Snapchat Ads and Measurement with Northbeam

Introduction to Measurement Challenges

0:00Brands tend to use various sources of truth for their measurement and if they have been burned by view data in the past, maybe they're using last click. Something like last click doesn't really represent any incremental impact that a channel like Snap or other secondaries have. >> 90% of our engineering resources over the last two or three years have gone exclusively to building optimization engines and signal environments that are going to drive not only click attributed conversions, but like last click. It's been our number one mission the last two years in a row.

0:31>> So we're trying to just have a simple system where it's conservative omni-channel safe views where you're not going to get burned by them and then use those to optimize your campaign ad set and ads. >> All right, check the mics. We good? >> Check, check. >> Hello, hello. >> Awesome. What who who who am I blessed to speak with here?

Meet the Experts: Ryan and Gabe

0:52>> Okay, so my name is Ryan Kovach. I'm at the head of media strategy over at Northbeam. How many folks here have heard of Northbeam, know what we do? >> A quick >> That's a lot more than I thought. I'm not a lot. >> I'll take those hands. So for everybody who doesn't know, we're first-party multi-touch attribution and that's the short pitch I'll leave you with. It's better data not cuz it's more conventional, but because it better maps to marketing realities. So thanks for having me and and then Gabe, I don't know how to say his last name, but I'll let him intro himself.

1:24>> Yeah, hey everybody. My name's Gabe Piacentini. I'm the business development

Understanding Snapchat's Role in Advertising

1:27lead for our performance and lead gen business at Snapchat, messaging, social media. You know it, hopefully you love it and here to talk a little bit about how to run ads on there as well because as you said, it's a great additional channel for a lot of commerce operators. >> So I didn't know this first, but Gary said he leaned over and said like, are you using Snapchat? What was he What did What was I don't I have no idea what that was. Spotlight? Yeah, what is that?

1:51>> Oh, Spotlight? So Spotlight is one of the mechanisms that if you want to post publicly as a story, you know, video, etc., it can go in. It's a vertical video scrolling tool. It just hit 100 million users daily on that, as well. So, it's just another way that people are consuming and posting content there, as well, that people really have responded well to. >> Okay. Who's spending on Snap right now? Is it Ouch. Is it No, no.

2:17People are spending on Snapchat right now. Who's spending on Snapchat right now? Jeez, this entire room could actually take this conversation, then. >> Perfect. Why? Why Why do you think that is? Why do you think people are not spending as much as they should be spending? I'll go both of you.

Challenges and Misconceptions in Ad Measurement

2:30>> Yeah, yeah, yeah. I I I have my guesses. I think measurement's a very interesting space right now, and brands tend to use various sources of truth for their measurement, right? And if they have been burned by view data in the past, maybe they're using last click. Something like last click doesn't really represent any incremental impact that a channel like Snap or other secondary have. So, so my my head goes there first. >> That might be it, because I don't know how we're measuring. Like, it doesn't really fit into the ecosystem of how we're measuring all the other things we're spending on. Do you think that's similar, too?

3:05>> Yeah, so I mean, I think it's twofold. So, first, in the macro, I think Snap gets forgotten about a lot. We're 13 years old. We're not the shiniest tool in the shed anymore. simultaneously, what Snap is is it's primarily a messaging service. It's Snapchat. It's It's in the name. And then, secondarily, social media. So, you're not going to have that viral Snapchat that people are talking about, the water cooler that they're sending around, "Did you see this?" So, top of mind, number one, but the reality is like, the users are there. We're around 100 million daily active users in the United States every day, which blows people's minds. That's like, a quarter of the country.

3:42so, the the audience is certainly there. And then, in the micro, I I will say I think we had some missteps in how we you know, measured a lot of our media, particularly when it came to, some of the last click items, but we we really made it a very, very intentional thing over the last 3 years to ensure that we are going to be able to drive click for focus conversions for our advertisers that could be measured in, you know, third-party analytics, MTAs, etc. I know a lot of people came on Snap when it was more that shiny toy, maybe didn't see the last click results they wanted, and, you know, now we're, I don't want to say apology to her, but, you know, we got some ground to make up, and, people like Northbeam have been really great in helping us showcase that.

4:26>> Yeah, I think that makes a ton of sense. I'm interested, little bit of show of hands for for the eyeballs we do have. How many folks have been burned by view platform attribution? Where they were like, "Hey, I'm seeing something look incredible. maybe I scaled some money into it, and then the reality was it didn't pan out." I got a couple hands in there, but does that happen to some folks? Yeah, so I I I'd say that's sort of the people who've raised their hand there nodding along, too.

4:53They're the type who've been burned by view view platform before. I think what what you're saying over at Snapchat is, "Hey, the machine underneath the hood is going to drive non-fluffy performance now. It It knows we want to see click data actually, as well." >> Oh, 100%. I think 90% of our engineering resources over the last two or three years have gone exclusively to building optimization engines and signal environments that are going to drive not only click attribu- attributed conversions, but like last click, it's been our number one mission.

5:26We actually saw median last click ROAS go up about 160% the last 2 years in a row. >> Is that because of the users growth, or the because of the more activity on the platform, or just it's just the machine is getting smarter? >> Machine. The machine and then the signal environments. One of the big ones, like, good signal equals good performance on Snap now. One of the other big things is like we integrated with Shopify for copy and pixel data. So, like anybody using that in a couple of clicks is able to send really good data in. Previously, the pixel was kind of a pain. So, that we're getting all of those signal points that we need to like actually drive the the business forward.

6:03>> A lot of the conversation that we have across our team are like, where we can allocate media dollars is based on like, well, if I know from running on Instagram, if I know if I'm running on Meta, if I know if I'm running on these other channels, my ads and my content will live on cuz it doesn't disappear. Now, real inventory on the other channels as well as whether it's Meta, whether it's Facebook, whether it's Instagram, you're you guys have been there for a very very long time. Is the users of the stories continuing to grow often? And are ads going to be like this is the thing that pissed me off the most on Meta. You get one ad and there's like four in a row.

6:38There's no way you guys are doing that as well, are you? >> Yeah, just to like, hey, we're going to vulture on this conversion or click quickly. >> Yeah, literally. >> Yeah, you like you're not generating. I think that's one of the big things we want to talk about is like the growth versus the efficiency. It's like, okay, maintaining the two of those. I would say specific to Snap, one of the other

Snapchat's Audience and Performance Insights

6:53things that surprises a lot of people is that we're thought of as the teenybopper platform. >> Yes. >> Right? but now we dominate that audience like, you know, 90% reach in like that 13 to 24 in the US. 54% of that audience is now above the age of 25 and almost 30% is above the age of 35. 35 plus has actually been our fastest growing demographic the last 3 years in a row and it's twofold. One, we're 14 years old.

7:21So, if you literally had a child, like you'd be almost 40. Right? That's how it works. It's the reverse Facebook problem. Our audience is like increasing in value for a lot of brands versus, you know, Facebook starting to, you know, get too old. And then two, we hear all the time that people who want to get in touch with their kid, their niece, maybe a a soccer team they coach, Snap is like the best place to do it. And whereas 5, 10 years ago, they'd be like, "What is this weird yellow app with puking rainbows?"

7:49They're like, "Actually, this is pretty sweet." And they stick around. So, like, the audience definitely getting more valuable. Definitely, you know, having more of that purchasing power. It's not just like whimsical trinkets to teenagers anymore. So, twofold in that audience. And then, like we said, that that really high-performance engine, that focus on being able to drive click-focused attribution and ROAS has been that other component that we're creating meaningful engagement, not just one that's like playing system. I know we're going to talk about some of the different measurement setups. You don't want to just play the system and get credit. You want to like actually drive meaningful engagement that that results in those conversions.

8:30>> Makes a ton of sense. I actually want to unpack something that that you had mentioned, right? And part of the topics I want to hit on today is like where are we as state of the union of measurement as a whole? How can we still have some secondary channels? What should we be looking for? What should What pitfalls should we try and avoid? But I know you have brands winning on Snapchat right now or any other secondary channel. And you said something interesting as we started. You mentioned it doesn't fit perfectly in like a way you analyze data.

8:59How do you kind of make decisions when two different measurement lenses are telling you two different things? How did you guys kind of scale up Snapchat in this instance with some confidence? What measurement lenses helped you there? What measurement lenses didn't help you there? I'd love to hear kind of how you think about what lens makes the most sense given you did scale up that Snapchat with a lot of success.

Effective Strategies for Ad Campaigns

9:21>> There's two gentlemen sitting in the front row that run our media on a gross gross scale that I love tremendously. We we do hourly reports on CAC spend per channel. And and like new customer CAC, overall ROAS, MER, we do it hourly every single day. So, we know to the T where the dollars are going, how we're spending and what's happening. And we're we're looking at we have blended CAC at the target that we want. Under 50 bucks, we can spend as much as we humanly possibly can. So, like from a tool measurement standpoint, we're across the board. We have Northbeam, we have Converge, we have Triple, we have all the different ones.

9:57But, at the decision that they're making is still going to be in the platform. Yep. >> Yep, still need to optimize in there. How do you think about you mentioned a $50 CPA. how do you think about the relationship between holistic outcomes and channel level performance, right? You guys are scaling that up, you got that $50 CPA. Is there any kind of referencing of the the holistic outcome in that thought? >> It's changing a little bit because we're we now the where our media spend is is we spend majority on subscription-specific offers or subscription-first offers. And so, it's actually a bigger conversation of LTV of how these how long these people last with us. So, it's like revenue per recipient on the page. And then it's how long are they actually going to stick with us cuz it's now it's actually a bigger issue in my eyes, maybe not in their eyes, more in my eyes when I'm looking at this thing cuz I'm on all the retention side of things. You know, my

10:52guys we just scaled an offer. The acquisition was incredible, it's the cheapest CAC we've ever had, but these guys are churning faster than ever before. >> Are they still reasonable customers for us, right? Are they just going to churn out and we actually don't care about the cheaper CAC? >> It's important cuz it's some platforms have we've learned in the past don't drive LTV over time. They might drive initial purchase. And some of the brands that are at least at our size when we're spending like a million to two million a month, you got to make a [ __ ] decision. Like, where is this money going to go? So, I do I do think our it's a day-to-day thing, week-to-week almost.

11:26>> Love it. have you ever considered testing any like incrementalities in Vogue a little bit? Do you guys ever do any incrementality or ever look to go validate any of that action or >> Soon. So, this is Q1 of us cuz Q right now we're in a cash crunch that we need to like be intentional around. So, like we're hoping Q4 our ad platforms shout out to ad loving I'm sorry. We'll we'll we'll figure it out. I'm sorry guys. But everybody else like I want ad credits. Like if you give me ad credits I will I will commit dollars to you guys in this Q4.

11:55>> Leverage is leverage baby. >> Noted. >> Yeah, Gabe do your thing. part of the reason I wanted to ask that line of questioning is because one of the things we've been working on and I don't want to sit here and pitch too much is like trying to bridge the gap between what's actually driving growth in that, right? So, sometimes I'm curious how you guys both think about like actual incremental impact. Sometimes you may feel that in the platform. Sometimes you may feel it in last click. But sometimes Snapchat can drive a customer journey with the click data is only branded search or organic search, right? And so, what we've been working on and focused on is trying to bridge that gap.

12:39>> I actually don't know how to think about it. And so, I would ask you what is the recommended way of thinking about that? >> Yeah, yeah. So, so what I see in the industry a lot is like brands if they have reason to believe like the channel is driving good growth for them. That may be a trigger to go maybe a Northbeam MMM or an incrementality, whatever it may be. You look to go validate that action. But part of the reason I'm up here and saying it with Gabe is we've been working on this thing called clicks and deterministic views.

13:03>> What is like >> And what this clicks and deterministic views thing is is it looks to capture more of the incrementality that click data as a measurement lens Northbeam or otherwise has historically never been able to catch. So, we plat partnered with Snap for for every platform or for specific platforms? There's a couple specific platforms we're starting with. I don't have material yet, but you can start to hear from us 10 so it's coming here really soon. But factual omni channel impression data that's deducted tied to a specific order and we're really excited about because for brands who are stealing and Snapchat are secondary channels and you had mentioned, "Hey, like trying to go do an incrementality study or an MMM even with us, it can be like this huge >> Yeah.

13:49>> huge cumbersome thing. we're excited for this new world of hey, let's go validate these secondary channels with factual conservative views and we think it'll help brands like you guys. >> Well, it scares me cuz like we have to run six to eight offers, rotate the pages, >> Yes. >> rotate the platforms and then report on LTV across the board and it's like really you're like, I don't know. >> Yeah, I don't know. I don't know how to think about that.

14:12>> It It gets really tricky. So, we're trying to just have a simple system where it's conservative omni channel safe views where you're not going to get burned by them and then use those to optimize your campaign ad set and ads. Use it for planning purposes, right? You got mad in there, you got all your secondary channels in there. Toggle it on and see how much credit got pulled away from branded search, right? Did it all go to Snapchat? If so, that that's probably going to be an incremental channel. At least I've seen that when we have run the studies on the back end and Gary was mentioning this earlier, but he's big on like organic social and different tactics and brand tactics.

14:46That's what I'm most excited for it. >> Do you think Do you think >> You're doing an organic social post and you're boosting that, right? I'm curious what the incremental omni channel views are on that. >> I think that'll justify more people creating on organic platforms that even So, say if you were to make series on stop, technically it goes away. Where does Where does that live? Can you continue to see it overall? If you can justify how much lift in organic you're creating from the brand side on any of the social channels, it's going to incentivize us to invest there rather than the dollars cuz we already know it's getting way more expensive on the big boys.

15:19>> And that may be organic social channels. That may be like a different tactic, right? Where you maybe you aren't putting the the offer call to action right in there, right? So, I'm really excited for it, but I I'm curious how you kind of think about measurement, MTA, last click, incrementality, any of the CDV stuff that we've been working on. Where's your head space at, Gabe? >> Yeah, so I think like the big three, 3.5 measurement setups like platform, we're not going to talk about that, right?

15:47Like we all know what what the downfalls are there. Within like the MMM universe, I think that does a really good job of balancing like looking at that growth and maintaining efficiency, but like you're predicting and you're forecasting at the end of the day, so like that's your drawback there. we'll say there's some daily read MMM stuff coming out that I do think has is is promising that that we like a lot. on the incrementality side, like it's such a volume game that structuring those tests across eight offer like nightmare.

16:16We are practically not offering CLS anymore. Like we really recommend a third-party geo lift as a good solution for that. >> Yeah. >> And then MTA is, you know, kind of what I think the majority of this room is probably most interested in, most utilizes. Again, like we're we're going to to really rely on that. I do think the growth component can be challenging there. Like you're kind of got 6 oz of steak on the plate and everybody's fighting for it. It's like there's a whole kitchen back there, there's a whole audience back there. Like you need to maintain ability to still grow and not just fight over the conversions that you now are happening. And so we really applaud Northbeam for taking this innovation with the clicks and deterministic views to say, "Hey, we know that there's far more interactions and there's far more capabilities and there's far more conversions out there for you. We need to find a way to be able to accurately measure that."

17:07And I think it's probably the biggest innovation that we're seeing in terms of these three sectors of measurement to take the step forward and conquer the con that that is out there for that sector. >> So, keep an eye on Zenly and we'll get Nick loaded up with the first. So, if it works for him, he can disseminate it and spread the good word. But >> Does it will still disseminate the good word? >> Exactly. Exactly. Like, let people know.

17:30I'm really bullish at it. I'm here cuz like I've been a growth guy and all I do is look at look at dollars, right? So, I'm excited. And I think it'll help us scale secondary channels, right? So, that's kind of piece I want to get into a little bit more a bit is like why do we need to even try? Curious from your lens, Gabe, like should brands diversify? We're heading into Q4. Should they not diversify?

17:54What's your read? >> Yeah, and I think this stems a little bit to the organic thing as well, right? Like, each channel is going to be very different. Like, you've kind of had two components about organic. Like, I'm going to tell you don't put a lot of time into organic Snapchat. >> That's real. That's real, yeah. >> Whereas TikTok, right? Like, keep posting, keep posting, find something that's getting traffic, put some ads behind it, right? Like, that should be your strategy there.

18:19but when it comes to that secondary channel with Snap, like, I mean, we heard Gary V rather loudly, like, even outside yelling, like, stop doing the same thing over and expecting a different result. Like, you need to get on these channels. like, we didn't I didn't bring it up in the earlier audience overview of Snap, but like, we offer about 30% incrementality to the meta channels as well. So, from an audience perspective. So, like, go there, right? You need that for that growth. The harder part and just being real with y'all, we don't have 10 years of pixel click data, right? We don't have that. And we need to build out your best practices. We need to work with you. We know that those first couple months can be challenging. Also, huge rewards, like we talked about with Breeze, right? Coming on and being able to derive really meaningful impact for you at critical times, whether you're early, mid, or even more late-stage brand. And so, that deterministic view component allows you

19:14to better understand what is the total engagement with these ads that I'm running on the secondary platform versus just this hyper-focused last click, right? That is going to favor other channels. So it enables you and and the data that we've seen is that Snap is actually very competitive with the other channels when you look at the deterministic views, and it enables you to say, "Okay, we're still going to look to drive your your click attributed revenue as best as possible." That's like I said, 90% of our product resources are going towards that, but that gives you a little bit extra boost to know, "Hey, there's something here plus the incremental audience, plus not relying on Meta to be able to get out there and and get that secondary channel going for your Q4, etc." And to be honest with you, like we didn't approach going back to Snap with like this is going to be perfect.

20:07This is going to be one-to-one because again, that's not the role that we wanted to play in it. We couldn't scale. Like when the time was needed of Q4 of last year, I could not increase my budget high enough and Meta would not take the dollars. >> Yeah, this is this is the conversation of you've maxed Meta. I like I think a lot of people have. >> And so this is I'm I'm trying to spend the money. I'm trying to give it to any platform that's going to take it that it gets me close enough. And then it's like a measure it's a ma- it's a matter of like I'm just pushing up on M-cap right here.

20:36I got to probably lower it down. Where am I going to get volume? Cuz if I just kill everything, then it all goes away. So I know I have to spend on these auxiliary channels to make my core channels be a little bit better. I know I'm seeing >> That's it. That's that's 100% it. >> Yeah, we see it at time too, right? When we pull out all the our tech allows us to track touchpoints indefinitely. It is very interesting if we go, "Okay, all the versions we saw from our pillar channel, how actually how many actually started on a secondary channel that maybe was 45 days ago that the secondary channel would miss and last click data would miss. It happens more than you think.

21:10>> Yeah, I believe it. >> I'm I'm curious like a you guys have scaled a lot of secondary channels. Pitfalls, traps, things like hey, this isn't working. any guidance you'd give the audience here on like things to try and avoid if you're thinking about launching in Snapchat or otherwise? >> We we the amount of spin we have on auxiliary channels is like it makes up all of the all together it's more than our core channels. We The easiest thing is we took our winning offer on Meta. We just took it in immediately onto the auxiliary channels.

21:38which is in our case it's like buy buy some buy X get Y or buy X get more X at Y discount. That's usually the structure we do. And then we just would duplicate over exactly what it is the ads were. Then phase two is you eventually will The ads that you take or the content that you made on core, Meta, Google, YouTube, then we have to actually make platform specific content cuz the experience is very native to the platform on all of them.

22:03Like and some people can be like, "Well, my Reels content is similar to what I can run on on Snap." And the short answer is yes, start there cuz it's the fastest. But you need to make unique content specific to the platform. And we we did not do that fast enough. Same thing on you, same thing on Apple. I'm like, there's different use cases and like gimmicky nick wrong word but you have to be gimmicky to the platform to get the the and or reach. That's so true but it I would say stick with the offer that works. And then we realized that some of our ALVs can't be as high as it was on Meta as it is on Snap or or app or even Twitter to that matter. Twitter we can say the most aggressive things that's humanly possible and they're like, "Awesome, give me your ad dollars." Whatever. But other places they're a little more sensitive to the the niche in which we're selling in.

22:48>> I would say that the biggest thing if you're looking at Snap

The Importance of Signal and Measurement

22:53like one through three is signal signal signal. Right? Like I think a lot of times people take for granted the signal and measurement setup that you've had on Meta or whatever channel for 10 years, 5 years, like the history of your brand. Like, you need to come and make sure that you're delivering a really good event quality score. When we see somebody take their score from poor, which would be like, maybe they're not sending click ID properly, user email, user phone number, those parameters are so important, to like good, we see 42% increase in CPA. So, like, people take for granted, "Oh, I have run ads on Facebook for all these years, but I actually never set up the pixel."

23:35Right? Like, you you need to set it up, right? We're making it as easy as possible. We're integrated with with Shopify, etc. >> So, so that you're saying even if you're on spending, just get just get the pixel and [ __ ] set up. Yes. >> Okay. Okay. >> Yes. And then, I'd say 85% best practice, right? Of the ads, you're dead on. Like, you know, you can get by with your first test with with your Reels, TikTok creative, but then getting involved. And then, we know it's hard, so like, we are offering far more account management support. Like, anybody in this room, like, I will get you on the phone with a rep today to like talk about it. If you've got a struggling campaign, if you want to get started, there's four of us here. We got the booth, like >> up. That's pretty cool. Thank you for that. That's cool.

24:13>> Take him up on that. Get the pixel installed. >> Like, you you can talk to me directly, like, anybody in this room. >> we heard from I want to I want to just fire like one more to try and wrap us up a little bit. the levers that matter most, and is that going to change as we move forward? Like, I I'm on North Beam. I believe measurements going to have a pleat a place as we move forward, but like, the levers operators can pull to try and drive growth. Sounds like you're a big believer in signal being one, and just make sure the data like the signal being clean is actually a bigger lever than you than you'd maybe think.

24:47I know what Gary talked about organic social affair a bit, but in your eyes, what's that number one lever still? >> I I think it's offer. Like, Zach Stuck, a good friend of all of ours, like, he I wish he was here, but he and I go back and forth and Jordan Menard will talk about this tomorrow morning. The offer like if something's going wrong and like there's like levels of like influence you have in the ad account.

25:08Things aren't working. Stick with me on this. I'll land this plane, I promise. Things aren't working. Change the creative. Awesome. Things aren't working. Change who's in the creative. Awesome. Things aren't working. Still not working. Okay, now I got to change the structure of the the landing page, maybe the the where they're with the experience of what they're going to see. And then lastly, the biggest lever that takes most amount of time is change the offer. The offer of what they're going to see from what they land on. It's the biggest impact A from a conversion standpoint and B from a like a a scalability standpoint. We first hand saw this. We were struggling, struggling, struggling, found one offer that cracked.

25:42Friday, Saturday, Sunday, it's a million dollars in spend and all of a sudden like very different scenario. And so I think that biggest impact of things that you're like, "Dude, nothing is working right now." You've got to go offer first. >> Yeah, I'm a big believer there, too. Right? Like that can change the financials of it. You're competing in an auction, you get competed down, it gets the competition. All of a sudden a little bit different positioning there can help. yeah, yeah, great questions. Very much appreciated you guys having me up there.

Closing Thoughts and Future Directions

26:08I'd say my closing thoughts and my hopeful takeaways for for folks in this room. measurement state of the union is tricky. You can baseline off any different lens you want in platform, last click, north beam, otherwise, but it's about kind of figuring out what lens is going to help you unlock that that channel and allow you to diversify. Signals going to be a lever that can be pulled, right? Keep an eye out for us, clicks and deterministic views if you are scaling secondary channels.

26:34There's a bunch of stuff you can do, but frankly that's going to be an easy way if you're already hooked up with us or or considering. any closing thoughts you'd like to add? >> No game. >> I enjoyed it. No. Phenomenal. Nice guys. Ryan, Gabe, thank you guys very much.