← DTC Podcast
DTC Podcast
Ep 556: Fix These YouTube Targeting Mistakes to Stop Wasting Money Building Top of Funnel Awareness
00:23:57
Loading summary
Transcript48 lines
- [00:00]
A
People are not targeting who they think they're targeting in a large variety of scenarios, auditing different players in the space. I just see quite a few things potentially going wrong for advertisers looking to build up awareness and targeting via YouTube. There is a sneaky setting in demand gen campaigns specifically so you can build out the perfect audience you want to target and then in the ad group level settings there's a box to tick. I bet most advertisers listening to this podcast are going to have their eyes opened when they see their significant discrepancies in what they are trying to target or what their most frequent converter is coming from in terms of demographics and audiences and where they're actually targeting in their upper funnel campaigns.
- [00:50]
B
It's all killer, no filler. And I'm here with Dougie who leads the Google team over at Pilothouse. Lots to talk about today our fantasy hockey teams are in full flight. The Blue Jays are in the World Series and we've got to talk about YouTube ads, advertising, targeting. What, what went wrong in the world, Dougie?
- [01:07]
A
Yeah, well I think we've looked under the hood at a few of the fun demand capture side of things in search and where advertisers are potentially going wrong. And we've been doing a lot of awareness marketing through the end of Q3, early Q4 as we try and stock the pond for Black Friday, Cyber Monday and through. Yeah, just a fair bit of auding different players in the space. I just see quite a few things potentially going wrong for advertisers looking to build up awareness and targeting via YouTube.
- [01:40]
B
Which is the big theme of the year is really understanding how to generate true incrementally new demand capture. And for a long time Google has been touting YouTube as it's a really effective top of funnel generator just because so much of what it's doing on the search side is more about demand capture versus demand generation. So practically, yeah, I guess, I guess what today we're talking about is the common mistakes that you're seeing when people are trying to use YouTube as that top of funnel, how they're not targeting always who they think they're targeting. It reminds me back of I was, I was doing gaming back in the day and the amount of times you'd find that your ads were just being advertised to children or to, to people. And I think that's a big one on YouTube too, right? Is kids just watching YouTube and if you've got ads going to them like that's not great. What are you seeing when you open up the hood on some of these demand generation campaign.
- [02:32]
A
Yeah, so I just think people are not targeting who they think they're targeting in a large variety of scenarios. To your point on children, that's one area that we've built custom exclusions for at the agency to make sure that as best as possible we are not serving on gaming content. Gaming content is such a huge portion of YouTube content in general, but generally for the majority of DTC advertisers it's not particularly aligned content. And we've touched on it in the past. But I'll touch on it again. Google doesn't have a good way to distinguish when for instance a parent is using a device versus when their child is using that same device. So you might be targeting Joe from Montana and he's 48, but he has a couple of daughters and maybe they're watching Bluey Highlights or whatever it is on YouTube. And so Google is trying to target Joe, but in reality his daughters are using the device and you result in serving impressions to that user, that child, which is not particularly approval. So that's one area of misaligned content is due to that lack of differentiation in the person using the device. There's also a ton of additional exclusions we try and put in there for people that watch like passive content like 247 ASMR. Right. The stuff you fall asleep to, there's quite a few of them we built out. But misaligned content and having topic keyword and placement exclusions in place are very important to try and reduce as much as possible that low quality traffic. And you can also go through your placement reporting to determine how relevant your ads are to the content they're serving to.
- [04:06]
B
How do you do those exclusions? Like how do you exclude kids on their parents devices for instance?
- [04:11]
A
Yeah, so the best you can do is just go by placement. So either building a list of specific YouTube channels that you do not want to show up for, you can leverage AI as well to try and pull up the most frequently served impressions on that type of content. Placement exclusions are one keyword exclusions as well. They work in a different way than negative keywords do on search. But generally if you're trying to avoid targeting gaming content, you might have Call of Duty as a negative keyword. Right. So that's reducing your ability to serve on videos that have Call of Duty in the title and description, that sort of thing. So that's generally the methodology is to build lists of placements, topics and keywords that are not particularly relevant to the content you want to go after.
- [04:59]
B
Can you walk us through? You gave us one example, but what's another example of a way that this audience can go wrong? I have a note here about you think you're targeting premium female audience, but it ends up again being teenage gamers. Why would that happen?
- [05:10]
A
Yeah, so there is a sneaky setting in demand gen campaigns specifically so you can build out the perfect audience. You want to target rich females who are 44 to 54, what have you. And then in the ad group level settings there's a box to tick you might be you'll automatically be opted into optimized targeting. So you have to. You can be optimized, you can utilize optimized targeting, but there's also an additional checkbox underneath that allows Google to automatically override any of the demographic targeting you're doing in the Demand gen campaign. And this is automatically enabled on all campaigns or it's automatically not opted into on all campaigns. So I've looked at examples where again we've been trying to target females, maybe middle aged, and we've ended up surveying targeting towards males instead for a great portion of impressions. Now we catch it super early on within a couple hours. Right? But still I've looked at other through my audits, other clients in the space and they are doing the exact same thing and they've left it for months if not years in some particular cases as well. So that sneaky setting and demand gen is one to look out for as well. Misaligned demographics are another big one. So if you are serving for instance a luxury product with a significant AOV and you are then running traffic campaigns and you are trying to drive low cpv, low cost traffic to site, try and build up your pond. In the vast majority of cases that low cost traffic is going to be the lower 50% of household income and typically that's not going to be very aligned for luxury shoppers. Right. So one additional way you can look to ensure that your audiences are aligned to your demographics is you can actually upload a customer match list to Google or use your all converters audience in Google and you can navigate to audience manager and data insights and Google will actually provide you the breakdown of your demographic. It will break down male versus female, the device, the location. You'll have to go into audiences to sort out the different household income brackets that your conversions are coming from. But cross referencing that existing audience data with the data your top of funnel campaigns are serving to I bet most advertisers listening to this podcast are going to have their eyes opened when they see there are typically significant discrepancies in what they are trying to target or what their most frequent converter is coming from in terms of demographics and audiences and where they're actually targeting in their upper funnel campaigns.
- [07:53]
B
Little sneaky. That's. We always talk about platforms always being a little greedy when it comes to things like that, but it's like little sneaky Google. I don't know, I don't know if I like it. One of the conversations I've been having recently is just on forcing these platforms to find new true net new people. So there's a lot of like exclusions when it comes to to past customers and stuff like that. How is Google for keeping things pretty airtight or is it a leaky process when you're trying to only find new people?
- [08:22]
A
Yeah. So I am of the opinion that Google the automated Google setup in terms of differentiating new versus returning customers is not particularly accurate. So in the past we've talked about okay, implementing server side tracking and rather than at the audience level, I think a lot of people do this at the audience level. They'll exclude all converters or all visitors if you're trying to really truly reach upper funnel. And that's totally reasonable. But in our experience, and I've actually got some numbers from a recent test where we rather than excluding only at the audience level, we actually tried to optimize the campaigns at the conversion level for new customers. Now you can only do this in conversion focused campaigns. So demand gen campaigns that are focused on driving action but that action doesn't necessarily have to be a purchase. Right? It is in some scenarios. But what we found in changing the conversion goal tracking towards new customers compared to just targeting any old customer at the conversion goal level is that we were able to increase new customers by 50%. We were able to do that at a lower end CAC and we were able to increase the overall percentage of new customers coming through Google by 15%. All of this as reported by Shopify. So really strong signals that actually changing and we had these audience exclusions in place for this particular example beforehand. Right. But what we did change is that we told Google hey optimize only count conversion value and conversions coming from a new customer purchase. We do not care about returning customer purchases in these campaigns and that resulted in us being able to drive much more incremental performance by way of acquiring more net new customers.
- [10:08]
B
That's a great tip. I was in a chat and I hopefully doesn't mind Me saying this with David Herman and he was seeing, I think it was mainly for meta, it was maybe not for Google, but he was saying that he was getting a higher match rate on his exclusions by breaking his spreadsheets into like. Instead of saying this email has all of this data, he would, he would break it into like a single column where he would repeat that email multiple times and only send one piece of data with it with each email. And was saying, I think it was more on the meta side, but he was saying that he was seeing a much higher fidelity to his exclusions by doing it that way, which I thought was interesting. Is that something that would be relevant for Google at all?
- [10:48]
A
So when you upload a customer match list, you can tie it to a zip code, you can tie it to a phone number, an email address as well as a home address. So there would be potential applicability worth a test in uploading that customer match list with each of those individually as opposed to all under one. To your point, I'll have to listen to that particular podcast to see how he's doing it.
- [11:10]
B
It was actually just a private chat on a Slack chat, so I don't.
- [11:12]
A
Know if he wants me to. There you go. Maybe breaking news here.
- [11:15]
B
Tip.
- [11:16]
A
But yeah, applicable for Google.
- [11:17]
B
Talk to me a little bit about how to use Google audience insights and CML uploads to audit actual converters.
- [11:24]
A
Yeah, so it's really interesting. You asked me last time I was on the pod, what do you think is the most underleveraged or interesting tool that people don't use often enough in Google? And I, I came back and I figured, you know, I think this is actually it because I don't think it's just applicable to Google as well. I think we've at the agency used this for insights on meta or additional targeting on socials as an example. So by uploading your customer match list or by strictly using all converters and you have to have data to do this right. A fresh startup won't have an all converters list, they won't have a CML to upload. But by navigating to audience manager, going to your data insights, as I was previously mentioning, you can look at which particular audiences are highly indexed in terms of what they're actively shopping for. You can see what their affinities are as defined by Google. Then again, you can get some demographic information as well. So sometimes we use this to help influence. Okay, what kind of audiences do we want to layer in? Whether that be on YouTube targeting or whether that be on Meta or elsewhere, it's just a generally useful snapshot of data to help you understand your customer a little better. And yeah, we've also used that to battle test some of the strategy we've got going on to determine here's what a brand thinks their audience is like. But when we actually look at the data, there are some points of misalignment and we can then analyze and discuss those areas of discrepancy and determine, okay, is the direction we're trying to go in aligned to breaking into a new demographic or did we have a misunderstanding in terms of what our audience truly is and what they're interested in? So highly recommend using that function to check things out there.
- [13:13]
B
I just wanted to ask about the evolution of YouTube because I know, are we spending more on YouTube than we have in previous years? How is that assumption that YouTube is the top of funnel demand gen for the Google suites? How is that playing out?
- [13:26]
A
Yeah, so we are probably investing more in YouTube now than we ever have before. And there's two buckets of YouTube for me. There's YouTube trying to generate conversion generally through demand gen. And then you've got proper YouTube for awareness. And depending on the scale of your business, you might want to do A or B or A and B. But we are definitely investing there more heavily than we probably ever have in the past. Part of it is that I think the search landscape is changing. And so because the search landscape is changing, I think that we also have to think about how Google can play more throughout the funnel. We as an agency have kind of aligned on the statement that creative is paramount. Creative should be a reflection of your strategy and you can't really reflect your creative strategy in 30 character headlines in many cases. And you're not going to create new demand for your product via Google search and shopping for the most part. Right. So there are often cases where there's existing demand out there that we want to pull in and we think, hey, we can show this person the right message of the right product at the right time and convert them even if they haven't heard of our product before. But there are also cases where we are capping impression share. We do not think that we can do much more on the search side of things. And that's where typically looking at that awareness bucket for YouTube can be pretty fruitful for brands as well.
- [14:48]
B
And then because it's all on one platform, you can see a really clear through line from those because I know as performance marketers quite often over the past five years, the focus has been on driving to conversion campaigns kind of all the time. And I think, you know, awareness and brand development is something that we're thinking about more and more. So in that case, where you're going to do an awareness campaign, it's, it's a part of a bigger campaign where you're doing the demand capture with shopping and with search ads and display as well.
- [15:18]
A
Yeah, probably not so much display these days. I'm a bit of a display naysayer for the most part, but definitely YouTube. They've also, they being Google have provided more metrics in terms of being able to measure upper funnel YouTube work. So if you were using last click attribution and you were also running YouTube campaigns, you're probably not going to get a lot of last click performance attributed to those campaigns because you've also got brand campaigns set up, right? So those brand campaigns are, someone's going to see a YouTube ad, they'll maybe search up your brand, they'll click and convert. YouTube's getting none of that attribution. So platform comparable conversions and conversion value specifically for demand gen campaigns can be useful for more accurately aligning on the impact of YouTube and comparing that to the impact of other socials platforms. So it's attribution that's more aligned to meta and platform reporting. As an example, we talked about, you.
- [16:17]
B
Know, optimizing towards new customer purchase actions and you mentioned your main point there. But you have a couple other ones here around how much you're trusting in Google brand lift studies. What, what do you have to say about that?
- [16:29]
A
Depending on your level of investment, Google offers a couple custom ways to measure awareness impact. One is through brand lift studies. And so if you've been on YouTube before and you don't have premium, you've had to at times fill out a survey instead of being served an ad. And generally it's given you a list of four different brands and which ones have you heard of lately, that kind of thing. So that is how a brand lift study is measured. I'm of the opinion that it's just not a strong way to indicate lift. I think people will click a button, they'll fill it out, they won't be thoughtful about their answers. What I do think is a strong use of measurement is measuring brand search. And Google has brand search lift studies as well that I think are much more useful in terms of analyzing the impact of awareness spend. And we also of course build audiences based on our YouTube campaigns, add them as observations, see, okay, are these people coming back through other campaign types, the halo effect and the other form of measurement, of course even before we get to conversion is just making sure that the videos are being engaged with the traffic that does even click through those videos is engaged post click as well, spending time on site, browsing different post click experiences, that kind of thing. But yeah, not so trustful of brand lift generally as a potential measurement option.
- [17:48]
B
I've been targeted with enough of them and just the way I treat them, the amount of times I'm like, no, I have no idea what any of these brands are or like, yeah, obviously I've seen three ads for each of these brands in the last week or whatever.
- [17:59]
A
Right.
- [18:00]
B
What's your note here about narrower geotargeting as it relates to forcing Google to find new customers?
- [18:05]
A
Yeah, so I guess that one is more around the measurability of awareness on YouTube. So I think a lot of brands out there would for instance just serve to the entirety of the states trying to get their message out there to the world. I think it's much easier to measure your impact when you're launching awareness in focused pockets of GEOs. And that can be even at the municipal level. That doesn't have to be, okay, I'm going to serve to California and Texas. That can be, I'm serving to Sacramento. Right. You can be very specific and that allows you on the back end. Even without custom measurement solutions, you're able to measure your customer increase from that particular geo, whether you're using Shopify or GA4 as a measurement tool as well, that kind of thing. So, and especially for brands that have lower budget thresholds for awareness or they're just trying to have their first foray into awareness. I just see narrower geo targeting as being a better step in than launching across the entirety of the states. As an example, just before we wrap.
- [19:12]
B
Up, give us I know you've mentioned this before, but give anyone listening this the exact steps to take to go in and audit to make sure that their top of funnel YouTube campaigns are actually or to tell their agency to do this audit to make sure that they're not wasting their money on bad impressions. What are the steps to take there exactly?
- [19:30]
A
Yeah, so you want to go into your audience report, you want to open up your audience report specifically for your upper funnel campaigns. You don't want to look at everything. You want to look at specifically upper funnel YouTube upper funnel demand, gentlemen. You then want to ensure that the actual served demographics so not the targeted demographics but the audience report, you're looking at the demographics Specifically and then you are also then zooming out, looking at your entire portfolio over the past 90 days, whatever it is, and determining whether your holistic customer base over the last 90 days in that audience report is aligned to the allocation of spend in your upper funnel campaign. So I think a lot of brands will go in there, they'll look at their audience report for the entire scope of the last 90 days for all their campaigns, it'll probably be relatively aligned and then they'll look at their demographic breakdown for those specific upper funnel campaigns. And I think there will be cases where you're targeting almost entirely the wrong gender. I've seen that before and I think most commonly you're going to get a lot of higher AOV brands that are just inundated with the lower 50% of household income bracket. And generally I think those will be lower quality. And that's why generally these upper funnel campaigns, when they're targeting a low cpv, low cpm, what have you, they get such vast engagement from that lower 50% audience. But that that engaged audience is not going to be potential converters. And in fact it actually dilutes the quality of your retargeting audience as well.
- [21:05]
B
Yeah, it was interesting example you had here about a women's undergarment brand serving mostly men. And is that just because men will click on that when they see that?
- [21:13]
A
Exactly. That's my assumption as well is that there was an optimization towards soft metrics by Google. Google saw hey these people are more engaged so I can get a cheaper cost per view but in reality that's not the ideal target demo. It's misaligned to 99% of converters that are female and it again just that sneaky demand gen opt out was not double checked. So make sure advertisers out there are double checking their optimized targeting settings for dimension.
- [21:43]
B
So if you're listening to this right now, talk to your agency or do it yourself and then email Eric and or Douglas at Pilothouse Co and let us know if you do this because I want to know if we're saving you some money. And then finally, what are your predictions for the World Series here, Douglas?
- [21:58]
A
Oh, I'm.
- [21:59]
B
Are you a Mariners fan? Are you one of those like west coast Canadian Mariners fans?
- [22:02]
A
Good, okay, thank you. No, no, no, I won't get my myself canceled here in Canada. No, I'm, I'm a medium Jays fan throughout the year, so I've been following them the whole year. I didn't expect them to be this good. I think most Jays fans didn't expect them to be.
- [22:16]
B
They were the worst team in the east last year.
- [22:17]
A
Yeah, it was it's been unexpected run very fun. I think that in some ways it feels like the Jays team has a bit of special sauce. They might fell the giant there in the Dodgers, but so I guess my My heart would tell me Jays My head would tell me Dodgers. My my hope is that it is at the very least a competitive series and one that goes six or seven.
- [22:40]
B
Well, I'm saying Blue Jays in seven. You're saying Dodgers in six or seven. What are you saying so we can revisit this?
- [22:46]
A
My head will tell me Dodgers in six. My My heart tells me Blue Jays in seven.
- [22:50]
B
My head and hard are both saying Blue Jays. So I'm I'm all in on this. Ohtani is an absolute unbelievable so is Wembayani. I'm just watching some clips of a Victor Wembayana. We're in this age of like absolute sport people just remaking the sports in all these ways, which is again, not what this podcast is about. But we could start a spin off soon, I'm sure.
- [23:10]
A
Well, I think we'd be remiss. Not to mention our our J's here. It's our.
- [23:14]
B
That's true.
- [23:14]
A
First World Series in 30 years. 30 some odd years. Right. So we've got to be able to be loud and proud about it for for at least the next week or so here. Nice.
- [23:22]
B
All right, well, thanks for connecting again, Dougie. Talk to you later.
- [23:25]
A
Thanks, Eric.
- [23:32]
B
Thanks for listening to today's episode. If you're not getting the D2C newsletter, you can subscribe for free at directtoconsumer. Co. And if you want to learn more about Pilothouse's all killer no filler services, take off to Pilothouse Co. I'm Eric Dick and this has been the DTC podcast. We'll see you next time.