to market product by product, but broadly speaking it's a good thumb roll. And what's interesting I guess about brand and performance is that they're both necessary and they're both very, very important, but they just operate on different timescales and when we sort of get too stuck with performance, then I suppose the risk there is just to focus too much on what's the next quarter going to look like without really thinking about are we securing our future for the longer term? Because at the end of the day, I mean brand marketing from a very cold ally perspective is really about efficiency and creating that sort of shortcut in our mind that helps you to deliver that next message in a slightly more effective way. And I think that's what smart brand teams and marketing teams generally aspire to do with varying degrees of success and I think we're sort of in the same ballpark.

So that's sort of the gist of it, the top level view, but there's a lot that goes on underneath that. I mean every channel, every marketing mixed decision comes with its own caveats, but I think generally speaking, you would always want to gravitate towards methodologies that are more incremental methodologies that are more controversial, exposed or have a holdout group things that help you to get to that sort of source of truth a little bit more credibly versus just basically doing something that is based on opinion or a qualitative study. So I think that's sort of the gist of it. But for smaller brands, I mean given that I believe this podcast especially also is listened to by a lot of smaller brands, then I think the tricky part there is to get the brand part or the brand measurement sort of credibly locked down. I think it's most businesses will start from, they will certainly focus on the website and I think there's a lot of performance tactics that are being shared, but what seems to be more tricky is to get a go on where is my brand and especially when it's tiny or small, how do you know what's a good indicator there?

But there's some sort of green shoots and I would say some optimism in that space as well. So for example, you could use share of search. I mean that happens to be Google, but applicable for a lot of businesses out there that are small and especially even regional mom and pop shops. So that's literally the queries or the searches within your category that you technically own, which has been proven in several categories to be sort of predecessor to market share. Doesn't apply for all categories, doesn't work for all categories, but for some it really does well and there's some interesting startups in that space as well that helped to get closer to what Miller Brown used to charge half a million for, but cheaper, more self-service like Tracksuit is a startup, interesting startup based out of New Zealand offers that sort of service also that is brand tracking basically for SMBs and mid-sized brands as well. And of course, even if that seems to be out of reach, then I think something like when you go to website analytics, you would have direct traffic for example, coming through a website looking at that as a potential proxy for brand awareness. I think those are sort of the areas where you need to get a bit more creative.

Evan Kaeding:

I like the way that you described using proxy metrics for example, to assess the value of brand because that's obviously going to be one of the tougher things to do when you're advertising in any case with a constrained budget where you know that you need some form of direct response in order to justify your media effect, your media spend. In addition to that, probably long-term for the growth and health of your business, you need to justify the investment in brand media as well. For companies who are just getting started and trying to collect the data that they would need to assess the effectiveness of their performance and brand, what would you recommend they start with From a data collection perspective?

Siim Säinas:

I would really start from experiment specific where you could get a brand lift, read these vary by platform. I mean can't remember the whole rate guard by default, but Indonesia you would need to spend at a single campaign with in around four weeks certain region frequency assumptions being made and around 10 to 13 K US on a single campaign. And then the brand lift measurement for example on meta platforms is already provided sort of free of charge. So similar mechanics are also offered by other platforms is the rate that which this is deployed or offered to you as an advertiser might be slightly different, but it's certainly accessible. So those are already, I would say the industry standard for controversials exposed and you're going to get two things from there. One is that you're going to get the baseline in terms of where your brand is if the activation was not there, and also the comparison between the activation period and what your assets actually did.

And the beauty of that is that the best in class methods will always use device IDs and won't resort to someone claiming that they seen your ad, but it will actually be pretty much a verified signal. So I would probably start from there if you've got already a little bit more budget, but the previously mentioned tactics that would also not resort away from there just to look at share of search, look at website traffic, especially direct what's coming in and yeah, look at those sort of startups that help you to bot a little maybe with a smaller sample size because usually that's where sort of the bulk of the budget is parked away when you go to someone like Kantar or Ipsos or Millwood Brown, but would be able to do a custom study that is perhaps smaller, cheaper, but you'd get a good sense already.

Good thing with custom studies of course is also that you get to add a few more questions. When I said my first go would be brand lift, but brand lift also has limitations. Usually they're very rigid in terms of setup of the questions. You've got only three slots and one is usually parked for recall, which sort of means something but not really something. It's sort of somewhere in between. I think every analyst would have their opinion, but those are some of the thoughts that I would go with. But it's really dependent also on what business you're in.

Evan Kaeding:

Yeah, makes total sense. One of the things that we talk perennially about is the role of creative. Now whether it's the role of creative and brand, the role of creative and direct response, I think there's a pretty exciting ecosystem of opportunities for creative testing that are coming up, which I think historically have been limited to large advertisers, but could you talk a little bit about some of the advents in creative testing that you're seeing in market today?

Siim Säinas:

Yeah, creative testing is an interesting one. I'm also fairly new in some ways as a client I think for the past three to four years, but it's clear that this industry in itself is going through quite a sizable shift and it's very exciting there is that first of all the use of machine learning to understand, let's say someone's emotions when they're watching through an ad. So previously all of that, or ever since from the eighties that I can recall the first study that was done on pretesting, it was always survey based and you're basically, you've got to watch a thing but you're not necessarily sure if in that moment you're already dealing with certain form of bias that we all have. So that's always the concern with surveys, but when the AI or the machine learning algorithm can sort of in real time look at your face and understand if you're frowning or fiddling with your phone, that in itself can be quite powerful because what's been clear from the past studies is that when there's ad likability and there's sort of watch through, let's say if it's a video asset, then those tend to correlate really well with brand lift and also with commercial outcomes such as revenue.

So those are the things you generally want to look for in terms of KPI. But there's a lot of diagnostics you could also learn from. For example, these days that's very powerful is also to know which card is going to be surfaced as the first sort of thumbnail when you're advertising let's say on YouTube or some other social channel that makes a big deal. I mean we all know that Mr. Beast I think has a whole team of people that are a dedicated on thumbnails, which is fascinating. So these things do for clickthrough but also down the line for commercial outcomes. And what these tools enable to do these days is to really pick out the moment where you see that there's, for example, being peak happiness. Most people out of all the watchers are indicating that their most mostly enjoying that specific moment. Then you could pick that card out and make that your thumbnail.

So you can even go that granular. And what's very exciting is these tools and these solutions and services have become more cheaper and this is really thanks to ML and AI and now the new development, I think what I've seen, especially as a client is that I'm seeing vendors that are coming up with synthetic testing solutions, right? Meaning that now it's based on all that body of evidence that we've got from human tests and the AI is being taught to sort of find those patterns and if previously from external studies, we got to know that pretesting as a category tends to be around 70% accurate. Then the synthetic version by all we know today could be very similar, which is sort of interesting. This I'm not talking about anything internally as far as I'm aware and neither I'd be able to share anyway, but this is really happening in the industry that all of us could get access to some of these vendors like realize human-made machine.

There's system one out of the uk, David, several different companies that are sort of in that space that are innovating in really interesting ways. Not to say the ad in itself, the one ad is always will determine the whole outcome of the budget, but arguably probably most campaigns it has the most weight. So you would want to test and spend maybe a few thousand dollars to be a bit more sure or maybe to optimize that a bit better, to increase your chances of success. So I think that's very exciting in sort of pretesting. So it's changing, it's becoming more available, accessible, it's becoming cheaper, it's becoming more insightful. What it won't do though is that some of the vendors will undeniably overpour you with metrics and I think that that's sort of very, this is all prevailing I think in the whole industry that the ability to siphon through those and say that this is actually what matters more is something that needs to come with certain expertise and perhaps a bit of discussion Also internally

Evan Kaeding:

Creative pretesting to me is super interesting because it fundamentally changes the economics of digital advertising because many brands will allocate a pretty significant portion of their advertising budget to testing creative testing media strategies on different platforms and with the opportunity to perhaps do that ahead of time, whether that's based on a real human panel or whether that's based on a synthetic panel, gives brands the opportunity to really understand how their creative will perform and how effective it is without necessarily allocating those media dollars and going through the traffick exercises. So if you're on a million dollar budget, you've got 50,000 allocated for your testing budget, well all of a sudden you can get quite a bit more out of that testing budget using pretesting than you otherwise would. Having to go through the mechanics of doing that trafficking process and doing that testing process, which is a pretty exciting economical change for performance marketers and I think brand marketers alike, would it be fair to assume now a lot of people in the industry are talking about attention metrics and I see that creative pre-testing could be a really interesting way to understand and maybe suss out the degree to which an ad captures attention.

Are you seeing those kinds of metrics available from these vendors as well?

Siim Säinas:

Yeah, absolutely. I think attention is sort of a little bit earlier in the row of things you're going to get back, but it's undeniably unless there's attention, not only it down the line will matter much. So that's the, I think universally agreed and everybody's looking at it as well as a strong proxy. So I think my mental model is almost like it's a bit of a funnel or a pipe and you're looking at loss along the way on every path. Attention is very early there not necessarily the earliest though because the earliest is probably the formats themselves and the channels and even brand that's also being shown as a very strong indicator whether someone is actually ordering the pay attention, even if it's a very high attentive environment. So those are interesting things. So I think that when you think of sort of ad effectiveness as a pipe or a funnel, whatever you want to visualize it with, then there's sort of leaks along the way all the time.

So it is sort of similar to the website world back in the day, but what is very powerful with pretesting is sort of what Abinash likes to say. Avinash ache is the wind before you spend. And I think that's exactly the point of it. For a lot of advertisers, it is a great way to make sure that you're really parking your dollars in the right place. And for smaller brands, my day-to-day is also selling some meditation cushions as a side project. Then obviously for some brands it might still not be feasible to even spend a thousand bucks, for example on a creative test. So there is definitely another school of thought here, which I find personally fascinating to always learn from, which is largely the DDC brand or the direct to consumer brands. And if you surface creative testing even as a word pair within those rooms, then you get a very different answer.

And in their world, creative testing will mean that you have a ton of assets in an ad set, let's say in ad manager on meta, and then it is basically a survival of the fittest process. There's no line at which you say that now the campaign starts or stops. It is basically marketing is always on. There might be a few seasonal moments like Black Friday, but other than that it's just continuous process. That to me is very fascinating in some ways, perhaps a merger or a blend of that world versus what I just explained is perhaps the future here, but that in itself also require or I think deserves a knot because I think that world will also show what sort of crappy creative ways are out there to actually find paths to effectiveness other than going for very analytical solutions.

Evan Kaeding:

True. And if you're operating at sufficient scale than having a variety of assets across a variety of different targeting parameters, you will end up in a survival of the fittest situation and ultimately find perhaps that couple of evergreen creatives that simply drive quite a bit of performance, whether that's brand awareness, whether that's direct response, whatever that happens to be for your brand. So multiple different ways to get there at the end of the day. I think that makes a natural pivot to one of the other topics of conversation that I wanted to touch on today, which is ai. And I know cm, you've worked quite a bit with ai, obviously being at Google, you've got probably a pretty interesting view that I think our audience would be interested to hear as well. So maybe from the top, if you could talk a little bit about AI and the specific marketing use cases that you see as being transformative or disruptive for our industry as it is today.

Siim Säinas:

Very popular topic these days. Generally I think resort in the world of pragmatism and see where sort of answers lie, where there's real gains in something. I think broadly speaking, most of the AI developments that are tested that are publicly available as well are sort of in the A of efficiency. They help you to save time and they help you to circumvent that day that you might have or maybe a party that you had really late night that you feel a little bit foggy in your head help to get to that sort of almost peer reviewed answer a little bit quicker. So that's very broadly speaking, but it will get more exciting very soon, but maybe about the future a little bit later. But the first, I think the obvious things today that in my mind are already adding incredible amount of value is coding any sort of code writing software development.

I personally, I'm not on the backend side of things, but I've tried different Python scripts, I've developed quite a bit of code on specific analysis pipelines, JavaScript, also sql, if that is necessary, some more complex things. It's a great aid for those things, some features that you might not be aware of. It's also a really good teaching assistant when you're trying to get into coding or trying to understand how do I extract data from this place in the most elegant or efficient way possible. So it can also be a great sparring partner when you're just trying to think of how would I approach this problem so it's a bit open-ended rather than just giving it a very direct sort of instruction. So that's sort of category one. Category two I would say is it seems to be that it's pretty good with creating narratives. So this could be a blog post, this could be a social post.

I'm working on a presentation for next week. You're probably not going to get something quite groundbreaking from there because there's something to be said about sort of the touch and the feel and the experience of the industry and what you've witnessed and what you think the anticipated answer will be versus what the AI spits it out to you. But it's again, it's a good sort of gorge on, there was a very famous TV show, I believe in the us which name I forgot, but when they always ask a hundred people, what would a hundred people say? Was it family food?

Evan Kaeding:

I believe so, yeah, that's the one.

Siim Säinas:

So it's sometimes it's like that. It's like when you're thinking about, oh, I wonder what does the general public think about this thing? It is very often you get a lot of these sort of answers, which I find also very helpful because it's easy to be locked up in your isolated world and think that everybody thinks the same way. And so those are the things that I think are useful. Also from a creation standpoint, I would say Dr. Creatives tend to be a little bit more, I would say, well obvious third bucket data analysis. I've done multiple different things to feeding it my insurance claims, to looking at let's say YouTube comments and mass breaking it down into themes into how many people are actually talking about buying our product versus talking about this influencer. There's many different applications where it will become instantly available to you.

These sort of answers that otherwise hiring an agency or time-wise even would take weeks at least to get back. So that's another, I think the third sort of category that I would say where there's immediate gains for user professional. But in the end of the day, a lot of this also, again depends a lot on the occasion at which you're working on, but I can certainly see that if you sort of paint out the marketing planning process, no matter how it looks in your company or agency, then I think every step of the way almost you will find some benefits, whether it's hoping to get to a crisper brief or refining a narrative maybe about being a sparring partner. Do you think this would be sort of a reasonable path at some point? Of course, I think you'll just need to resort to experience gut feel and so on, but I find it an incredible aid already today and multimodality.

I think that's sort of what everybody keeps on talking about. It is coming. It is already sort of there, meaning this is fancy for different types of inputs. So x video, pictures, audio, everything that you can sort of imagine I think where everybody's sort of cautious about or if I look at my friends and people around me, then I can imagine a world where not everybody's okay with it carrying around a device that is sort of clearly always just on the lookout of analysis. I think there was a good joke some years ago that the guys that are wearing the AI Google glasses, I think there was a good meme about it online that it is the best way not to get a date is wear one of these glasses. And I found that funny. And I mean you can clearly see that with all these developments in tech and ai, they'll very quickly be sort of a counter movement.

And the counter movement will clearly say that these are spaces where we don't want to see any AI or very minimal level of ai. So that's an interesting thing to observ see and how can we make it in such a way that it is disclaimed, it is safe. Also as a marketing professional, there's another caveat here that what you're sharing and then what are the user rights of these tools? What are you going to do with that data that you're feeding in? Are you okay with that, that it's going to be training other AI as well? Those are things that I think there's no universal answer to it. It really depends on the company. Chances are if you're a small company though and you don't really care that much, there's a lot of people out there like that as well, then you get immense benefits out of it.

But if you're a bigger company, these questions might actually become real roadblocks and consulting with your legal partner, your compliance team, whoever that might be is probably a good idea as well. But overall, I guess I wouldn't be here, but otherwise, but I'm definitely an optimist. I think a lot of things will become a lot more efficient. I think it lowers the bar of entry for a lot of people to get into marketing, get into data analysis, maybe creating content even perhaps. So those are good things. And I mean in the end of the day, where does it lead societally? I mean, that's up for smarter people to debate and figure out, but that's what I'm seeing at the moment.

Evan Kaeding:

Yeah, I like the way you categorized it. Is it going to be one giant shift all at the same time? No, not necessarily. But what you just described is AI making meaningful contributions at each step in the process, whether that's coming up with a tighter brief or using it as a sparring partner to align on what a consistent message or narrative might be using it to some degree for creative asset production. And then on the tail end, being able to use it to collect and maybe synthesize that data as well. Is it one kind of landmark change all at the same time? Not necessarily, but providing incremental benefits along the way. Now obviously at Google, I'm sure you get the opportunity to use Gemini in your day-to-day. I'm curious to hear if there are specific things about Gemini that you're either using today or are upcoming that you'd like to share with our audience.

Siim Säinas:

Yeah, I think most of my opinions, to be very frank, are really based on using tamini to begin with. I tend, because I work with the marketing team there closely. So I mean part of that is to test and make sure that I know what the product does. I think what's a good indicator is what are the things that you keep on using? So there's this sort of initial excitement about just even a plain response or maybe a witty response or a humorous response. But we had that also with assistant for example, or with Siri. But what's really powerful, what I've gone back to is code writing is analysis of things that could be a little bit annoying, like personal finances on the work front. I've definitely, I'm using it all the time to create summaries just to sort of shortcut and say, here's a larger piece of body of information, whether it be a deck or maybe a long document.

Can you help me summarize that? So I look slightly better in the next meeting that I haven't had the time to look at this huge pile of information and I'm just trying to catch up and do my best to be helpful to other people. So these are the things that I keep on going back to after the code is done very often. Obviously you'll find that if it works and there's not all of amendments necessary, then that's sort of done and dusted. But it's equally, I tend to go back for different learning, learning different new features or functions within the code that I wasn't aware of, and that I find also particularly useful. So yeah, you're maybe a little bit less of an annoying colleague, like always tapping on someone's shoulder, but you've got this sort of proxy in form of an AI that helps you to learn and have that sort of growth mindset that I think we all need in this relatively complex world.

Evan Kaeding:

Yeah, sounds super excited. I certainly use a combination of AI tools in my day to day, both the ones behind the company firewall and outside for my personal tasks as well. So important to keep in mind the privacy and your company's guidelines for using AI tools as well to assist with the day-to-day. We'll see. I've certainly enjoyed our time together. Is there anything else you'd like to leave the audience with today?

Siim Säinas:

No, thank you so much. It's been a pleasure and in the spirit of this podcast, I think as much as there's really exciting developments outside, I mean, I think reassuring people that you still have a job and a very meaningful job if you're in the field of analysis and marketing measurement, I think is also very important. So yeah, all the best of luck to all of you, and thank you for having me. Thank you. See much appreciated.

Edward:

Thank you for listening to the Marketing Intelligence Show brought to you by Supermetrics. If you're enjoying the podcast, then we'd love for you to tap that subscribe button, leave a review, and share with your colleagues and peers. We'll see you in the next episode.