Transcript
Intro:
Welcome to Unpacking the Digital Shelf, where we explore brand manufacturing in the digital age.
Peter Crosby:
Hey, everyone. Peter Crosby from the Digital Shelf Institute. When measuring Amazon DSP performance, if all you are looking at is ROAS impressions and views, you are not using the full range of metrics available to understand the impact of your spend across the full range of the consumer buying cycle. Kiri Masters, head of retail marketplace strategy at Acadia, recently published a report outlining 23 Amazon DSP performing metrics to measure, optimize and communicate the impact of your campaigns, both on and off Amazon. Kiri joined Lauren Livak and me with the highlights. Kiri, thank you so much for coming. We are so happy to have you on the podcast.
Kiri Masters:
Thanks, Peter. Great to be back.
Peter Crosby:
The world of retail media, we know everyone's spending on it. We're making big investments, we want big returns. And you recently put out some new research, which is really digging into the wealth of data that's sitting around in Amazon DSP to help people understand the true, full impact and performance of the spend that they're doing. And it's really an exciting sort of brain shift on how to think about what performance of the campaigns on DSP means. So I would love to just dig into it with you. It's exciting research.
Kiri Masters:
Yeah, absolutely. Yeah, excited to get into this. I'm not the first or last person to say that ROAS in particular is a metric that is pretty one dimensional. And I think a lot of brands are very open at this point to thinking outside of those standard metrics and what they could be, and considering what else is out there. But if that's up your alley, we've got a whole lot of things to talk about.
Peter Crosby:
Great. So tell us a little bit about what you were seeing among your clients and out in the market that led you to try and sort of reconfigure how to think about these metrics.
Kiri Masters:
Yeah, there was two things. One was a lot of brands relying on out of the box metrics, like ROAS and impressions, as markers of success on Amazon. And we were finding that as an agency we would see really good returns from the DSP, but not necessarily on the basis of those individual metrics. We would see incremental sales, we would see pay-per-click performance improve with the addition of DSP. But we got the sense that impressions and ROAS alone were not really telling the full story. And the reality is that DSP advertising and pay-per-click advertising is very different in the way that it's calculated and the way that success is measured. So we wanted to dig more into how were these metrics really measured on the DSP, and how could we get a little bit more sophisticated around measuring success beyond those standard metrics, because it's an impression-based ad platform, right?
So you could see an [inaudible] over a second and then that would be counted as an impression. But think about your own habits as well, and all the ads that we see every day. That may or may not consciously register for you, and it is counted as a conversion. So there's a bit of skepticism, including from my colleague and co-author of this research here about are we counting what would have organically been a sale anyway? If you were considering buying a Stanley Cup and you saw a DSP ad [inaudible] you bought one, did that DSP ad actually really impact your consideration? So the first thing we wanted to learn more about was how the DSP results were calculated. And then the second piece was what kind of metrics might be more illustrative, more forward-looking, a bit more comprehensive for brands who have different kinds of goals on Amazon as well, not just purely related to ROAS or profitability?
Lauren Livak:
Kiri, would you say that the shift away from ROAS is also looking at more of the effectiveness of the content and the ads and what the organizations are running, rather than what goes in, comes out? I feel like that's a big shift that we've seen from a retail media perspective and also from a digital shelf perspective. It needs to be how effective and how future forward-looking are my metrics versus just what's going in versus what's coming out.
Kiri Masters:
Yeah, that's a great point. And it was actually something that we did consider in developing some new metrics, which we've got 23. And they can each answer one of three questions. Is it showing, when you think about the customer life cycle or the marketing funnel, and the stages that we all go through as consumers to become aware of the problem, consider alternatives and ultimately purchase? One question that a good metric can answer is, are we seeing success at the current stage that we're focusing on? So if we're focusing purely on conversion, then a metric can tell us if we're being successful at conversions. If we're focused on awareness, building brand awareness, does the metric show that we are successful at building brand awareness?
The two other questions are, is the metric bringing us down the funnel towards a conversion, so moving someone along the consideration cycle? Or the third question is, does the metric tell us how the ad can be improved? So is the creative resonating? Are we targeting the right audience? So yeah, that's a really good question, Lauren, and something that I think when you look through that lens, you can think more holistically about what metrics that you're looking at. Because it's not just about that first question, are we being successful at the current stage? We can also look at things like are we moving someone ultimately towards a conversion at a later time?
Lauren Livak:
So Kiri, let's dig into some of these metrics. You said there were 23, but there's a couple that really stand out for you as you're working with your clients. And I think the first one was competitor audience acquisition. So how can brands look at that? What does that mean? How does it help them effectively measure?
Kiri Masters:
This metric calculates how many unique users from our competitors' audience we've managed to reach in a given period. And so it tells us if we're chipping away at people who've previously bought a competitor's products. And to that end, it's a great early indicator of market share. So we might see the market share ultimately play out, but we're actually getting a sneak peek ahead of time. If we're Nike, are we actually successfully conquesting Adidas customers, for example? And so this can be very effective at giving us an early preview of that.
And so this metric is telling us if we're being successful in the lifecycle stage of conquesting. And that's one tactic or strategy that we can use for brands is conquesting competitors or taking market share from them or targeting people who've bought from competing brands. Those are things that are all possible within the DSP, so it makes sense that we should be able to measure them as well. And so this is a view of monthly cumulative reach over the forecasted reach of our competitor audience.
Peter Crosby:
And so Kiri, you were talking earlier about one of the uses of these metrics is to help optimize, to help get better. Are you seeing clients take this competitor audience acquisition metric and making changes to improve that metric? Are you actually seeing people use that as an area of focus?
Kiri Masters:
Yeah, so I think with the DSP, and just a reminder that the thing that's really attractive about Amazon DSP as opposed to pay-per-click keyword-based targeting is the audiences that we can reach. So we can reach people with standard sort of demographic audiences, but we can also reach in-market shoppers, so people who have actively been searching for espresso machines, for example. We can also reach affinity segments, so young, health-conscious parents for example. And that's based off of past purchases. Amazon knows how old your children are, because when was the last time you bought diapers or when you set up your wishlist when the baby was born?
And so they know all these things about demographics, and then also interests as well and we can bring those two things together. And so we've got a variety of different strategies that we can use at the top, middle and bottom of funnel. And conquesting competitors is one of those strategies closer towards the top of the funnel, where we're attempting to draw in those audiences and show them an alternative from a brand that they may have previously bought from or are actively researching. So by looking at... This metric really tells you how successful you are at doing that, if that is part of your DSP strategy.
Lauren Livak:
Have you seen competitor conquesting be more important in different categories, like let's say the beauty category versus the cereal category? I'm just curious if there's a place where this kind of metric would be more impactful.
Kiri Masters:
That's a good question. I think rather than category, it's more of a question of the objective of the brand. And I think that we've spoken about this along the way, is being really clear on what your objective is as a brand or even within a subset of products that you sell. So if we're talking about a CPG company, for example, they may have a cash cow brand where they're just focused on profitability. It's a mature brand, people know it, and it's more about defending that brand, brand equity, and being profitable. Because that is the reward of getting to that position in the market. And really now you're defending it and making sure that it keeps returning for you.
And that same company may have a new line that they're launching or a new product that they're launching to be innovative. And it's an awareness play at that point. It's a new brand people have not heard of, there's not the affinity yet. It might be an innovative new product that we need to educate the market on, or it might be a substitute for a well-known brand. So getting clear on what the objective is there and what kind of tactics could we use based on that objective is really, really important. And I think a big trap that a lot of brands fall into is they may have some growth goals, they may want to grow market share or introduce a new product line, but they're overly focused on bottom of funnel metrics like ROAS. And that is just going to kill any innovation that you might have in looking at things like this, at looking at conquesting, because it's too expensive when, going back to the goal, the goal may not be to drive a sale that day.
It might be to get more traffic to your product pages. It might be getting prospective customers to look at your product reviews, adding products to a wishlist, things like that. Things that they're not a sale right away, but they do share... You've got someone's attention, they're now aware of you, and they will start to consider that product. So I think that those goals are universal to any category. We could be talking about pet, could be talking about apparel, or we could be talking about electronics. There's always going to be some products where it makes sense to focus on profitability given the lifecycle or market position of that product. And there's some products that are going to need to have some awareness strategies built in to help that brand or product get to where it wants to go.
Peter Crosby:
Well, this is what I loved about reading your research, is that it almost felt like the unveiling of these metrics that are possible inspires a less blunt instrument strategy. Because you can actually dig into these things and think about what it... Once you're aware of these metrics, then you have to sit back and go, "Oh, what am I really trying to do here in this moment with this SKU?" almost. I mean, it's really that detailed, right?
Kiri Masters:
Totally. And I think if nothing else, 23 metrics, at least that tells you there is more out there, and to at least get a little bit more ambitious with what you're measuring and what the data could tell you. Absolutely. I think we need to move beyond just looking at whatever the Amazon dashboard or AdTech solution serves up for you, and understand all of this data is in here. We can see how many customers who looked at your ads, their product reviews. How many of them added a product to a wishlist? How many added a product to the cart? This data is in the DSP, and we can, just by swapping out numerators and denominators, get some much better early indicators and metrics that are going to suit, be fit for purpose, which is the name of the report. It really is what are the metrics that are fit for the purpose of your brand or SKU?
Peter Crosby:
You mentioned product reviews. Tell me about the product review page view rate. I can imagine that's sitting there in the DSP and you're looking at it and you just glaze over. You just scroll right past it. But you've really put in the context of where that can have some meaning.
Kiri Masters:
Right. Yeah. Again, this is not a metric that is displayed in a dashboard anywhere. But you can source the product review page views, and divide that by the number of detail page views, and get the percentage of people who visit your page that look at your product reviews. And I don't know about you, but for me, when I start to read product reviews, I'm pretty deep in considering a product or an alternative. And so that's a great forward-looking indicator of consideration and ultimately purchase. If you've got a high-price item, you might be looking through reviews and then making a purchase at a later date. So it's a good early indicator for those high consideration items.
And so this is another great example of are we being successful at that consideration stage? Are we getting people to linger on the page and consider it and go and look for social validation from other people? And so, yes, I think that this is really just an unknown asset sitting in the DSP that a lot of brands just aren't aware exists and aren't aware that they could actually go and measure.
Lauren Livak:
I think the thing about reviews that always surprises me is that the brand doesn't always think about using it for themselves. Because there's also that gold mine of information that you hear from the reviews when the consumers are providing that, that you can then feed back into your product development cycle or how you write your content or even how you target your ads. So are you seeing more brands really adopt that methodology as well and feed that into their retail media campaigns based on what people are talking about?
Kiri Masters:
Yeah, that's a good one. I see it more on the content side of things. So if we're [inaudible] a product page, there's lots of different ways that we can provide social proof and validation. And some are through things like awards or accreditations. You've got a vegan certification, or your beauty product was an Allure winner or something like that. And then there is PR and media. And then social proof from actual customers is pretty huge as well. So that's one way I see brands mining product reviews to take little snippets of positive reviews and attributes and featuring that on the product page. And then with some of the DSP ads, I believe that one of the template options is to show the product with a five-star customer review. I think that that's actually one of the template options there.
Lauren Livak:
Another really critical piece is understanding how the campaigns are actually reaching the consumers. Are they reading it? Are they listening to it? Or is it impactful? You have another metric called audience growth. So can you talk about how brands can use that metric and what it means?
Kiri Masters:
Yeah, so audience growth is the growth of unique users that make up a DSP audience over a period of time. So this tells us to what extent our DSP campaigns are reaching incremental users over time. When you're setting up a DSP campaign, you have an audience [inaudible] especially if you have a growth objective and you want to be reaching new customers or new in-market customers, you want to be growing that over time. You don't just want to be showing the same ad to the same audience over and over again, although we do find that repetition helps, which is not a new concept. But with Amazon Marketing Cloud, which maybe we can talk a little bit more about, we're seeing all these paths to conversion where we can identify an optimal path to conversion through a combination of DSP performance advertising. And so this metric helps to understand or measure whether you're growing that top of funnel audience.
Peter Crosby:
Looking at this whole report, as I think about the 23 metrics, that's a lot, but I think what you've done with the report is you've provided a quiz that people can go to so that it can just help them understand where they're at and where they might go. Can you explain how that quiz works?
Kiri Masters:
Yeah, absolutely. Great point. And the quiz asks three questions and then spits out a handful, like six to seven metrics that are suitable for your brand given the answers. And the questions really are, what's your competitive positioning? If you're a top five brand in your category, your approach is going to be different to a brand who has either just launched, and that can be a new product launch as well, or a brand that is seeking to break in. So just in general, those approaches are going to look a little bit different. We also ask, is the brand or product setting question focused on profitability or growth? Which is important because that, to a large extent, dictates top of funnel versus bottom of funnel strategies, essentially, at a very basic level. And then the third question is around purchase frequency. So if the product is [inaudible] frequently, like every [inaudible] days or is not generally repurchased, that also will govern what type of metrics are most suitable as well.
So with those three questions in hand, we have 19 different scenarios and different combinations of metrics that are going to be most suitable for your brand. And we explain why as well. Why is audience growth important for a company that's focused on growth? Why is subscribe and save subscription rate important for a high repurchase frequency brand, et cetera? [inaudible] I'll say is take the self-assessment, get the answers back of what the metrics are, and then you won't need to bother going through all 23 of them, you'll just have a few to start with. And that'll give you an easier entry point into the world of alternative DSP metrics.
Peter Crosby:
If I could just ask you maybe how you use this framework with a client. At what frequency are you considering introducing new metrics into the program? When you think about that overall structure of how you decide to lean into, is it a quarter of testing different metrics, or how does your strategy to execution path work with your clients using these?
Kiri Masters:
Yeah, that's a great question. I think the first step is really making sure that the brand is aware of, and that we're really understanding what their objective is. And that can take some time to get to. And often I really have a lot of empathy for our clients out there who are e-commerce directors or VPs of e-comm. And they're really still needing to do a lot of education and advocacy internally about e-commerce and marketplaces. And where does the funding come from, is a simple question with a whole lot of history and answers there. One interesting thing that I think is relevant here is, depending on where the budget comes from, whether it comes from a historical sales budget or a trade budget or from a brand budget, that's going to really impact what kind of metrics the executives are going to want to see.
So if e-commerce comes out of a sales function, then ROAS and ACoS and these direct ROI metrics are what those sales organizations are really used to seeing and demand seeing. They want to see performance within a quarter. And the concept of brand building and building awareness and consideration doesn't feel as relevant to those leaders. So long way of saying I think coming to common ground of what is the goal here, and what are we prepared to give on? Because with these metrics, which generally, roughly equate to stages of the marketing funnel, it's a balancing act. If we're going after ROAS as the ultimate metric, we are usually sacrificing some growth. Because to go out and acquire new customers is more expensive. And the reverse is true. If we're looking to grow, then often that ROAS suffers at least for a little bit.
So that's really the first step, is establishing what that objective looks like and then what strategies are we going to be using for that? And then what are the relevant metrics from that strategy? So in the report, it's sort of hard to explain on a podcast, but there is a very detailed look at the marketing funnel, which I know not everyone's a fan of. But we do go through stages of awareness, consideration and purchase, whether you're talking about a journey or a funnel. And there's lots of different DSP audience capabilities within that funnel. So we talk about different strategies. We're talking about competitor targeting, for example. Each stage also has some relevant metrics that we can look at there. So really starting at objective for the product, the strategies that we're going to use to fulfill that objective, for example, conquesting competitors, and then what are the relevant metrics for that particular strategy.
Lauren Livak:
Kiri, you had mentioned Amazon Marketing Cloud. And I think that really is a bit of a game changer in what we're dealing with in the retail media space, especially with the clean room offering that they have, and how brands are really trying to look more holistically at the analytics that they can get from the retailers and the analytics that they have themselves from the brand internally. Can you tell us what you see from the Amazon Marketing Cloud perspective, where there's an opportunity for brands, and anything that really kind of stands out to you?
Kiri Masters:
Yeah, absolutely. I mean, we could talk for a long time about Amazon Marketing Cloud. But we at Acadia have been using it really aggressively over the past year. And relevant to this discussion is how DSP advertising and pay-per-click advertising intersect. And actually, for the first time, being able to see across those two different modalities how a customer is progressing through the journey, what the best volume and viewability path is, and what is the best return on investment path as well. And we've found, and it depends on the brand and the category and everything, but we found that certain conversion paths that were not super profitable or interesting on an ad level basis, when you combine those ad types, we end up with a really attractive return on investment, or much more new-to-brand shoppers, for example. So the issue with both DSP and pay-per-click advertising in isolation is that it's a last touch attribution model.
So if a customer converted after seeing... The last thing they saw was a DSP impression on Forbes, they clicked over and bought the product from that. Well, then that DSP ad gets the credit for that sale. When we can layer in Amazon Marketing Cloud, we see that that DSP ad was the last in a journey of five other ads that a customer saw before actually making that purchase. And so we're able to get a much more holistic view of the whole journey and understand, hey, that ad that we've been serving doesn't look super profitable or interesting by itself. It's actually serving a really important role in building awareness or warming up a customer. So that's one really important use case, is really understanding the return that you're getting from your advertising outside of just that last click attribution model.
There are lots of other things that we've discovered with Amazon Marketing Cloud that are surprising to our clients. Like some product, for example, that might be flying onto the radar is actually a great introductory product to a whole product line that we never even knew before. So yeah, we should really do justice to Amazon Marketing Cloud.
Peter Crosby:
We need an episode.
Kiri Masters:
But relevant to what we're talking about today, and that's also part of what gave us confidence in this model, is we could see from Amazon Marketing Cloud that DSP really contributes to building awareness and growing sales. And Amazon Marketing Cloud really can show you beyond a shadow of a doubt exactly how that's happening.
Peter Crosby:
All right, so clearly we're going to have a part two of this, which is Amazon Marketing Cloud. We're going to stitch it all together for optimization, performance and communication of the investments that our listeners are making. So Kiri, I want to point people to the URL to get this report, because it's really, really impressive work and really clear. And I think the shift in how people are thinking about the strategy and helping educate within the company, it just can power so many important conversations, I think, in an area that's still developing massively as we speak. So acadia.io/fitforpurpose is the URL. Take your quiz, download the report. And thank you so much for, A, doing this research and then coming on to talk to our audience about it.
Kiri Masters:
Pleasure, thank you.
Peter Crosby:
Thanks again to Kiri for this literal wealth of data. Go take that quiz. And then head on over to digitalshelfinstitute.org and become a member to stay up to date on all the DSI is up to. Thanks for being part of our community.
