Ep 603: Why Most DTC Brands Fail on YouTube (And How to Fix It in 60 Days)

0:00What are the best brands we're working with on YouTube doing right now?

0:03Dougie:They're approaching YouTube in a slightly different manner. People increasingly need to view it as more of a awareness and consideration platform than pure conversion. What are other go-to guardrails that you have around setting up YouTube campaigns? If you don't get your brand across in that first five seconds, there's going to be a lot of wasted impression. If you don't understand who you're trying to reach, if you give the algorithm too many signals, the wrong signals, too wide a breadth of available targeting, it's going to have a much harder time building meaningful awareness for your brand.

0:45It's all killer, no filler. I'm Eric and I'm here with Dougie who leads the Google team at Pilot House. Today, we are talking about YouTube, a topic we don't talk about a ton on this podcast, but one that's always interesting to me. What's going on with YouTube? What are the best brands we're working with on YouTube doing right now?

1:02Dougie:Yeah, I think they're approaching YouTube in a slightly different manner than a lot of DTC e-commerce brands do. I think increasingly, YouTube has kind of taken the role of what has traditionally been TV marketing, right? I think its use case is significant from a quality reach perspective. And I think people increasingly need to view it as more of a awareness and consideration platform than pure conversion. Super smart. And why do you think, I think the story that I hear a lot is brands will test it to a small extent and say, oh, this doesn't work and then give up. You think that happens because they're looking at it as this pure performance play?

1:45Dougie:I think that's often the case. And absolutely, you should expect some level of performance from a retargeting perspective. When you're targeting warm traffic, I intend you should expect a return. But if you are truly reaching people for the very first time on YouTube, your first impression is via YouTube. Like you can't expect that to directly result in a purchase from that initial touch point. So I think it's a colder audience targeting platform as compared to maybe like Instagram or Facebook. And I think that because Google and its nature is grounded in search and shopping, which are high return channels, I think Google advertisers, but also advertisers outside of just the Google platform have that natural equation of, okay, Google is just for direct high intent return.

2:37Dougie:And therefore, when they don't see that direct high intent return from YouTube, they kind of pass on the opportunity after initial testing, despite potential indicators that might show it has an effect upper funnel. You mentioned earlier that more and more advertisers are treating it like TV. And I think that's really apt because I was just looking up 150 Americans watch YouTube on their TV every month. I actually don't watch YouTube on my TV very much. I'm one of the few North Americans who doesn't, I guess. But that's almost 50 % of the US population watching YouTube on their TV. So it's like, and you don't think of TV necessarily as always direct response by now, although Amazon is doing a good job with all the by now prime setups.

3:14But yeah, I think that's a good comparison.

3:16Dougie:Yeah, and your point on TV screens, like we're seeing it in how placements are playing out in the device breakdown as well. Like people come to or utilize YouTube on TV for long form content and people use it for shorts as well, but primarily long form content. And so the other added benefit of TV is it's harder to skip a YouTube ad if we're using skippable in-stream. There's non-skippable, of course, but hard to skip a YouTube ad on a TV that is on a mobile app or on your computer as well. So you see slightly inflated view rates from that on TV. But generally, there is high quality traffic coming through YouTube consumption via TVs in the United States.

4:01Dougie:All right, let's back up a minute here. And then what would you say has to be true for a brand before they should consider YouTube as an investment channel? I think two things. One, you have to be confident in the creative that you're utilizing for the platform, right? Like you can find the perfect person, but if you don't reach that perfect person with the perfect message, it's not going to pan out, right? So creative is one component. so having really dialed in creative strategy and therefore high performance creative assets maybe grounded in testing done even outside of the platform right is essential and then the other component so who are we serving to right what's the ad the creative component who are we serving to if you don't understand who you're trying to reach if you give the algorithm too many signals the wrong signals too wide a breadth of available targeting it's going to have a much harder time I'm building meaningful awareness for your brand.

4:54Dougie:So I think knowing your customer inside out and also making sure that you have a message creatively that resonates with that customer. You mentioned it as a top of funnel or it has a lot of value top of funnel and especially probably within the Google suite because as we've mentioned, Google is a intent capture platform. Google ads traditionally are more of an intent capture platform than a demand gen platform. So are there metrics you look for or is there like a threshold of Google ad spend that really makes a lot of sense to make sure that you have, you know, the top of funnel YouTube aspect?

5:27Or is it just more about the brand being ready themselves?

5:30Dougie:Yeah, I think the brand being ready is like you could have no ad spend on Google all time and still find success on YouTube. If you have the budget to invest and you know what you're looking for and what you're trying to do, like a startup might think, okay, my, my consumer base is more often consuming content on YouTube than they are consuming content on Instagram. Therefore, it is a more effective channel to deploy brand awareness budget against because of my understanding of that consumer. And in terms of measurement, we don't look at true conversion or not exclusively true conversion attributed to the platform.

6:07Dougie:In fact, my view is that the attribution Google has for YouTube is generally quite poor as compared to like a mountain or a CTV provider. So I think you miss out Just looking at the conversion view, as already mentioned, what we are looking at primarily, which I may have mentioned before, is cost per brand search increase. And they're actually starting to roll this out as a metric you can track in platform. You don't have to launch a brand search study. You don't have to go bug your reps. They're slowly rolling out the ability to track the impact of your awareness campaigns via demand gen campaigns or true YouTube video reach campaigns in platform themselves, which is really interesting to me.

6:48Dougie:That's great. Wait, you actually haven't mentioned, I don't think that on the podcast before. So I think that's a cool update in the metrics that matter for YouTube. The attribution piece doesn't make a lot of sense to me. It's just like Google's whole thing is attribution and it's all their sandbox. You have them in. Why do you think they're getting attribution wrong? Yeah, I think they have a hard time translating cross-device conversion. Personally, this is speculation, but yeah. Speculation. I think that the mountains of the world, they do attribution differently, right? Like they are plugged into looking at, okay, we served ads at XYZ times and this was the XYZ lift in purchases and session data for your store.

7:30Dougie:Whereas Google is more trying to draw a thread just purely based on kind of the Google sign-in, right? They are looking at someone who viewed and then they can track cross-platform or cross-device, rather, if they're logged in on their TV and they're logged in on their mobile phone. But they have a real difficulty, I think, tracking that cross-device attribution piece. And I think that they just don't have the same measurement format as some of the CTV providers. And I think that they recognize that as a gap. And it is something that they're increasingly trying to focus on as evidenced by the rollout of brand search as a metric that they're looking to, to kind of help bridge the gap between awareness and meaningful performance impact.

8:15I think cost per brand search is a great show. Are you looking for anything else in terms of you're looking, obviously, the direct traffic that's driven there? Do you ever see increases in conversion rates from people that have maybe seen an ad on the good screen?

8:27Dougie:Yeah, absolutely. So we're like other avenues for tracking YouTube performance are building observation audiences within your campaign structure to measure, okay, are we seeing increases in this audience pool in our brand campaigns, in our even our non-brand campaigns, in our retargeting, how are they performing? So that's one component. I think additionally, like when it comes to things outside of brand search and in platform, we do also look at true kind of net sales, what's happening in the areas that we've pushed. And primarily what you're looking at there is just a true geo test, right? Right.

9:05Dougie:So whether that be launching in certain kind of series of zip codes in the states that you believe your customer is aligned to, or whether it be a true kind of state versus state test, you have to take into account that if you are true, again, first touch discoverability on YouTube, the tail to conversion on that's not going to be immediate, depending on your niche, of course. But measuring that over time and deploying budget towards a specific geo, it's not limited to YouTube as a function to measure performance at any level of awareness. But that is another kind of methodology to try and reinforce and build a case for performance from YouTube.

9:45We may have talked about this in the past, but I can hear the music that plays with them right now. But the brand recall ads that I see on YouTube quite often, like which of these brands have you seen an ad for or whatever? Or that's part of their sort of attribution stack or how they quantify brand resonance?

10:01Dougie:Yeah, so it's called a brand lift study. I personally will disagree with most Google reps you meet out there because I don't believe it to be a very strong indicator of additional consideration. This might be personally biased because I find myself never filling them out. You'd think the Google guy would maybe pay for YouTube premium, but I like being able to be exposed to all these advertisements and what have you, right? to see what's going on in the space. But yeah, my opinion on brand list studies, I'm generally out on them. I would far prefer centering around a true data point like brand search volume where you know, okay, someone was served an ad and they took the intentional action.

10:39Dougie:It wasn't forced upon them. They didn't pick whatever was closest to their cursor at the time. They went out and they went, okay, I'm looking for this brand after being exposed to a discoverability format facilitated by YouTube. Let's talk a little bit about campaign setup here. So you've got, you maybe have your search campaigns running, you've got your marketplace or what am I, the shopping, Google shopping campaigns running. You want to jump into YouTube. What's your rule of thumb for how to set up the YouTube campaigns within Google Ads? Yeah, I think you can honestly get away with a relatively basic campaign structure, right?

11:15Dougie:The campaign structure is not the hard part of getting YouTube to work. I think it is setting the right guardrails in place within the campaign. And again, that comes back to knowing your consumer. It also comes down to just being familiar with the platform. Many, many times I see that we're not putting enough guardrails in place and the customer we're trying to hit is not the customer that we're actually hitting in YouTube. So YouTube does a pretty good job of over-indexing into the lower 50 % household income audience. I see that very frequently. And I would argue that the vast majority of DTC e-commerce brands are non-essential goods.

11:54Dougie:They might be great goods, but they're non-essential goods. And those non-essential goods tend to perform much stronger in kind of the upper household income demographics. demographics. So I think you can get away, going back to your true question around campaign structure, you can get away with a basic retargeting campaign, reach campaign, maybe some middle funnel if you were kind of focused on bridging the gap between a viewer and a site visitor. But I've seen success even running a strict two campaign setup. Guardrails, you mentioned those. That's a really good one that you mentioned there about household income.

12:28What are other go-to guardrails that you have around setting up YouTube campaigns?

12:32Dougie:Yeah, so I personally don't love expansion and optimized targeting. I just think it tends to over-index into that category. And I think people set up these audiences and maybe they're very good audiences. They're very strict, they're custom intent, they're using a lookalike from high quality first-party data, what have you. But they've opted into expanded and optimized targeting And that eats up the vast majority of their budget. And you don't get really much visibility at all into who that person is outside of demographics. So that's a flaw. Again, household income is a flaw, I think, when it comes to guardrails as well, ensuring that you have true customer exclusion lists properly implemented.

13:12Dougie:I see too many times people have set up a customer exclusion list and they forgot about it. So it hasn't been updated in at times years. or they're using the wrong list, especially if you're on Shopify, the Shopify audience integration allows you a really easy way to exclude 90 % of your existing customer traffic. So I see that as a primary gap. I also see that having application in performance max campaigns that are servicing on non-search-based formats as well, because Google tells us, hey, give us a retargeting signal so I can help find the right person and maybe a prospecting signal to kind of accompany that.

13:52Dougie:So people input, okay, here's my customers. And then they give that as a signal, but they don't have that as an exclusion. And they think, okay, my PMAX campaign or my demand gen campaign is performing really well. Look at this high return, but it's highly indexing into the existing customer segment. I see that so often the case with CPG brands. Yeah, I think everyone is looking to reach new customers, especially top of funnel. So is it like in Google these days, Is there a setting where you're essentially saying like, don't reach people that are already in my CRM? Yeah, so you essentially just import a CML and can exclude it.

14:28Dougie:You can also have that imported CML, whether it be manually or through Shopify audience integration or Klaviyo or what have you. Like there's a million and one ways to do it. But you can also at the account level define here's my existing customer segment and then provide the signal to only target net new. But generally, my recommendation is just to, at the campaign and audience level, just exclude existing customers. Again, through whatever implementation you so choose, the Shopify one is easiest for many. And then when it comes to audience targeting, what are your insights there? I guess it's going to depend campaign to campaign, but their interest targeting.

15:05I know they've got a lot of different kind of targeting options. What do you recommend?

15:08Dougie:Yeah, so a few things. We're trying to start with the nucleus of highest intent, leveraging custom intent audiences. So people who have looked at competitors, people who have converted or not converted, your highest converting non-branded terms are ones to build audiences around, right? As well, you can input for demand gen specifically a CML with your highest confidence buyers. So maybe not anyone all time has purchased, but rather let's pick out, okay, just the women who have purchased and they've purchased two plus times. That's a very, I guess, precise signal to provide the algorithm to chase after.

15:45Dougie:So when you tell Google, hey, find the 2.5 % users on YouTube that are closest to this list, it has a much more precise signal to work off of as opposed to the broader signal of multiple genders or multiple personas or multiple intents. People might buy products for many different reasons, right? So trying to provide a high quality signal of this is my core segment is pretty important. And then just talk to me about maybe expectations when you've got a client, for instance, that wants to jump into the channel. How do you set expectations around, you know, ROAS or just general return that they could be thinking about?

16:22And maybe the length of engagement, like how many months or weeks does it take to kind of get up and running to a place where you can feel optimized?

16:29Dougie:Yeah. So if we're talking ROAS on YouTube, we're probably already, I think, in the wrong discussion. Yeah. With the caveat being retargeting YouTube very much has a place where you can see high returns. And there's maybe an argument to be made for lower cost goods, having that lower price sensitivity barrier to entry. But generally speaking, if we are assessing YouTube campaigns purely on ROAS, I think it is not a worthwhile investment for that business at that time with those intentions. There's a reason why the larger brands are not viewing YouTube through that lens. they like coca-cola is advertising on youtube because it's mass reach right they're trying to reach as many people as possible it is again akin to tv they're not measuring the the bottles sold from someone who clicked on uh that particular youtube campaign nobody's clicking on uh on youtube ads regardless right so i think that's a flaw if we're already focusing on the return alone and i i've already forgotten your your first question but i've had that discussion so many times uh that I'm sure.

17:35Dougie:I want to nip it in the bud. And then what do you, okay, so that's nipped in the bud, ROAS, okay, that's not the platform I'm looking at. So what does success look like then for a brand that sticks with it? In terms of timeframe, I think the four-week mark is a really important milestone to kind of assess what's been done to that point. Typically, you should be able to measure some degree of brand search lift from the efforts by that point in time. If not, I would make the assumption that your efforts have not been impactful to the business. For the smaller advertisers out there, you can simply go to something like Google Search Insights to indicate, OK, are we building awareness for my product in this space?

18:18Dougie:But that four-week mark is important. You can also back out through brand search your expected return, right? So if your brand search converts at 10 % rate and you know that you drove 100 additional incremental brand searches, okay, there's 10 sales right there. Okay, what's my AOV? Apply it to those 10 sales. Okay, this is roughly my return associated with this effort. If we're really trying to do the work of kind of backing out, this is resulting in sales, which we should. But I think just the window on that tends to be longer than people expect. And so you have to be ready for a 60 or 90 day engagement when it comes to YouTube.

19:01Dougie:And even more so if your budgets are small. I think if your budgets are significant, again, you're going to reach that point of statistical significance much more quickly than if you're advertising with a 3k a month budget on YouTube. And in fact, I think that 3k a month on YouTube without very strict targeting, and I mean very, very strict targeting, you're not really going to be able to measure any kind of impact from that kind of investment. That's the minimum. What is the minimum then? If it's not 3K, you should be doing at least 5, 10K? Yeah, my general rule of thumb is we should be able to measure a 10K a month impact on YouTube over a three-month span.

19:40Dougie:That is not necessarily grounded in a specific kind of benchmark metric. That is, to some degree, a gut check on experience in the space. What I would say for users who have a more limited budget but know that their potential consumer is present on YouTube is to leverage contextual targeting. So target channels that you know are highly relevant to your niche, and you can do that at a much smaller scale. And you can, in theory, see a bit more measurability out of that than you can kind of going mass market reach or more audience-focused metrics or methods of targeting, rather. What's the first thing you'd fix if you inherit a struggling YouTube account?

20:22Dougie:Most of the time, I'm looking at who we're hitting first. And oftentimes, I'm cutting out expansion-optimized targeting. I'm cutting out low performing demographics as well that aren't aligned. I'm auditing the placements as well pretty closely. So I think that and maybe I've talked about this in the past, but there's a lot of device sharing that goes on in family households. Right. So you can see a lot of campaigns that are targeting, in theory, a middle aged affluent man. But in reality, they're serving on like bluey content, right? Because the child in that household is consuming under that same profile.

21:01Dougie:So auditing placements is important. If we're running any kind of traffic campaigns, I'm kind of, if we're setting a low CPM, low CPV campaign and telling Google, get me as many people as possible. My position is there's a reason that traffic is so cheap and it's because it's low quality and no one else is trying to acquire it. So auditing placements is part of it, removing low performers. I don't mind. In fact, I encourage using demand gens with max conversion bid strategies, just with the assumption that, hey, we're not actually going to get a ton of converting traffic out of this, but using the algorithm to help find people who are most likely to become that converter over time can be valuable.

21:43Dougie:So a few places that I look at first to try and trim things back. I think of myself, and I'll admit, I quite often will try to fall asleep with a YouTube video on just let the ads play. If I'm going to do that, I should probably get a no-ad account. But do you ever do day parting for people that are doing for that use case? We've played around with it to minimal success, I would say, because we've done some day parting for parental audiences, given the device overlap concern. I haven't done a ton of it for the specific purposes of trying to get people who are falling asleep. Although, I will say, ASMR content, passive content, fall asleep content is very common to see in the low-quality placement report, right?