Meta’s New Attribution Model Will Blow Up Your Ad Strategy, Here’s Why

[00:00:00] Richard Gaffin: Hey everyone. Welcome to the Ecommerce Playbook Podcast. I'm your host, Richard Gaffin, Director of Digital Product Strategy here at Common Thread Collective, and I'm joined this week by two very special guests, uh, one we've had before, one who's brand new to the pod. Uh, so first up we have Tony Chopp, who's our VP of Paid Media. Right. I always miss

[00:00:17] Tony Chopp: You got it.

[00:00:18] Richard Gaffin: Our VP of Paid Media here at Common Thread Collective. Tony's joining us again and our first time guest, Mr. McCay Rasmussen, who is recently promoted to a social media. What's your role again? Paid Social Manager. That's right. Okay. I gotta get these straight. But uh, so we're bringing McKay on the pod with Tony to talk about something very exciting, which is Meta's new incremental attribution. Tool that it's rolled out, uh, that or is in the process of rolling out into its ad product. Um, so honestly, this is sort of like a breaking news type of segment here. We got Tony McKay on to talk about specifically what this product is, um, how it's gonna change things, and then maybe some specifics of how Tony McKay have been running some tests with this tool for our clients.

So, uh, I'll kick it over to you to start, Tony, talk to me a little bit about what. This product is, uh, and then we can kind of go into the implications.

[00:01:17] Tony Chopp: Sure. Thanks Richard McKay. Great to have you here. Super excited. So it wouldn't be a a pod, a media centric podcast if we didn't talk about the idea of incre mentality and. I, I've been sort of exploring this idea in at CTC for like the last year, and I keep racking my brain to come up with like a metaphor.

'cause it's like a confusing idea. So I, McKay and I were talking this morning. I, I had one. I don't like it. It's pretty clunky, but I. This idea of incrementality measurement about sort of validating the results of a media investment. I, I had a phase in my life where I was like, um, paying attention to my weight and I was using a scale at home.

Uh, and, uh, and then I, and then I happened to have a doctor's appointment and I, I went to the doctor's appointment and it turns out I, I weighed quite a bit more

[00:02:10] Richard Gaffin: Mm.

[00:02:10] Tony Chopp: than,

[00:02:11] Richard Gaffin: Yeah.

[00:02:11] Tony Chopp: and, and it's a bit of a clunky metaphor, but, but I think it represents this idea of. Sort of maybe an unfamiliar idea of like, we look in these media platforms and they give us a measure.

They say, this is the, this is the return that you're getting, the equivalent of my weight on the scale. And when I went to the doctor and I used their more, I, I'm assuming accurate scale. Um, I, I got a, a different reading, which is a little bit, a little bit jarring, I think was probably one way to describe the experience.

So, so CT C's been on this journey for, for the last year or so of. Doing our own incrementality testing and, and the way that, the way that we execute this is with geo holdouts. So we run the media in one area. We don't run the media in another area, and we measure the, the impact on the business results on, on revenue.

And basically what we're trying to ascertain is if a media platform is telling us we're getting a two to one return, how does that compare to what we measure in these tests and the difference between those two things we call the incrementality percentage? Okay, so if the media platform tells us two to one and in the incrementality test we get the same result, two to one, we would call that a hundred percent incremental and effectively, if I would go back to my, my doctor's office analogy, that would be like my scale said the exact same thing as what happened when I went to the doctor's office.

Okay? So, um. We've been doing all these tests and there's a, there's a connection between what you, the settings and the tactics that you deploy in a media account and the incremental output of that. Endeavor. Okay. And one of the, one of the questions that we, um, that we're asking and answering all the time is, um, which, which attribution setting do we use in a media platform?

Okay. And CTC. Typically, our, our best practice is to use click only attribution. And, and most often we use seven day click attribution. Okay. So we, we don't use view attribution typically. The, what we see when we do the incrementality tests is seven day click reported in platform. When we do the incrementality test, we see a greater business result, so we'll get an incrementality reading that's above 120%.

Okay. But let me connect back to what, what we're here to talk about today, because the incremental attribution feature or product within the Meta Ads platform eliminates that, that whole choice. Do we use one day click? Do we use seven day click? Do we use seven day click plus one day view? It's, uh, an option that we can choose instead of any of those.

And it's, and it's really, I think it's really fascinating because, uh, what, what we're after fundamentally is pretty simple. Uh, we, we do all this measurement stuff to get there, but what we're after fundamentally is to invest the maximum amount of media dollars at the highest incremental contribution.

[00:05:36] Richard Gaffin: Mm-hmm.

[00:05:36] Tony Chopp: think this, this tool, if it turns out.

To measure well and work effectively for us gives us an opportunity to. Spend less time figuring out, Hey, should we use one day click or seven day click, or seven day plus one? Like, which one of these should we use and spend more time, spend less time on the measurement and the setup, and spend more time on what we know is actually the most important thing to do in, in any advertising platform for that matter, which is feed it fuel, feed it creative.

Okay, so we're gonna talk much more specifically about. How, how this product is, it's not, it's not just a different attribution setting. It's like fundamentally different way of measuring and, and attributing value. So we're gonna get into more specifics. Um, and, and the last thing I wanna add is we're really excited, uh, to be testing this on two accounts and we have a couple, uh, some early stage results to share.

Um. And it should be available. It, it's recently rolled out to all advertising accounts, so it should be available for, for anyone to test as well. Um, McKay, let me kick it over to you. So that was my super long preamble, but gimme your 2 cents. I.

[00:06:52] McCay Rasmussen: I think, um, I grabbed this from in platform. So straight from the horse's mouth, from meta Incremental attribution enables optimization and reporting based on incremental conversions. It uses sophisticated model, uh, model. To predict the incrementality of conversions. And this prediction determines how much to value conversions and optimizations and reporting. So a lot there. Um, but essentially meta is just using its modeling tools to decide whether or not the, the purchase it's going to attribute, uh, was, was incremental. Um, so would the revenue have happened without serving the a. And I think it's really interesting to, to get a gauge on whether or not this will become the new standard.

Um, and, and what level of reporting this gives us in data. so.

[00:07:44] Tony Chopp: Yeah. Yeah. And Richard, I want to jump into this sort of one more piece of this, this, these questions that we ask and answer ourselves when we're, when we're setting up campaigns. So there, there's a trade off. Uh. In the old, in the old settings of one day click or seven day click, or seven and one, there's a trade off between those two spectrums.

Think about one day click as the, the most conservative, the tightest way to measure, uh, think about, um, seven day click plus one debut as the the most open, right? And so let's, let's talk about some of the pluses and minuses on either side. One day click the. Tightest attribution window is likely gonna be the most incremental.

We'll see compared to incremental attribution, but historically likely to be the most incremental. But in order to get the necessary conversion volume is gonna require a significant amount of, spends, significant amount of budget and conversion volume for the algorithm to optimize. So that, that's the, the strength is, um.

Solid reporting or reporting that's likely under reporting business impact. The limitation is you need tons of volume to make it work

[00:09:04] Richard Gaffin: Mm-hmm.

[00:09:04] Tony Chopp: all the way On the flip side of the coin, seven and one, the widest open, so you're gonna get the most signal from, from that. Um, but the challenges, uh, are gonna be, um, potential for lower incrementality.

And I think McKay and I were chatting about this this morning. One of the things that we observe in accounts when we, when we do have seven and one is the amount of conversions that are attributed to view conversions can vary really widely. I. Right. So if we have, uh, we've, we've seen it, um, for large accounts in some campaigns be 10 or 15% of conversions attributed to view in some instances.

Mc McKay mentioned an account he, he was working on recently that had, um, over 50% of the conversion volume was, was as a result of the view conversion. So. If you think about it, that presents a really difficult mana, uh, measurement challenge when we have to go dig one layer deeper on the seven and one and see, well, how much of that is view conversion?

So we're, I'm really excited about, about this product because it, it aims to address that challenge hat on and say, Hey, you guys don't need to, um, we don't need to suffer either of those limitations. Instead, we can have this product that sits in the middle. So that's one thing I'm gonna shut up. Mikay.

I'll let you, I'll, I want to go to you in a second, but I have one more point to add. Another challenge for us when we're, when we're, uh, doing our setups and we're thinking about how to structure accounts and structure campaigns is around this idea of new versus returning customers, right? All sorts of challenges contained with this, within this idea.

Like, number one being like, how. Effectively, can we actually exclude existing customers using the pixel, using email lists? Um, there's all sorts of evidence that we see that our, our ability to fully exclude existing customers is, is actually quite limited. Another piece of the new versus existing customer.

Equation is, should we have different ROAS targets based on the incre incremental contribution of either of those cohorts of customers? Well, here's another place where this product can sit in, in the middle and, um, have us be, um, spend less time on asking and answering that question and more time giving meta input about the business outcome that we need, regardless of whether it's somebody that's purchased from us before.

Or is new to the brand, what do you think about that? McKay Less, less having to decide on seven in one or seven day click, and a potential world where we, um, you know, ha have true like a CQ campaigns that, that are, we're not, we're not doing, um, any sort of hard exclusions. What sounds lovely, huh?

[00:12:01] McCay Rasmussen: I love that. I hope, uh, I hope we get to a point where in platform we're just looking at IRO as rather than, uh, in platform roas.

[00:12:08] Tony Chopp: Yeah.

[00:12:09] Richard Gaffin: So actually that that kinda like leads to, I had a couple of sort of layman's questions I guess about this and sort of understanding this incremental attribution setting. And it sounds like you've kind of maybe just summarized what you've laid out is there's obviously there's one day click and then there's seven in one. in a sense it sounds like incremental attribution. The idea there is it would replace it

[00:12:28] Tony Chopp: It does.

[00:12:29] Richard Gaffin: sort of an algorithmically judged. Like the algorithm, judging where on that spectrum, the attribution setting should be for the most accurate result or the most incremental result. I guess that's where I'm sort of struggling to understand like, what, what, what exactly is the mechanism here, I

[00:12:48] Tony Chopp: Yeah. Let's, so let's talk about it. 'cause it's actually quite fascinating.

[00:12:51] Richard Gaffin: Okay.

[00:12:52] Tony Chopp: so. Yeah, it's it. Okay, so we're gonna get into the, the technical details as, as far as we're understanding, we're still learning about the product, but we're gonna get into the technical details so far. So all the old stuff uses, um, time-based attribution.

So it looks for conversion, uh, within seven days within one day to include view or not to include view. This doesn't work that way. There is no time-based consideration. Instead, what. What it's actually doing is running, uh,

[00:13:24] Richard Gaffin: Oh.

[00:13:25] Tony Chopp: holdout tests continuously under the hood. So we don't see any of it. So you can't see any of the, the results of the holdout tests, where they're running, um, the holdout sizes.

It's, it's all sort of. Behind the, behind the curtain, if you will. But functionally that's what's happening. And what they're doing, what the algorithm is doing is showing, um, ads to a control, uh, not showing ads to a control group and showing ads to, um, a uh. I forget the, the word, but, uh, treatment.

Treatment group. Okay. Happening continuously. And it's looking at the, um, conversion rate of these two groups. So the control that the group that doesn't get served ads, let's say it has a 1% conversion rate. The group that does get served ads has a 1.5% conversion rate, and it's considering that 0.5% as the incremental lift.

So f fundamentally different way of doing attribution, not time-based, but rather looking at and making, uh, an ascertain around did this media actually have an impact on converting these users and happening continuously and, and under, under the, behind the curtain, so to speak, really, really different.

[00:14:39] Richard Gaffin: Yeah, that's interesting. McCade, do you have something to add there?

[00:14:45] McCay Rasmussen: No, I think Tony covered it. Um, yeah, I think the, the main nuance I'd say is like when advertisers start to split test this, um, just to keep an eye out that your in platform wise will likely look lower on the incremental attribution, but that's, that's as expected because it's going after incremental purchases.

Um, so that makes sense.

[00:15:06] Tony Chopp: Yeah.

[00:15:07] Richard Gaffin: I is is the rollout here then, like, and it sounds like you were alluding to this a little bit earlier, Tony, but that really, this is the first step towards just ROAS as a metric call together with IR Oass or just a better roas. 'cause that se sounds like, it's like, why would you not have an incremental attribution setting?

[00:15:25] Tony Chopp: Well,

[00:15:25] Richard Gaffin: maybe there's some, some way in which me is incentivized not to give everybody, or have everybody use I oas. But tell me about like what, what the, yeah, what do you see rolling out here in the future?

[00:15:34] Tony Chopp: well,

[00:15:35] Richard Gaffin: Mm-hmm. Yeah. Yeah, exactly.

[00:15:53] Tony Chopp: much money you, here's how much money you put in, here's how much money you get out, right?

So that's like the sort of logical extreme of this thing. Um, but more to your question, Richard. Yes, this represents a step on that journey of saying, my business makes money if my media investment returns at a two to one. I don't care if it's an existing customer, I don't care if it's a new customer. I don't care if it's somebody in Florida or Texas or I don't care what product they buy, but.

Obviously there's nuance in that margin, et cetera. That's a, that's a whole nother conversation that we need to get into. But, but yes, Richard, this, this, this represents that, and I think to the point, to the points that I was making earlier, like right now, we, you know, we're over here on the media team at CTC, just like

asking, spending a lot of energy on questions like seven, eight click, or seven, eight, click one day view. New customer campaign versus existing customer campaign percent, percent cap, like, which we can't even include in the, in the newest campaign. So we're spending all this time and energy on,

um, what's the word I want to, what's the word I want to use? Like steps along the journey

[00:17:20] Richard Gaffin: Mm-hmm.

[00:17:23] Tony Chopp: where the platforms are going, what, what we actually know to be true is we wanna spend more time and energy on inputting fuel into the system. Giving the system clear indicators of what, what the outcome is that we need.

[00:17:35] Richard Gaffin: Yeah. So I mean, it sounds like basically what you're saying is it removes. More elements of the kind of more, uh, opportunities for human error. Those are kind of getting subtracted again from the equation in a sense of saying like the media buyer making the decision around what attribution setting makes sense across what time window, and sort of spending brain power making tweaks there.

That's something that ostensibly is going away if this works out. Uh, plus kind of thinking about, yeah. Hard excluding or only going after new customers with this campaign in existing with that, whatever, like those, that's, those set of choices is no longer something that's necessary. Is that kind what you're saying?

[00:18:13] Tony Chopp: Y Yes. And

[00:18:15] Richard Gaffin: Alright, let's go.

[00:18:16] Tony Chopp: like meta with this, with this shift, meta is functionally bringing. Let's call it, uh, causal attribution in, into the algorithm itself. So when you think about things like, um, new versus returning or, you know, seven versus one. All of those choices are made at a, a moment in time, right? Based on some information that we have, based on some tests that we've run recently and whatever.

And, and then we just, we sort of continue to operate on, on that understanding until we get, until we get some sort of better understanding. Right. The thought that comes to mind is like, um, you know, we have some experiences where we will have like a big sale moment, um, and we'll use like a. Like one day click, like highest value sort of set up to just really pump some spend through, because we know, we know something about the business in, in this moment.

Right. So

suffice to say, bringing, bringing continuous testing like as a mechanism to feed the bidding algorithm. Is a fundamental shift from how we do things and how, how every media, media person has done things historically where we, we set these things at points in times based on our, you know, what's Taylor calling it?

Our ci ci circadian.

[00:19:54] Richard Gaffin: rhythms. Right. Well,

[00:19:55] Tony Chopp: Yeah.

[00:19:56] Richard Gaffin: the, the media buyer or, or rather, these sorts of decisions no longer have to be made on. According to human rhythms, right? And if they can be made the way a computer can make 'em, which is to say constantly all the time and with the best possible information, then this is just gonna create a much more, uh, hopefully a much better scenario for everybody in terms of both like workload and then results. Um, but I'm kind of curious to talk then about what our actual interactions with this has been so far. So I know, McKay, that you're setting up some tests with this on a couple of client accounts. So why don't you walk us through like what. those tests look like and what the initial results have been.

We can go to YouTube first,

[00:20:35] Tony Chopp: I, I, I just wanna throw one more thing in before we go to the test, design and the setup. So one, one other thing that's really fascinating about this so is, okay, so this brings like causal, um, attribution into the algorithm. Another thing that we're reading and learning about is, um, that it can and should be informed by other tests that are running in, in the account.

So, um. The idea of having. Like a conversion lift test in the account from some previous iteration of testing, actually being able to inform this incremental attribution. Um, the way, the way that we're, we're reading and understanding this, it's like a, a suite of testing tools that work together that provide meta with the best available information to optimize for the business result that you're going for.

I think that's kind of a cool, like part of how. Um, they're thinking about how this product integrates to the broader suite. Um, but McKay, go ahead and tell us about, uh, what, what we're cooking up over here.

[00:21:42] McCay Rasmussen: Yeah, absolutely. I think the only other thing I'll add is like, to your point, Tony, on the incremental attribution, uh, always running under the hood is like yes. We run, let's say like quarterly. Um, incrementality benchmarks on our standard attribution campaigns. Um, but that's quarterly and the idea is like, in theory, the incremental attribution is doing this all the time on every campaign. Um, so yeah, really interested to get a read on this. Um, right now we have this set up for two clients. Um, we just did an in platform conversion tests. So two cells, essentially. Cell one is standard attribution. two is incremental attribution and we're just trying to get an idea of the overall lift between the two. Um, because we wanna see if, if incremental attribution is doing what meta claims is, is going after incremental conversions. so I'll pause. Does that make sense or do you have any questions? We're gonna test that up.

[00:22:35] Richard Gaffin: Yeah. Yeah, no, so it's, so you have one set up with one attribution setting, one set up with the other, or the standard versus this new incremental, and then my supposition then, I guess, is that you're running incrementality tests against the incremental attribution campaign in order to determine if, as you said, meta is actually. This is functioning properly or if they're just sort of blowing smoke at this point.

[00:22:58] Tony Chopp: In progress. So this, this is stage one, uh, is the ab split in in the meta ads account. Um, Steve is working on the test design for CTCs incrementality testing tool as, as stage two of this journey,

[00:23:12] Richard Gaffin: Gotcha. Okay. so then talk

[00:23:15] Tony Chopp: I.

[00:23:15] Richard Gaffin: a little bit about, I mean, I know I. We don't have the final results of, of these tests right now, but just a little bit initially about what you're seeing in terms of how it's functioning, how it's performing, all that type of thing.

[00:23:27] McCay Rasmussen: Absolutely. Um, so right now what we've seen is essentially incremental attributions falling between one click and seven click. In terms of. Um, so it's interesting. I mean, basically one day click being the most incremental you could get. Again, Tony, the down. Tony mentioned earlier, the downside of that is, um, meeting higher levels, higher levels of spend and volume, um, to make that work essentially. Um, we typically would see that running for sale days when we have, uh, insight the algorithm doesn't. Um, and then seven day click, uh, which is like our standard, uh, BAU right now. We're.

[00:24:11] Tony Chopp: Yeah. Yeah. I, we have a chart too, Richard, that we can share, share with you. So for, for anybody that's watching on YouTube, it, it is really fascinating to see. Um,

so let's see. Um, we have it running for two accounts right now. Um. We'll talk about, um, I'll just hit you all with some numbers. Um, we have it running for a golf brand and the seven day click Ross is 1.5. And the one day click Ross is 1.1. And right now we're, we're getting 1.3 out of the incremental attributions.

So it's like literally right in the middle.

[00:24:53] Richard Gaffin: Hmm.

[00:24:53] Tony Chopp: Um. And almost a, almost an identically similar result for a, a pet products brand. Um, seven day click at 0.75, one day, click at 0.4 and incremental attribution at 0.5. So like sitting, sitting right in the middle.

[00:25:10] Richard Gaffin: Interesting. And we'll, we'll, we'll share that visual, uh, in the, in the video version, this final version here, so we can kind of review it. But, um, are there any, do you feel like there's any conclusion, I know it's really early, but any conclusion at least initially to be drawn about what this, this result means?

[00:25:26] Tony Chopp: Well, um, I think the thing that we're, the question that ma the burning question that McKay and I are a asking is, you know, we, we have a. We have a default setting, uh, that we apply to meta for our accounts in, in lieu of having, um, of doing incrementality reads, which, which, which we always do to sort of refine our defaults, but we have a default setting.

That seven day click is, um, has a positive incremental percentage depending on, uh, what it's reporting in per in, in the account. Um, and this would. This result would actually be, we, we would've to tangle with this result because it would be giving us a little bit different signal being the incremental attribution and platform is reporting lower.

So not to say that we haven't seen that, um, or, or, um, found that result. We find, we find variants in incremental contribution depending on, um, the brand and, and when we run the test. But. So I think that's the first thing that stands out is, and we're really excited to go look at the, once we get through these in platform tests to go do our second round of tests using our, our incrementality testing tool and status.

Um, but the other thing that sort of pops out at this, about this for me is, and I think this like connects back to this idea of. Just not having to choose an attribution setting like seven, seven and one, or having to sort of decide new versus existing customer. Uh, if, if we develop a lot of confidence in using this tool.

Um. And, and the measurement that it's outputting. I, I'm excited to go, uh, totally nuts with account consolidation and inflated budgets and, you know, let it rip sort of thing. I, I guess one caveat to that is, McKay correct me if I'm wrong, but we, we can only run highest value right now. Is that, is that correct?

[00:27:33] McCay Rasmussen: Yeah, we can only run highest volume right now. It is, uh, I've heard from some other reps that cost controls with incremental attribution is coming later this year. So hoping to get, uh, on the alpha stages for.

[00:27:45] Tony Chopp: Yeah, and this is the same thing happened when, you know, when they rolled out a SCI think it's really, it's typical for meta to, um, when they launch a new product to, to, uh, focus on getting volume through the new product, um, prior to adding in adding in cost controls, but. Um, but assuming we, assuming we validate the measurement, um, with our tools, so, uh, trust but verify sort of idea.

Um, you know, I, I think this represents a, uh, opportunity for us to continue to push into the idea of consolidation, continue to push into the idea of having a lot of, uh, confidence in inflated budgets. Assuming we get to, um, a product, um, an additional feature that that gives us some cost control.

[00:28:34] Richard Gaffin: Yeah. So let's, uh, we'll, we'll keep everybody kind of up to date on how this goes. Again, obviously, it sounds like there's two phases of this test. We'll keep you updated on both, just so you can kind of get a sense of how we are, uh, our understanding of this tool is evolving and how our use of it is evolving. Um, but, uh, before we go, I, I just wanted to say a. Is there anything else you wanna hit on this? And B, what I usually like to talk about at the end of these conversations is like, what is the one thing that you as a business owner should do now relationship to this tool? So first I'll like open up, is there anything else kind of details about the tool, ways we're using it that you think we missed that, that you wanna talk about?

Tony?

[00:29:14] Tony Chopp: Um, I don't know. I mean, I kind of already hit this point. I think some of the criticisms that I've seen online of the tool is like it's black box. Like, you know, we don't really understand. Maybe you said another way, like it's hard to generate like, actionable insights from it.

[00:29:31] Richard Gaffin: Hmm.

[00:29:31] Tony Chopp: Um, but I don't care. The, the actionable insights like are, don't go make, like, make less tweaks and accounts is kinda like what Luke was talking about on podcast recently.

Like, that's not where the value is. The value is in, um, gi giving the, the media platforms more fuel to work with. It's all, it's it's product, it's marketing, it's offer design, it's creative. Um, it's not attribution setting. So, so I'm, I'm excited to, um, for us to, to move further in that direction.

[00:30:00] Richard Gaffin: I mean, it seems like if, if this works out, this would be an enormous, I mean, it would be a relief in a lot of ways. I think in terms of just like out a lot of the decision making fatigue, I think that comes into media buying, but also just like getting an accurate result and having it be easier, I think is gonna be better for all of us.

So, um, I. Back to my second question then. So what, what do you think, like as a business owner, if you're listening right now in relationship to this, it sounds like it's been rolled out widely, or at least across almost all accounts. What should you be doing with this right now?

[00:30:30] McCay Rasmussen: I think because it's available to everyone, uh, the first thing I'd be doing as a business owner was, uh, a if you can get like a clear incrementality test, a true incrementality test going, I would do that. Um, just to get a read versus your standard attribution. Otherwise, uh, using. Metas and platform tools.

So just the experiment tab and setting up a two cell test, um, would be a good start. Just so you can get a gauge of, of how impactful this new incremental attribution is.

[00:30:55] Richard Gaffin: Yeah, makes sense. Tony,

[00:30:58] Tony Chopp: Yeah, yeah, same. Give it a shot. Um, met, met is hosting their annual performance marketing Summit, um, next week. Uh, so I, I recommend everybody watch that there, there's gonna be a lot of conversation about, from what I can tell about, uh, measurement and, and incrementality and I, and I'm sure this product, so I.

So that'll, that'll set you on the right path. But yeah, to McKay's point, um, give it, give it some touches and, um, put it, put it into an AB test. And McKay, I liked your point about like, if you're using a tool like House or Measured, or if you have an incrementality testing platform, like getting those reads before or, um, after, at, at some point during this journey to, to get a comparison, like, you know, Tru Trust, but verify sort of concept.

Um, that, that's, that's the way that we're doing it.

[00:31:41] Richard Gaffin: That's right, folks. Yeah. Trust, but verify. That'll be the, the takeaway from this conversation, I think. But wise words. Well, folks, thanks for listening. Tony and McKay, really appreciate you joining us. We'll probably have you back on to discuss how this has gone in the next, uh, few months here. Um, but until next time.

Thanks everybody for listening. Take care.