How to Set Meta ROAS Targets Before Black Friday

[00:00:00] Tony: a large percentage of the conversion that you're gonna capture from the advertising investment that you're making right now is going to happen through Black Friday and Cyber Monday. And, and after that the lag is going to go back to more of a, a normal representation.

[00:00:16] Tony: So all of this is going to this idea around how. Aggressive, can we be with our ROAS targets at this moment in time, leading up to Black Friday?

[00:01:00]

[00:01:06] Tony: Hello. Welcome to the E-Commerce Playbook podcast. I'm your host, Tony Chopp. Richard, the professor is on PTO today, I believe. So you got me. And you got Steve. Steve, how are you today?

[00:01:19] Steve: I'm doing great. Tony you know, saw the consumer confidence, the, the future purchase sentiment jumped up to a record high Yes. Yesterday.

[00:01:27] Tony: America's feeling good.

[00:01:29] Steve: Yeah. Feeling good about that. So, turns out consumers are just holding, they're waiting.

[00:01:35] Tony: Well, you know, that's actually the perfect segue into our conversation today. Steve and I are, are, have been digging, doing a bunch of, digging into. Our favorite topic, one of our favorite topics the ideas of attribution and measurement in the, the time windows in how we measure and capture value and how that [00:02:00] relates to our spend and a MER models and our spending power and all of it through the lens of this moment that's coming up Black Friday and Cyber Monday.

[00:02:09] Tony: So. Steve, I want to, I want to get into some of the conversations that we've been having over the last several weeks around let, let's actually start on the topic of measuring attribution, measuring the full value of the media investment. And I, I'm gonna start with the exploration that I've been doing around this idea of like, how do I set my raw target today knowing that I have this big event coming up. In in several weeks for Black Friday and Cyber Monday. So we've been going through a bunch of accounts and, and looking at the difference between seven day click attribution, which is what we measure for, for most of our accounts. Or it's the optimization setting that we use for, for most of our accounts. And comparing that to [00:03:00] the reported conversion value from. A 28 day click window, which is only available as a, a reporting function in meta not, not available as an optimization goal. So it's, it's been really interesting to me in this exploration because I'm seeing something that I can best describe as symmetry. And I want to get your take on it, Steve. When we look at an account and we compare these two seven day click to 28 day click in, in almost all of the cases that I've looked at in most of the cases that I've looked at, I see this relationship. Where approximately 80%, maybe 85% of the total value capture is happening within that seven day window, and the remaining 15 or 20%, give or take is happening in the day eight, between day 28 window. And the [00:04:00] symmetry here for me is in relationship to our incrementality testing. So Steve, I'm curious if you're on the same page as me as thinking that what our incrementality testing is doing when we often find 120 or 115% incrementality. From those geo holdout tests is sort of the inverse of what I'm talking about.

[00:04:30] Tony: When we measure the value capture between seven and 28 day click as being 80 or 85%, what, what do you think about this? Do you see, do you see the same symmetry that I do, or am I chasing ghosts?

[00:04:44] Steve: Yes, I think you have the right idea. I do think incrementality is catching that longer latency from the ad spend that you're talking about the the eight to 28 day window that where we pick up an additional 20 per 15, 20% that we're seeing [00:05:00] on of. Somewhat regular basis between that seven day click and the 28 day click.

[00:05:07] Tony: Yeah. Yeah. It's, it's such an interesting thing, so, so. Now I've been approaching this, looking at it in through this lens of the difference between seven and 28 day click and you, and, and then comparing that to incrementality. Share a little bit about some of the way that you've been approaching this question around looking at the, the daily spending power and potentially some of the overlaps that we're, we're coming across in our thinking there.

[00:05:38] Steve: Yeah, so of course we're concerned with coming up with daily forecast at CTC, we're, we want to. Budget and see how we track versus a brand on a daily basis. So it becomes really interesting to us to be able to create what should be spent in a day on what we can model spending in a day. So [00:06:00] in order to kind of come up with a decent model for that. Basically what I thought is what if I took what we do on the monthly basis and get that even more granular, going down to a daily basis. So how much should I spend in a day and what can I expect in return in new customer revenue? Basically a spending power or a MER versus spend model on a daily basis. So that can kind of give you what's going on. By the day, but what it doesn't account for really as much is the latency that we're talking about, and that's kind of the one factor that we, we need to include in considering such a granular approach to spending power is what happens from the money I spent today on November 12th, further down the line, like during a Black Friday and Cyber Monday.

[00:06:51] Tony: Yeah, it's such an interesting idea. So you're so the idea of getting to daily spending power correct me if I'm misunderstanding this in any way, but is [00:07:00] around the thesis that potentially your brand has significantly more spending power on weekends. Or potentially there's significantly more spending power at you know, I, I don't know.

[00:07:12] Tony: What are, what are some of your other ideas for, like, why this would be different on a daily level?

[00:07:17] Steve: Well, I, I think some of it is obviously gonna be around product drops or events, but some of it might also be related to holidays as well, or to paychecks or to you know, like the, the Monday. Two weeks before Mother's Day actually sees a KickUp or something like that where we see an interesting gifting increase for a particular holiday, or we have spring break coming up and we didn't understand that. Oh, the Monday before spring break start beginning, people start buying more swimsuits. And then we'll see, see a tick up with brands that are in that space. So we. I think there's nuances here that we'll learn when we start looking at [00:08:00] this, that there's advantages that we can kind of, in a way arbitrage opportunities for brands to spend more on certain days and then other days when it makes sense to like, Hey, pull back a little bit there, there might not be that much available interest in your brand at that time.

[00:08:18] Tony: Yeah. Yeah. It's such an interesting idea. So. I am gonna, I'm gonna try to connect some dots here that I think are I think they're gonna bridge these two worlds. So I'm over here looking at the difference between seven day click and 28 day click. And what is that? What is that gap? And fundamentally, what, what I'm seeking from that information is can I be more. Aggressive with my ROAS targets. Can I, can I deploy more media dollars because the value capture is gonna happen later. Incrementality testing that, that gets us to a, a incrementality percentage that's over a hundred [00:09:00] percent is effectively. Accomplishing the same thing. It's helping me understand that I can push my meta ROAS targets down, or any channel for that matter. I can push my ROAS targets down with confidence because the full value capture is going to be more than what's reported in the platform. Okay. And so as, as I've been sorting out this idea for myself, I am, I am, what I'm recognizing is that. I can't, I can't combine these two ideas, right? I can't say for an account that we have that we're already running the meta meta target 20% lower because we've gotten an incrementality read. I can't go and then double add in this math from like looking at the difference between seven day click and 28 day click.

[00:09:52] Tony: 'cause at, at that point, I'm gonna be effectively. Sort of double counting that, that latent impact, right? [00:10:00] So let's let's come back to this idea of like. You are trying to look, you're trying to get to a daily spending power. How are you thinking about factoring in either the incrementality tests that we already have for or this idea of latency between seven day click and 28 day click?

[00:10:18] Tony: Like what, what's on your mind for how we don't, how, how we don't go too far and, and I'm, I'm the guy here at CTC that's always trying to push the media spend as far as possible. So keep keep me honest.

[00:11:00]

[00:11:12] Steve: Yeah, I, I think that this is a really challenging question because what I'm concerned with not. Necessarily is not necessarily what the media is going to report. So that seven versus 28 day, and I think you're right, a lot of that is captured through incrementality. There might be some kind of halo effect in friends, telling friends or word of mouth basically. Were other kind of impacts from media spend that, that, that's incorporated into incrementality factors as well, not just the latency of the spend. I think. There needs to be, and, and we're going down this path with considering the latency on Facebook because that's kind of our primary channel that we spend a lot more on, and it's also one of the best reporting channels. It, it has [00:12:00] pretty good data that I can rely on far more than some other channels. We're, I'm really concerned with trying to find out the overall impact of it of. Spend not just in meta, but in all other channels. So I'm also, I also took a peak at like Google 30 and 60 and 90 day windows. And the, by that point in time, you're, you're pretty close.

[00:12:20] Steve: It's not that much more that you're gaining by going from 30 to 60 and 60 to 90. It's a, it's a few percentage points I was seeing for brands. When I was kind of looking at that data, going a little bit deeper in that. So I, I, what I'd want to do in combination with all these is kind of come up with a decent curve for each brand of the latency of that consideration period.

[00:12:41] Steve: How much how far out does spend continue to impact you? And, and it might be indefinitely, but it's probably a small percentage point. After a certain amount of time where it's almost in, it could almost be ignored.

[00:12:56] Tony: Yeah,

[00:12:57] Tony: we, you.

[00:12:57] Steve: of spending.

[00:12:58] Tony: You and I were talking about this earlier this week, [00:13:00] and my, my, so Steve and I, Steve and I were drawn on the whiteboard and we, so Drew drew a curve.

[00:13:06] Tony: And so you know, the cutoff line on seven day click of that curve capturing approximately 80% of the value, and then the cutoff line at 28 day capturing. As far as the meta reporting is concerned, a hundred percent of the value. But we're wondering about that, that tail of the curve and, and my point, Steve was like, what do you think the value of a click was in like 2017? You know, I was like, probably pretty close to zero. And you're like, well, I don't know. So.

[00:13:48] Tony: I, I think for, for me, I just like speaking as an operator. Like these, these questions of value capture and like how we connect to measurement are, [00:14:00] it's a, it is a question of like, how much fidelity is enough fidelity for me to go do what I need to do, which is ultimately set a raw target in meta. And if I believe that seven day captures 80 or 85% of the value. And 28 day captures 90, 95% of the value. That's gonna get me pretty darn close to what I need to do with my Meta Ross target. So the, the debate's still ongoing here between Steve and I, around what is, what is the value of that, that really long term tale that that 30 to 60 day window and that 60 to 90 day window. And I'm, I'm at this point, Steve, I'm not. Yet I haven't yet been convinced by any data that it's large enough for any of the brands that we've looked at, that it's, it's going to be a big differentiating factor for how we fundamentally, how we go and set the rawest target on meta for today. But maybe you disagree.[00:15:00]

[00:15:00] Steve: I think the cumulation of all that from previous periods. Has a small impact individually, but when you sum that up to me, that almost represents what brand can be. Said as like, oh, they, they have good brand. It like adds up from all previous PR events you have done from all previous spend on all the media channels.

[00:15:24] Steve: It's because it's a summation over like a decade maybe of ad spend and promotions.

[00:15:32] Tony: Yeah. Yeah.

[00:15:34] Steve: when you edit all up, all those percentage points from the past actually wind up being significant for a brand.

[00:15:41] Tony: Yeah, Yeah, that, that's a super fair point. And I think, you know, we have we have several projects that we're, we're working on right now. That are, that are involved in like, expanding our scope of measurement to include upper funnel and, and brand activities. And I, I think this is the, [00:16:00] this is a space that is I, I'm really excited about how we push our, push our measurement, like further, like how we further measure impact.

[00:16:09] Tony: So by no means am I, like, am I like sort of out on like, Hey, I don't, I don't care about 60 day. Value capture, a 90 day value capture. I, I, I think that we, we need to continue to press into how we understand this and, and connecting the dots between what is the value capture of a click 90 days after and how do we set our ROAS target today specifically on, on meta, and how do we use. Steve's daily spending power model to think about even more specifically, how do we set our roast target on meta for this coming Saturday? I think that the, the different sort of the, the, the gap between those two ideas of super long-term value capture, like really brand media investment like ad stack. how how much are we spending over, over months or quarters? All the way down [00:17:00] to what does that mean for us today in, in our meta ads account is like, that's like prime CTC measuring philosophy. How do we connect these big ideas to these daily actions?

[00:17:14] Steve: Yeah. And, and that's a difficult problem to address because the, the stuff further out, the 90 days that the stuff beyond 28 and even the stuff beyond seven days becomes far more noisy and challenging to measure and challenging to put a particular value on in terms of what you're getting in return And, when you need to hold every ad dollar accountable there's a lot of noise that you're trying to hold accountable there.

[00:17:40] Tony: Yeah, for sure. So I want to, I want to hit, hit us with one more question here, one more idea. So let's bring it back to Black Friday and Cyber Monday. What, what, if anything, have you noticed from this exploration to daily spending power as it relates to this upcoming seasonal moment? Is there anything popping out of your. Early [00:18:00] stage analysis that we could think about trying to incorporate into this, this coming event.

[00:18:07] Steve: I think there's a, a significant opportunity, well, we saw with a few different brands that product drops or product introductions or any sort of moments leading up before Black Friday and Cyber Monday stuff in earlier November can actually have a good holding pow pattern. Like it can increase your spending power and then kind of stay pretty high prior to Black Friday.

[00:18:28] Steve: And Cyber Monday kind of keeps it, it basically. Increase the baseline. That was kind of a wonderful thing to, to see for the couple of brands we looked at. The other thing is that within kind of that seven day window, as you're getting within seven days of Black Friday, that spend is pretty conducive to creating conversions during the that special moment. So, yeah, black Friday doesn't necessarily start now, but it, it definitely starts before it does in terms of media spend.

[00:18:59] Tony: [00:19:00] Yeah, it, yeah, it's starting now for me. So I, like,

[00:19:05] Steve: Yeah.

[00:19:05] Tony: think about, for me it's always an interesting, like, juxtaposition between my own, like anecdotal behavior in life and like how, how I respond to advertising and as, as person who does this professionally and, and. The juxtaposition with the data science that, that we do. I, I have one example slide that I wanna share that we did from our our monthly client meeting earlier, earlier this month that I, that I thought was really interesting. I'm gonna talk us through this, this slide and the analysis that we put together, the very top bar represents a standard distribution, and I'm gonna call it the baseline, the total value capture over a 28 day period. Okay, the darkest blue section of the graph is one day click, and the second one over represents seven day click. And the vertical blue [00:20:00] line represents this baseline, this 85% of the value captured within a seven day click window, and the remaining 15% captured within a 28 day click window. So what was super fascinating to me about this analysis is the, the second. Chart the bar second down from the top, which is the period pre black Friday last year. And what this is illustrating here is the value shift into 28 day click. So notice the shrinking of the one day click and the seven day click, and the expanding of the 28 day click attribution. And in this case, it's, it's significant.

[00:20:43] Tony: It's 47% more. Of the value is captured over this 28 day click window versus normal. So this is, this is the illustration of the longest lag. Happening pre-Black Friday. The other thing that really jumped out to me from this analysis [00:21:00] was the bottom chart. So Cyber Monday actually represents the shortest lag between seven day click and 28 day click. And if you think about it, it makes, it makes perfect sense. The vast majority a large percentage of the conversion that you're gonna capture from the advertising investment that you're making right now is going to happen through Black Friday and Cyber Monday. And, and after that the lag is going to go back to more of a, a normal representation.

[00:21:29] Tony: So all of this is going to this idea around how. Aggressive, can we be with our ROAS targets at this moment in time, leading up to Black Friday? And my recommendation to everybody listening to this is to go and analyze the difference between your seven day click and 28 day click attribution. Establish a baseline for yourself, for your account, for how that behaves, and then go look at that behavior for for Black Friday and Cyber Monday last year to [00:22:00] see if you can arrive at, hmm.

[00:22:01] Tony: This period actually behaves quite a bit different. All in service of, Hey, we have more spending power right now in this moment in time than we do in a normal period. While, while Steve helps us refine these models and get them into our, into our CTC lexicon. That's Tony's sort of first pass at how to go think about this for for your brand for this season.

[00:22:24] Tony Chopp: So Steve, on the media side, we're looking at all of our Meta Ads accounts. We're looking at the difference between seven day click and 28 day click. We're working to understand the, the gap between those two. And considering how that affects our raw targets, at the same time we have to consider if we're already operating with an incrementality test of 120%, because that's gonna be factoring in some of this as well.

[00:22:48] Tony Chopp: So this is how we're working towards arriving at this this conclusion for what we do now, today. My question to you is, what is the, what do you [00:23:00] see as the next steps? For this daily spend in a MER model, and when can we hope to see some of this on our side?

[00:23:08] Steve: Oh, so I'm still working on the daily. I think starting in like Q1, well. We're already producing a few of them. And kind of like using that a little bit as guidance. Also kind trying to get feedback on how to improve it. But the latency that you're talking about, the difference between one and seven and 28 and then on the Google side 30 16 90, are kind of important things that I'm thinking about and trying to reconcile those versus the spend overall. With that seeing that revenue in the future. So there are different data science things. We're, we're gonna try to kind of get these models good on the daily basis, but also including some latency effect. So it'll be the summation of all the latency that you have from previous spend, and those are kind of included and and cooked into the model.

[00:23:56] Steve: So you could consider, what should I [00:24:00] spend 50 days before Black Friday.

[00:24:03] Tony Chopp: 50 days before, oh boy. I can't wait.

[00:24:06] Steve: Well, even in this case where you're like more than a little more than two weeks out from Black Friday.

[00:24:11] Tony Chopp: Yeah. I mean, I think, I think that is the, I think that is the anchoring position that, that I'm, I'm really ultimately after like we're asking and answering these questions in late October or early November. But yeah, to think about like how, how our prospecting in September ultimately affects our Black Friday is like the holy grail.

[00:24:29] Tony Chopp: I think that's like, how, you know, would, would be an am an amazing thing to be able to understand more concretely. I'm really excited Steve, to keep doing this work with you and I appreciate every time we get a chance to interact and think about how we tighten our data science and all, all in service of.

[00:24:46] Tony Chopp: The, the principle, the the, the primary goal for CTC, which is how do we create predictable, profitable growth for our brands? And for us, that means getting really sharp about measurement. So any last thoughts, Steve?

[00:24:59] Steve: [00:25:00] Yeah, that's it. It's getting sharp about measurement not just in the shorter window where it's easier to measure. It's like, now let's take on the bigger challenge of looking at longer windows to measure those and how effective our spend is in those longer windows.

[00:25:12] Tony Chopp: That's great. All right, here we go. How many days until Black Friday 17.

[00:25:19] Steve: Yeah.

[00:25:21] Tony Chopp: Go measure the difference between your seven day click and 28 day click conversion value in your account.

[00:25:27] Steve: Yeah. Hopefully we see those future purchase conversions.

[00:25:31] Tony Chopp: Thanks everybody.