Artwork

เนื้อหาจัดทำโดย Blake Beus & Greg Marshall, Blake Beus, and Greg Marshall เนื้อหาพอดแคสต์ทั้งหมด รวมถึงตอน กราฟิก และคำอธิบายพอดแคสต์ได้รับการอัปโหลดและจัดเตรียมโดย Blake Beus & Greg Marshall, Blake Beus, and Greg Marshall หรือพันธมิตรแพลตฟอร์มพอดแคสต์โดยตรง หากคุณเชื่อว่ามีบุคคลอื่นใช้งานที่มีลิขสิทธิ์ของคุณโดยไม่ได้รับอนุญาต คุณสามารถปฏิบัติตามขั้นตอนที่อธิบายไว้ที่นี่ https://th.player.fm/legal
Player FM - แอป Podcast
ออฟไลน์ด้วยแอป Player FM !

Why campaign objective matters (and when it doesn’t) EP-028

44:32
 
แบ่งปัน
 

ซีรีส์ที่ถูกเก็บถาวร ("ฟีดที่ไม่ได้ใช้งาน" status)

When? This feed was archived on October 28, 2023 04:09 (5M ago). Last successful fetch was on February 18, 2023 00:39 (1y ago)

Why? ฟีดที่ไม่ได้ใช้งาน status. เซิร์ฟเวอร์ของเราไม่สามารถดึงฟีดพอดคาสท์ที่ใช้งานได้สักระยะหนึ่ง

What now? You might be able to find a more up-to-date version using the search function. This series will no longer be checked for updates. If you believe this to be in error, please check if the publisher's feed link below is valid and contact support to request the feed be restored or if you have any other concerns about this.

Manage episode 329026442 series 2987873
เนื้อหาจัดทำโดย Blake Beus & Greg Marshall, Blake Beus, and Greg Marshall เนื้อหาพอดแคสต์ทั้งหมด รวมถึงตอน กราฟิก และคำอธิบายพอดแคสต์ได้รับการอัปโหลดและจัดเตรียมโดย Blake Beus & Greg Marshall, Blake Beus, and Greg Marshall หรือพันธมิตรแพลตฟอร์มพอดแคสต์โดยตรง หากคุณเชื่อว่ามีบุคคลอื่นใช้งานที่มีลิขสิทธิ์ของคุณโดยไม่ได้รับอนุญาต คุณสามารถปฏิบัติตามขั้นตอนที่อธิบายไว้ที่นี่ https://th.player.fm/legal

Listen on:

Transcript

Blake Beus 0:00 Okay, and then we'll just make sure we look up there. I did pretty good at looking up there during the I only good. Yeah. All right.

Greg Marshall 0:08 Well, we are live

Blake Beus 0:10 live ish. You're gonna see a recorded recording. But you you have some questions for ya. So cousin was like, let's do it. Let's do this.

Greg Marshall 0:20 We've had this conversation before off recording I think it'd be valuable now to do a recording but also to kind of go through strategy live since Yeah, yeah. So So yeah, so one of the things that I have seen the I want to challenge this is, so there's different bidding strategies, right? Yeah, cost per click CPA Max conversion all in each one of these ad platforms. And that's assuming that the algorithm is going to match up the result that you want. Right, right. Here's my question. So I ran a test the other day, and you know, me, I'm always testing, essentially my job as marketers. Yes. Yeah. 24/7. I ran a test the other day. And I bid for video views, but this is on Facebook, okay. And I measured it against a traffic campaign, where one objective you say, I'm looking for video views. Okay. The other one, you say I'm looking for traffic

Blake Beus 1:25 traffic. And to clarify traffic is pageviews. Right? Like someone clicks on a link? The page loads? That's the event that we're optimizing for? Correct?

Greg Marshall 1:34 And what I actually found was the video views campaign. were slightly better. Oh, really? Then the traffic, okay. And the here's the kicker, same ad copy, same audiences. So interesting as it was, so nothing changed the other

Blake Beus 1:51 variables with the same worked better in what way? Like what was your end goal,

Greg Marshall 1:55 the end goal was to see like the test was, which one would be better to run video views or traffic, vert to get to the page, right? One of them you're looking for a higher watching rate. Right? Because if technically, if you bid for video views, you're getting higher washers versus traffic. I, in my base off this test, I saw deeper washy rates and better click through rates to the website versus the traffic.

Blake Beus 2:28 And so when you say better your your definition of better is more more people going to the landing page. You're not like adding to cart purchasing. Correct? Correct. Because this is a blog post. And then it also says two blog posts. Yeah. So you set so your objective was to get people to hit that blog post? Yes. But then you told Facebook, on one campaign, I want the objective of video views. But that video has a link in it to go to the blog page. And then the second campaign was you told Facebook, I want people to go to the blog page. Right. So and then you measured which one actually got more people to the blog page. Correct. And the video views got more people to the blog page.

Greg Marshall 3:10 Yes. And so I don't know if that was just and I believe both ads were a video ad. This literally the same little Okay, identical, like no difference and no change in work, because there were existing posts. Okay, she's existing posts, okay. And so it made me think because then I did a dive into, well, what if you go to YouTube, YouTube ads, that's you can do video, essentially, video views or web, right? Traffic leads wherever. And I looked at a couple of clients that are running, and we were starting with video views. And their cost per clicks. Were basically the same as if you were bidding for like web traffic. Oh, really, or CPA, but the CPMs cost per 1000 impressions. Were lower on the video view of video in YouTube and YouTube. Okay. This is where I was really trying to think, Okay, the next test I'm running currently to see if there really is any difference is targeting the same audience, changing the bidding strategies and see where the ads show up. Okay? And the reason why I'm testing where the ads show up is my what is it? The hypothesis, okay, is if your if your targeting is what you want it to be, and you change bidding styles. Is it really different to bid on one versus the other algorithm wise? Okay, that's the question, okay. Meaning, if I'm going after the same audience, I keep getting the same cost per clicks or maybe not cheaper. Okay, on video view versus web tracking or CPA. Then what is the difference of between one versus the other. Okay, and why would you read The why would it matter? Yeah. If you're bidding one versus the other? Yeah.

Blake Beus 5:03 Okay. That's the hypothesis. Okay, here we go. So I, I have some thoughts. So a lot of people talk about algorithms, and we talk about algorithms a lot on here as well. And many of the people that talk about algorithms, I have never written an algorithm before. Which is fine. Like, that's totally fine. But if you've, if you've written some code and sat down and thought, Okay, I need to write, I need to bang out some code that's going to take a few things into account, and then provide a result. How do I do that? You've had to approach the problem a little bit different. And you have, you have to have a better understanding of how algorithms can work. Now, I've never written an algorithm algorithm as complicated. As well, Facebook and Google ad platforms have I mean, they're literally taking 1000s, maybe even 10s of 1000s of data points, and running it through their machine learning. To to to pop out, okay, here's the people that need to see this at whatever, right. But I have written algorithms before, for some, some basic use cases, which I feel like gives me some unique perspective on how algorithms work. Because a lot of people, they're just like, well, I'm testing this and this and this. And then they say, this is how the algorithm works. And they're like, Well, maybe, yeah, maybe. But we've got to think about the decisions that the coders were making and how, how they can detect data, right? So the very first thing and we say this over and over and over again, is you've got to make it easy for the algorithm to to make a decision. Yep. Okay. Because, again, there's there's 10s of 1000s of data points, is one of the reasons why if you have so you have a purchase conversion and fires at the end of a big long funnel, right? You're you're doing the Russell Brunson Click Funnels thing where, where they buy this thing, then there's a bump sale, and then there's an upsell offer, and another upsell offer. And then like, if they turn down this one, there's like a, another offer that's maybe a discount of one of the first offer has and, and all of this thing, and then they go through six or seven steps, and then they purchase. And that's where the event fires. That's very difficult for any algorithm to determine what the hell happened here. Yeah, like like, because it's so far removed from the ad, click, or from a measurable ad event, that there's so many places for the ball to get dropped. Right.

Blake Beus 7:39 And so one of the things, I think the very clearly the easiest thing right now to say is, it's so much easier for Facebook, Google, whatever platform to measure events that happen inside their system, as opposed to something that happens off their site. So like a page load that happens off YouTube, you go to a different platform, even if you're on your phone, and you're inside the YouTube app, you're going to a separate webpage. But if you're measuring a result that happens inside their platform, it you they have access full access to that data, even with iOS 14. Yeah, right? Like, iOS 14 is not going to block Facebook's ability to know how much of a video you watched inside of Facebook's app. Because when you install the app, you give that app permission to track these things. And iOS 14 can block them doesn't block that right. And so whenever you have like a video view objective, it's so much easier for the algorithm to determine objective has been met. Yeah. Which is why in my opinion, you're gonna see always see lower CPMs got it. Now, we've talked about this before. And if you're thinking about this, you the listener CPMs, are essentially how you're charged. Yeah, that that is is an A very important metric. And the the lower the CPM, it's a stands for cost per 1000 views or impressions, right? The lower that CPM mean means you get more people to look at your ads for a lower cost. Yep. Therefore, you have more of an opportunity for clicks, more of an opportunity to determine in you know, that that person's intention, all of those things. So I think the difference in CPMs is probably why you're getting cheaper cost per clicks, and why you're getting more pageviews because more people are actually seeing that. And I think that's probably most relevant inside Facebook because they were really destroyed with iOS. And now iOS 15. Google was like, and we've been through this before we got this not a problem, but Facebook was absolutely right. They'll completely blindsided right So I think that's probably why you're seeing the best results, and probably why you'll very likely continue to see better results.

Greg Marshall 10:07 Well, in that, here's, here's my question then. And maybe I have always looked at it incorrectly, okay? Do you get a less lesser quality audience bidding on different strategies? So, for example, let's say I'm going after the same audience. Do you believe that the algorithms would show like by, say, video views versus purchase? Does that just completely show it to a different audience, almost like my mind in the past would think, well, if I do video views, it's essentially never going to show it to a buyer. Right? And if it is only going to show people who bought Yeah,

Blake Beus 10:51 so this is how this is how I look at when you set up targeting in any any ad platform, it will give you an estimated audience size, right? Yeah. So let's say you you set up targeting, targeting several different interests. And Facebook says you have an audience size of a million people based on your parameters, right. And then you run two ads to the exact same thing with different objectives. One is a purchase objective, and one is saying, video views objective? Yeah, your budget is probably most people's budget probably isn't big enough to blow through a million people have that audience in 24 hours. Now. It's just like, that's not gonna happen. So So part of what the algorithm does is decides, Okay, we've identified this audience based on the parameters you've set, but which people in that audience do we actually show it to? They don't just randomly pick people they're trying to optimize, like, they really do want to get you good results, because they want you to keep spending money with them. Right. And so I would guess that it's quite possible that they will show purchase objective campaigns to a different subset of that audience than a video view objective. And the reason is, is because one of the data points in there is is that they're using as an algorithm is, how likely is this person to to make a purchase? How likely is this person to watch a video, and we all have slightly different behaviors on Facebook, and some people are more likely to watch 1015 20 seconds, five minutes of a video, and some people are more likely to click on a button and add something to the cart and make a purchase. And so Facebook's going to say well, this person feels like they're ready to make a purchase. So let's put it in front of them. And, and they're also like, well, this person is the kind of person that tends to like to watch videos, so let's put it in front of them. But I think where I think there's a disconnect there. Yeah. Because I think it's much harder for an algorithm to say this person seems like they're likely to make a purchase. Yeah, I don't think that's an easy call for an algorithm.

Greg Marshall 13:01 Here's my argument. Yeah. Okay. Are my agreeance on what you're saying, but an argument to the general marketing population? Wouldn't you say that if there's less data, because of the privacy issues being fed back in? What did that mean that the algorithm has lost a significant amount of those n data points? Yeah.

Blake Beus 13:24 Right. Absolutely. Especially on Facebook.

Greg Marshall 13:27 So with that being said, Wouldn't it be you have you should change your thought process with the changes that have happened? Yeah, absolutely. So that's kind of my thing is I'm curious to actually, that just kind of sparked my curiosity, because I was like, Well, if you notice, like, if the algorithm or the datasets or whatever, are just not as intelligent and what is it 80% has been taken away? Don't they say? Like automatically 80% of that out? Yeah, maybe even more? Yeah,

Blake Beus 13:59 because iOS by default is opting out of ad tracking. So so it like, what is it? It's like, on my website, it's almost 90% of all traffic is mobile traffic these days? And the difference between iPhone and Android? It's probably 60%. iPhone 40%. Android, depending on industry, some industries are, are more, right, like Realtors way higher up on iPhone than Android. But developers may be higher on Android. But so it kind of kind of depends. Yep. But yeah, because by default, iOS 14 opted everyone out of ad tracking.

Greg Marshall 14:38 So what does that make sense that if they've lost 80, because I'm making the assumption that we were all optimizing for these events in the past, because of how powerful the data was, right? Yeah. But if you strip all that data away, would you be shortchanging yourself by optimizing for something that It only has 20 or 10%. Of Yeah, versus maybe something a little higher.

Blake Beus 15:04 Yeah, I, I would I definitely agree with that. But here's, here's the thing. Like, if you have a campaign with a purchase objective, and it's working, and it's profitable, don't turn that off based on what we're seeing here like that can work for you. A lot of this has to depend on aggregated data that you've that you've obtained inside your ad account. That historical data can it is used to make decisions, per ad account or whatever. But if if purchase events don't seem to be working Yep, then you need to start reimagining how you're doing and start optimizing for events inside the Facebook app. Got it. And then, and then find some sort of a way to follow up with some sort of direct offer or whatever, right? So you might have to think of things in two stages. And this isn't a new concept like this concepts have been around for a long time. But it's probably more relevant now than it has been is because a lot of those purchase events aren't getting reported back into into Facebook.

Greg Marshall 16:10 So here's, here's a question. So like, I've seen with a lot of my clients that have, I guess, you know, seasoned pixels, or you know, they've got a lot of data, the end goal of purchase or lead works perfectly. Yeah. And that's because they have 1000s and 1000s of those events that have happened previously. And even happened now. My question would be, what if someone has a brand new ad account? Yeah, I have heard opposing positions on this over and over again. And I'm curious on what your thought is, because my thought is, logically, if I have no data, and AD account, I should be starting at the highest level, and then work your way down. Yeah. But then you do I hear at least I don't know about you. But I hear from people that just say, always optimize for the bottom of the funnel, even on a brand new account. Yeah. And I don't know if I always agree with that. And I've heard this from like, legitimate people that I would consider good marketers. Yeah, I'm not saying they're completely wrong. Yeah. But it's an it's an idea that I don't know if I 100% am on board right with that.

Blake Beus 17:27 So here's what I would say. I would say there's a lot of seasoned marketers out there media buyers that are talented and really good at what they do. But they can still not have a great understanding of kind of inner workings. The reality is, is Facebook and Google, they don't publish. They don't publish the details about how everything specifically works. They will publish some best practices and give you some concepts, that data and information tends to always be skewed towards making them money. So keep that in mind. So sometimes their best practices. Their best, right? are best practices for them. And sidenote, how many times have you talked with like a Facebook? Like, they'll reach out to you if you spend enough money, they'll reach out, Hey, have a free consultation call. I do it every time. And every time I've had a call with them. I'm like, I'm pretty sure the information you're giving me is lining Facebook's pockets and not mine.

Greg Marshall 18:20 Well, I don't Okay. To go. I don't know how many times I actually instruct my clients. If you get reached out by a REP. RON. I remember thinking of this one client, where she was her Azur apps. I mean, I'm talking absolutely crushing. Yeah, cold trial. Spending. I want to say she was around $400 a day. Okay, three or 400. So not huge amounts. But also that's, that's a significant amount for a lot of businesses. Yeah. And so she was spending that and of course, Facebook calls, right? They reach out to her, they get on the phone call, and they tell her to change all of her bidding. And so she changed that. And the the conversions just like stopped, because of how she changed, you know, they made they're optimized for different things. And that was a big eye opener for me. This happened a few years ago, where I was like, This is not good. Because several times when I talk to him, some of the advice that they give you doesn't feel like they know they're talking it feels like they're just randomly telling you anything. I whatever the sales sales manager told him to say when you call

Blake Beus 19:35 everyone well, I think I think the reality is, is if you're not spinning serious money with Facebook, and when I say serious money, I'm talking like $10,000 A DAY PLUS, you're not getting someone who's on the phone with Facebook, you're not getting someone who's actually run serious as you're getting someone that's been through a six weeks training and they're regurgitating what they've heard and That's fine. Yeah, but in a test Yeah. And it's okay. Usually I take those calls, because I want to pick their brain about something and see what they say. And I just have this morbid curiosity, like every time, but we kind of got off topic that what were we talking about? So optimizing? Well use you said you have you've seen serious advertisers say you always optimize for the end result. Yeah. So their line of thinking is essentially, it's like, you always want to tell Facebook or Google exactly what you want. And that makes sense at a high level saying, I actually want purchases, like, I really do want that. So I want you to use your algorithm magic to find purchases. And I don't think that's entirely wrong. But if it's not working, it's not working. Yeah. The when, when the rubber meets the road, and you launch that campaign, and you're spending $100 a day and you're getting $20 in sales, that's not working. You can't keep up with that. And a lot of them, and I've talked not, not a lot of them. I've heard some say, Well, you just got to spend more spent, you got to lean in and that spending and I've tried that. And and sometimes that works, but not always. And then you're and then you you lose several $1,000. And you're thinking okay, this, this is not working. So it's all about testing what's going to work for you, your ad creative, your audience, all of these things are completely unique. Yep. Even if you have a brand new account, your picture that you're using is probably unique. The ad copy you're using is unique. The landing page or on your cart software that you're using. All of those things are unique, and means that your situation is unique. And so if you want to try testing something bottom of the funnel right here on a brand new account, go for it. And if it works run with that. But if it doesn't work, you need to turn that off. Don't spend more money. And then and then work on kind of building engagement audiences and then retargeting retargeting those engagement audiences? Well,

Greg Marshall 22:03 here's, and this is where the stats from? So you hear well, ad accounts are unique to your business. And I believe that because it does learn who your customers are. So what did you say if if I'm going into a new ad account, and it has nothing? And I tell it to get me a customer, but it literally has never had a purchase? Right, then? Are we shooting blind? Because the the flip side of that was I hear? They say Yeah, well, the algorithms have so much data on everyone else's ad account that they know how to put in front of someone. Yeah. But my question is, but do they really like so let's say I sell an $800 washer and dryer? How many $800 washer and dryer purchase Do you think Facebook is really getting throughout the world on ads can't be can't be that many,

Blake Beus 23:06 that's a pretty unique offer. And so I would say I would say that, if your offer is in line with a lot of other people doing the same thing. So let's pick on Realtors for a second. Like a lot of there's a lot of realtors out there. A lot of them use Facebook ads to get leads, it's a very short funnel, right? Like you can say, here's this house, check it out, give me a call, you know, enter email in so you could even use the on the on, you know, in paid Facebook lead little tool or whatever. And that works for realtors, partly because whenever they make a sale, their Commission's are so big compared to their ad spend that it totally works. And if you have an offer like that, that's very similar to lots of other things out there. But unique because it's you and you're in this geographical area, I can see putting a purchase or putting that lead at that event. That objective. At the end of the funnel, I can see that totally working on a brand new account right off the bat and because of historical data from from everybody, like that's a very easy thing for the algorithm to say to figure out like, it's easy for them to for me, it's easy for the algorithm to see that I've been looking at every house, someone posts on Facebook, and I'm definitely in the market for a house like that is an easy thing for them to determine. And it's an easy funnel, and it's an easy event for them to track. And there's lots of people doing that kind of strategy that could totally work. But if you're selling something unique, like a course that covers something or a membership, like the SM3 membership that we do. Those are those are definitely more unique things and there's not there's probably no historical data on other ad accounts relevant to that. So when I think

Greg Marshall 24:54 and here's here's a thought that I'm curious about, which is cool. Did you almost customize your ad account? Meaning? What if? Okay, let me go backwards. When you talk about ad accounts, right? Does an ad account, even if they don't get to your website, learn from video views and engagement? And all of that? Does that store somewhere? Yeah,

Blake Beus 25:23 yeah. Yeah. So Facebook, I know for a fact that I've actually read this, they can track if you just stop scrolling and look at a post, even if it's not a video or whatever, they don't give marketers access to that data, they do give marketers access to video view data, like, you know, 10 seconds or 25% figures that they give they give access to that they don't give access to the data have always stopped and read this post, right, they do give access to the data, have someone hit the like button, or they made a comment or engaged in some way, shape, or form, or on Instagram, if they viewed my my profile page, they give us access to that data. Oftentimes, it's high level access, but that they let us target those people. But they know if I stop scrolling and read something they know, if I opened the comment section and read some comments. They know if I know those people, and they use all of those data points to put something together. So they do learn a lot on your ad account. Even if some people aren't clicking on the ads and going to some

Greg Marshall 26:26 and they and they do they store this information for their own learning? Like, see, what I'm saying is do they tie the info? Kind of like if you had 1000 purchases last month? You know, it's like kind of in the ad account? Does that ad account also store that information? Or does it only store the events? Like you tell it to based off of view content Add to Cart purchase? Yeah, so

Blake Beus 26:50 they don't publish this? But I I'm very confident saying that? Yes, they do store Okay, all of that stuff. I mean, if you think about it, Facebook's specifically Google to their they are data brokers, like their data collectors, data warehouse platforms. And they definitely store that they may

Greg Marshall 27:17 aggregate that data a little bit. So they're not still storing every single little data point. But they may kind of aggregate that into some collective internal data they can use for decisions, but they do they do store it. Again, they don't publish this. Yeah. But I would be very surprised if they didn't store that and relate that to your ad account. Got it. Okay, because because in my head, I'm structuring a strategy of like, so you could take an ad account, and almost customize that ad account early on, like a brand new one, you could train it up, like an employee, and show it and start with here some, you know, engagement, video views, clicks, blah, blah, blah, then landing page, and then view content all the way to the purchase? And if you're patient enough, would it? In my opinion, wouldn't that be the most cost effective and more customized approach than because here's what I've seen before with new ad accounts. I'm thinking of this one particular ad account, I remember when I first launched, we were going after purchases, and I was almost getting no clicks on the ad saying, Okay, same ad. So remember, anytime I testing, it's with the same exact thing, same audience, you will always get no clicks, no engagements, no anything. I took the same ad post, Id moved it into a view content, suddenly started getting a lot more clicks, engagements, and people even moved down closer to purchase, versus starting out the purchase. That was like one of the first times I was like, that's interesting. It's the same audience same ad, nothing's changed. But when I told it to get a purchase, nothing happened.

Blake Beus 29:04 And I'm, I'm gonna guess you didn't say this. But I'm guessing the CPMs were way lower on the view contents, way, way, way, way, way, way lower. And that's the other thing, right? Like, a lot of companies out there with very, with a lot of data in their ad account, are optimizing for purchase because it's working for them. And they've put that time in. But those are more expensive eyeballs. Yeah, because it's harder for Facebook to determine who clearly fits in that group number one, and number two, there are more valuable set of eyeballs. Yes. Right. And so you're going to be bidding against people with bigger budgets when you optimize for purchase, which again, if you're optimizing for purchase, and you're getting a 2x or 3x Return On cold traffic, like keep doing that, yep. But if you're optimizing for purchase and you're not Yeah, then we got to read. We got to rethink that. The eyeballs for view content or whatever, those are going to be cheaper eyeballs to get to take action. And some people interpret that as a lower intent audience, which you'll hear, you'll hear a lot of advertisers talk about their lower intent intent audiences. But the reality is, is oftentimes, they're lower intent, simply because Facebook can't determine what their attempt intent is. They haven't done the things to give Facebook a clearer idea of what their intent is. And with the right message, the right the right hook, all these things we've talked about, we're still talking to people, even though we're using algorithm with the right hook and the right message and everything you can make that make that work and determine that intent for yourself. Through actual version.

Greg Marshall 30:55 Remember, over the summer, I share with you how I had a client that I think it was over last summer where I said, Yeah, we're actually optimizing. Well, we had multiple objectives that we're going after, right? But I had shared with you that one of the campaign's was the objective was view content. And it was actually getting purchases for the same cost as my purchase, right? Which that threw another like, like, that's interesting, because the view content, which is considered and the CPMs are way lower, is always, not always but mentioned many times from other marketers saying, Yeah, you don't ever really want to optimize for that, because they're not. They're lower quality eyeballs, right? That's essentially what they're implying. But the data in this account was showing otherwise. And so I was like, I don't know, in my mind, I've always interpreted Well, if an account was maybe early on, or whatever, it's better to go after that anyways. Because it's going to be cheaper. Because the other thing that you want, I think, for you, is you want to keep in mind, of what stage is the person giving you that advice? And definitely don't bring up that too. Yeah, because one thing that I have started to challenge a little bit is the people that give a lot of the advice, there are by no means purposely trying to misdirect you. There are only speaking from their current spent, right. Some of these companies are spending 1000 2000 3000 10,000 The horse a day, you're gonna do things a lot differently. It's kind of like, if the CEO of Coca Cola gave you advice on how to grow a business. It's almost like not applicable to you if you're started yesterday.

Blake Beus 32:50 Right? Yeah. CEO of Coca Cola could come in and be like, Okay, we've got this. I don't know craft store yet where you can buy like rubber stamps and things. And we're going to take this multinational, yeah, they're probably not going to do a very good job at data like, and that's the same thing. So you get a lot of these, these gurus and many of them have good information, right, like, and good intent, like they're not trying to screw you over. But many, many of them are will say things like, you know, this is what I did, and spending over $2 million in ads, yes. But if you're only spending 10 hours a day, that your strategies are totally different data different. Let me ask you this, though, I want to circle back to what you're saying you had the purchase event over the summer, and the view content event, and they were getting you similar purchase results, cost per purchase cost per purchase results. Let me ask you, in your opinion, which of those campaigns would would be more valuable in long run? And your Yeah,

Greg Marshall 33:50 and my opinion, it's just gonna go against everyone's opinion, I believe the view content on to me? Yeah, because it's gonna be a bigger audience. It's gonna be a bigger audience. I could feed it more data faster. Yep. It can optimize quicker. That's just my opinion, though.

Blake Beus 34:07 No, I would agree with you on that. And that's why I brought that up. Because a lot of people say cool, well, these are both working. So we'll shut off the view content, cuz we're gonna go just the purchase, because that's what these guys are saying. And it's working. But I would say the view content is probably going to get you more consistent results, like you. And I've run into this with purchase conversions. Sometimes it's all over the place. Whereas if you're optimizing for something else, sometimes it's way, way, way more consistent. Because you're using bigger, your CPMs are lower, so you're in front of way more people. So statistically, it should smooth a lot of these things out. And you're creating retargeting audiences that are much larger, that you can then retarget and you and I both know this, oftentimes, your initial cold traffic audiences, you'll get maybe maybe a 1x return on your row and sometimes a little less, whatever. Sometimes to x like if you've got a cool offer whatever, but then your retargeting ads will get 567 10x return on your row as row as. And that's where that's where all the magic happens. So if you can create that retargeting audience for cheaper, yes, and you're getting sales, I think that audience is bigger. And then you just launched a retargeting

Greg Marshall 35:18 campaign. Funny, funny follow up to that, that client forced me to turn off the view content. Like maybe back in September or whatever, okay, because they heard some other gurus say should only do purchases, right? And so we turned it off, and their sales tanked. Now we have to rebuild it back up. And the database is the data right there. We were getting that cost per purchase. It was working fine having the view content was in there. But once again, maybe a well in tenant guru pension for that to only optimized person get rid of the view content. They basically strong army to turn it off, even though I did. And then they saw an immediate drop and say,

Blake Beus 36:04 oh, so let me ask you this. You have the you were used to running both our purchase and the view contest, you show off the view content. And you saw a drop in so did you see a decrease in performance on the purchases? So so it's almost like the purchase campaign was being propped up? Yes. By the view content campaign because Facebook was basically stealing data from the view content campaign to determine who had the proper purchase intent. Correct. So it was essentially an an unintentional maybe retargeting. Yep, campaign is what might have been added. We

Greg Marshall 36:39 talked about this, where we run side by side campaigns with different objectives, and that they seem to give more signals and feed off each other. Yeah. And I don't think that's a bad thing. I think that's a good thing. Because you're simultaneously growing those retargeting audiences while giving the machine enough data consistently, to actually optimize versus when you do the purchase. And you don't run into this as much as into lead generation, lead generations easier. It's a pretty easy conversion, when you're asking people to pay money right away, you know, you get a lot more fluctuations and your results with the cost versus where one day it will be $15 A purchase for three or four days. And then another day, it's 68. And then another day, it's 30 minutes

Blake Beus 37:31 before and I've been there before. And that is frustrating, right? That's the stuff. That's the stuff that keeps you up at night. Yep. And that's the software. As soon as you wake up in the morning, you're checking your ad accounts, see where you're at that day. And that like that is that is frustrating

Greg Marshall 37:47 goes up and down. And I think one of the one of the thoughts when I set up that side by side campaign was what would happen if I feed enough data constantly to stabilize the account? And that's one of my theories on why, while those were running side by side, we almost never had a jump. And I think it's because of what you're saying it's stealing from the other campaign. And it's able to maintain that level of performance, right. And every day

Blake Beus 38:18 will end in traditional advertising wisdom, with with Facebook specifically was you don't want overlapping audiences. And every Facebook ad rep will tell you that and they'll say, Oh, you're

Greg Marshall 38:32 you're stealing from currently running? Yeah. Overlap. Right now. It's working very well.

Blake Beus 38:38 I personally do. Yeah, I think the only time that matters, is when you're using big budgets, right? Like, if you're spending maybe $1,000 a day, $2,000 a day. You you might run into where you need to stop overlapping audiences, especially if your audience isn't super big. But you can actually a lot of people don't know this, you can actually hop into Facebook, I can't remember where it's at. But you can see how much you're overlapping on those audiences and make that determination. The reality is on smaller, much smaller spends a few $100 a day or whatever. Your audience size is so big. Yeah, that that even if it's like a million, you know, million person audience, oftentimes we're targeting 2 million 3 million person audience. You're you're not you're not gonna overlap in a way that's going to make you a make it make an impact or waste your money, right.

Greg Marshall 39:31 It's overly I think it's like, what is it like an over exaggeration, right of the overlap? Because I have found even with clients spending just two or $300 a day I don't think overlap becomes a problem until you even hit 1000. Right? Right. So that doesn't apply to most average

Blake Beus 39:53 law. A lot of businesses will never never go past $1,000 a day because they simply just don't need to Like, you think about, I saw some ads for a roofing company, right? We've talked about roofing companies before, I had my roof redone a few years ago, because it was damaged in a hailstorm. And the bill was $25,000, or something like that, right? They don't need to spend $2,000 a day to get roofing clients, right? They can spend $200 a day and be profitable all day long, all day long, every day, at $200 a day and no need to spend more than that, because they can get more work than their crews can accomplish in any given amount of

Greg Marshall 40:35 time. And you bring up a good point to strategy because what I do think when we make these podcasts, my intent, and we've actually never really had this conversation gold, nice. The intent is to help more of the early stage to mid stage advertise. Right, right. So I want to speak almost on those terminologies to help that person either start and get going, or if they're getting going where to go. I don't necessarily intend to try to reach the $10,000 day spent.

Blake Beus 41:12 Right. You don't expect McDonald's to reach out to you and say, Hey, Greg, Greg, I was listening to your podcasts about we're spending $100,000 a day, how can we optimize that?

Greg Marshall 41:22 Exactly. I'm not intending for for that particular audience. Because at that point, they're typically not the audience that needs help. Right? The person who needs help, is the person to starch events here, and how to grow that to be a profitable

Blake Beus 41:39 business, right. And I wouldn't even say at that point. They kind of maybe don't even care about results, like exact, there's so big, and they're so big. And it's like that we have a marketing degree, have a marketing department and we have a marketing budget, let's spend it on ads. And we'll write we'll run some reports to show that we've done some stuff. But the reality is, they're making money regardless. And so they kind of they they don't need to know the nitty gritty details, because they're all getting their paycheck and going home at the end of the day anyway.

Greg Marshall 42:09 Yeah, that's a great point. Because they have an ad budget. They just need to spend it. Yep. So they're not really worried about

Blake Beus 42:14 now their job is to spend that much budget and make it look like they're doing something I've worked in corporate. That's really what that's really what most corporate jobs are, is to make it seem like to whoever it matters above you, that you're doing your job.

Greg Marshall 42:25 Yep. So I think that's what our podcast is all about is helping that person that's from the start to mid level tier that, you know, I would, we could almost put a number two, as far as running paid traffic. We could almost say the person that wants to get up to 1000 hours a day. profitably, yeah, yeah, up to 1000 hours a day profitably making that work. Pass that I think that's a different audience. That, you know, they're obviously they know, these standards, they wouldn't get to that.

Blake Beus 42:57 And that's something we could speak to pass to that Absolutely. A bit. Right. But you know, a lot of the principles, you if you're spending $1,000 A day profitably, you've already kind of worked out what's working for your ad account.

Greg Marshall 43:09 So that's kind of the target person. Yeah, that we really want to serve?

Blake Beus 43:14 Well, Greg, we got to wrap this up. This is one of our longer episodes, I'm just looking at the timer right here. And but I think this has been a really interesting one to go over. And it's something I feel like a lot of media buyers need to have a better understanding of, because it's very easy to just kind of throw stuff out there and see what sticks without having an understanding of sending the right signals to the ad platform, how to what the differences are and the nuances for a brand new account versus one that has conversions. We didn't talk about this a whole lot, but versus one that has like messy data. Yes. Right. They've run all these different offers that are not related at all. And now we've got some messy data, right? Yeah, definitely. Since we didn't really talk about that, I would kind of treat that like a new ad account. Yep. Is what I would do. And I would kind of start from the beginning and go from there. And then try to keep all the offers in line. And go from there. But yeah, so let's let's wrap this up. Greg.

Greg Marshall 44:12 Well, if you want to reach out to me book a free strategy session, go to Greg marshall.co. And Blake, Blake beus.com.

Blake Beus 44:19 There's the SM three group in there on social media marketing, organic and paid. Hop in there. Check that out.

Greg Marshall 44:27 Thanks for joining us. Talk to you later.

Blake Beus 44:30 All right, bye. That was that was a good one.

  continue reading

55 ตอน

Artwork
iconแบ่งปัน
 

ซีรีส์ที่ถูกเก็บถาวร ("ฟีดที่ไม่ได้ใช้งาน" status)

When? This feed was archived on October 28, 2023 04:09 (5M ago). Last successful fetch was on February 18, 2023 00:39 (1y ago)

Why? ฟีดที่ไม่ได้ใช้งาน status. เซิร์ฟเวอร์ของเราไม่สามารถดึงฟีดพอดคาสท์ที่ใช้งานได้สักระยะหนึ่ง

What now? You might be able to find a more up-to-date version using the search function. This series will no longer be checked for updates. If you believe this to be in error, please check if the publisher's feed link below is valid and contact support to request the feed be restored or if you have any other concerns about this.

Manage episode 329026442 series 2987873
เนื้อหาจัดทำโดย Blake Beus & Greg Marshall, Blake Beus, and Greg Marshall เนื้อหาพอดแคสต์ทั้งหมด รวมถึงตอน กราฟิก และคำอธิบายพอดแคสต์ได้รับการอัปโหลดและจัดเตรียมโดย Blake Beus & Greg Marshall, Blake Beus, and Greg Marshall หรือพันธมิตรแพลตฟอร์มพอดแคสต์โดยตรง หากคุณเชื่อว่ามีบุคคลอื่นใช้งานที่มีลิขสิทธิ์ของคุณโดยไม่ได้รับอนุญาต คุณสามารถปฏิบัติตามขั้นตอนที่อธิบายไว้ที่นี่ https://th.player.fm/legal

Listen on:

Transcript

Blake Beus 0:00 Okay, and then we'll just make sure we look up there. I did pretty good at looking up there during the I only good. Yeah. All right.

Greg Marshall 0:08 Well, we are live

Blake Beus 0:10 live ish. You're gonna see a recorded recording. But you you have some questions for ya. So cousin was like, let's do it. Let's do this.

Greg Marshall 0:20 We've had this conversation before off recording I think it'd be valuable now to do a recording but also to kind of go through strategy live since Yeah, yeah. So So yeah, so one of the things that I have seen the I want to challenge this is, so there's different bidding strategies, right? Yeah, cost per click CPA Max conversion all in each one of these ad platforms. And that's assuming that the algorithm is going to match up the result that you want. Right, right. Here's my question. So I ran a test the other day, and you know, me, I'm always testing, essentially my job as marketers. Yes. Yeah. 24/7. I ran a test the other day. And I bid for video views, but this is on Facebook, okay. And I measured it against a traffic campaign, where one objective you say, I'm looking for video views. Okay. The other one, you say I'm looking for traffic

Blake Beus 1:25 traffic. And to clarify traffic is pageviews. Right? Like someone clicks on a link? The page loads? That's the event that we're optimizing for? Correct?

Greg Marshall 1:34 And what I actually found was the video views campaign. were slightly better. Oh, really? Then the traffic, okay. And the here's the kicker, same ad copy, same audiences. So interesting as it was, so nothing changed the other

Blake Beus 1:51 variables with the same worked better in what way? Like what was your end goal,

Greg Marshall 1:55 the end goal was to see like the test was, which one would be better to run video views or traffic, vert to get to the page, right? One of them you're looking for a higher watching rate. Right? Because if technically, if you bid for video views, you're getting higher washers versus traffic. I, in my base off this test, I saw deeper washy rates and better click through rates to the website versus the traffic.

Blake Beus 2:28 And so when you say better your your definition of better is more more people going to the landing page. You're not like adding to cart purchasing. Correct? Correct. Because this is a blog post. And then it also says two blog posts. Yeah. So you set so your objective was to get people to hit that blog post? Yes. But then you told Facebook, on one campaign, I want the objective of video views. But that video has a link in it to go to the blog page. And then the second campaign was you told Facebook, I want people to go to the blog page. Right. So and then you measured which one actually got more people to the blog page. Correct. And the video views got more people to the blog page.

Greg Marshall 3:10 Yes. And so I don't know if that was just and I believe both ads were a video ad. This literally the same little Okay, identical, like no difference and no change in work, because there were existing posts. Okay, she's existing posts, okay. And so it made me think because then I did a dive into, well, what if you go to YouTube, YouTube ads, that's you can do video, essentially, video views or web, right? Traffic leads wherever. And I looked at a couple of clients that are running, and we were starting with video views. And their cost per clicks. Were basically the same as if you were bidding for like web traffic. Oh, really, or CPA, but the CPMs cost per 1000 impressions. Were lower on the video view of video in YouTube and YouTube. Okay. This is where I was really trying to think, Okay, the next test I'm running currently to see if there really is any difference is targeting the same audience, changing the bidding strategies and see where the ads show up. Okay? And the reason why I'm testing where the ads show up is my what is it? The hypothesis, okay, is if your if your targeting is what you want it to be, and you change bidding styles. Is it really different to bid on one versus the other algorithm wise? Okay, that's the question, okay. Meaning, if I'm going after the same audience, I keep getting the same cost per clicks or maybe not cheaper. Okay, on video view versus web tracking or CPA. Then what is the difference of between one versus the other. Okay, and why would you read The why would it matter? Yeah. If you're bidding one versus the other? Yeah.

Blake Beus 5:03 Okay. That's the hypothesis. Okay, here we go. So I, I have some thoughts. So a lot of people talk about algorithms, and we talk about algorithms a lot on here as well. And many of the people that talk about algorithms, I have never written an algorithm before. Which is fine. Like, that's totally fine. But if you've, if you've written some code and sat down and thought, Okay, I need to write, I need to bang out some code that's going to take a few things into account, and then provide a result. How do I do that? You've had to approach the problem a little bit different. And you have, you have to have a better understanding of how algorithms can work. Now, I've never written an algorithm algorithm as complicated. As well, Facebook and Google ad platforms have I mean, they're literally taking 1000s, maybe even 10s of 1000s of data points, and running it through their machine learning. To to to pop out, okay, here's the people that need to see this at whatever, right. But I have written algorithms before, for some, some basic use cases, which I feel like gives me some unique perspective on how algorithms work. Because a lot of people, they're just like, well, I'm testing this and this and this. And then they say, this is how the algorithm works. And they're like, Well, maybe, yeah, maybe. But we've got to think about the decisions that the coders were making and how, how they can detect data, right? So the very first thing and we say this over and over and over again, is you've got to make it easy for the algorithm to to make a decision. Yep. Okay. Because, again, there's there's 10s of 1000s of data points, is one of the reasons why if you have so you have a purchase conversion and fires at the end of a big long funnel, right? You're you're doing the Russell Brunson Click Funnels thing where, where they buy this thing, then there's a bump sale, and then there's an upsell offer, and another upsell offer. And then like, if they turn down this one, there's like a, another offer that's maybe a discount of one of the first offer has and, and all of this thing, and then they go through six or seven steps, and then they purchase. And that's where the event fires. That's very difficult for any algorithm to determine what the hell happened here. Yeah, like like, because it's so far removed from the ad, click, or from a measurable ad event, that there's so many places for the ball to get dropped. Right.

Blake Beus 7:39 And so one of the things, I think the very clearly the easiest thing right now to say is, it's so much easier for Facebook, Google, whatever platform to measure events that happen inside their system, as opposed to something that happens off their site. So like a page load that happens off YouTube, you go to a different platform, even if you're on your phone, and you're inside the YouTube app, you're going to a separate webpage. But if you're measuring a result that happens inside their platform, it you they have access full access to that data, even with iOS 14. Yeah, right? Like, iOS 14 is not going to block Facebook's ability to know how much of a video you watched inside of Facebook's app. Because when you install the app, you give that app permission to track these things. And iOS 14 can block them doesn't block that right. And so whenever you have like a video view objective, it's so much easier for the algorithm to determine objective has been met. Yeah. Which is why in my opinion, you're gonna see always see lower CPMs got it. Now, we've talked about this before. And if you're thinking about this, you the listener CPMs, are essentially how you're charged. Yeah, that that is is an A very important metric. And the the lower the CPM, it's a stands for cost per 1000 views or impressions, right? The lower that CPM mean means you get more people to look at your ads for a lower cost. Yep. Therefore, you have more of an opportunity for clicks, more of an opportunity to determine in you know, that that person's intention, all of those things. So I think the difference in CPMs is probably why you're getting cheaper cost per clicks, and why you're getting more pageviews because more people are actually seeing that. And I think that's probably most relevant inside Facebook because they were really destroyed with iOS. And now iOS 15. Google was like, and we've been through this before we got this not a problem, but Facebook was absolutely right. They'll completely blindsided right So I think that's probably why you're seeing the best results, and probably why you'll very likely continue to see better results.

Greg Marshall 10:07 Well, in that, here's, here's my question then. And maybe I have always looked at it incorrectly, okay? Do you get a less lesser quality audience bidding on different strategies? So, for example, let's say I'm going after the same audience. Do you believe that the algorithms would show like by, say, video views versus purchase? Does that just completely show it to a different audience, almost like my mind in the past would think, well, if I do video views, it's essentially never going to show it to a buyer. Right? And if it is only going to show people who bought Yeah,

Blake Beus 10:51 so this is how this is how I look at when you set up targeting in any any ad platform, it will give you an estimated audience size, right? Yeah. So let's say you you set up targeting, targeting several different interests. And Facebook says you have an audience size of a million people based on your parameters, right. And then you run two ads to the exact same thing with different objectives. One is a purchase objective, and one is saying, video views objective? Yeah, your budget is probably most people's budget probably isn't big enough to blow through a million people have that audience in 24 hours. Now. It's just like, that's not gonna happen. So So part of what the algorithm does is decides, Okay, we've identified this audience based on the parameters you've set, but which people in that audience do we actually show it to? They don't just randomly pick people they're trying to optimize, like, they really do want to get you good results, because they want you to keep spending money with them. Right. And so I would guess that it's quite possible that they will show purchase objective campaigns to a different subset of that audience than a video view objective. And the reason is, is because one of the data points in there is is that they're using as an algorithm is, how likely is this person to to make a purchase? How likely is this person to watch a video, and we all have slightly different behaviors on Facebook, and some people are more likely to watch 1015 20 seconds, five minutes of a video, and some people are more likely to click on a button and add something to the cart and make a purchase. And so Facebook's going to say well, this person feels like they're ready to make a purchase. So let's put it in front of them. And, and they're also like, well, this person is the kind of person that tends to like to watch videos, so let's put it in front of them. But I think where I think there's a disconnect there. Yeah. Because I think it's much harder for an algorithm to say this person seems like they're likely to make a purchase. Yeah, I don't think that's an easy call for an algorithm.

Greg Marshall 13:01 Here's my argument. Yeah. Okay. Are my agreeance on what you're saying, but an argument to the general marketing population? Wouldn't you say that if there's less data, because of the privacy issues being fed back in? What did that mean that the algorithm has lost a significant amount of those n data points? Yeah.

Blake Beus 13:24 Right. Absolutely. Especially on Facebook.

Greg Marshall 13:27 So with that being said, Wouldn't it be you have you should change your thought process with the changes that have happened? Yeah, absolutely. So that's kind of my thing is I'm curious to actually, that just kind of sparked my curiosity, because I was like, Well, if you notice, like, if the algorithm or the datasets or whatever, are just not as intelligent and what is it 80% has been taken away? Don't they say? Like automatically 80% of that out? Yeah, maybe even more? Yeah,

Blake Beus 13:59 because iOS by default is opting out of ad tracking. So so it like, what is it? It's like, on my website, it's almost 90% of all traffic is mobile traffic these days? And the difference between iPhone and Android? It's probably 60%. iPhone 40%. Android, depending on industry, some industries are, are more, right, like Realtors way higher up on iPhone than Android. But developers may be higher on Android. But so it kind of kind of depends. Yep. But yeah, because by default, iOS 14 opted everyone out of ad tracking.

Greg Marshall 14:38 So what does that make sense that if they've lost 80, because I'm making the assumption that we were all optimizing for these events in the past, because of how powerful the data was, right? Yeah. But if you strip all that data away, would you be shortchanging yourself by optimizing for something that It only has 20 or 10%. Of Yeah, versus maybe something a little higher.

Blake Beus 15:04 Yeah, I, I would I definitely agree with that. But here's, here's the thing. Like, if you have a campaign with a purchase objective, and it's working, and it's profitable, don't turn that off based on what we're seeing here like that can work for you. A lot of this has to depend on aggregated data that you've that you've obtained inside your ad account. That historical data can it is used to make decisions, per ad account or whatever. But if if purchase events don't seem to be working Yep, then you need to start reimagining how you're doing and start optimizing for events inside the Facebook app. Got it. And then, and then find some sort of a way to follow up with some sort of direct offer or whatever, right? So you might have to think of things in two stages. And this isn't a new concept like this concepts have been around for a long time. But it's probably more relevant now than it has been is because a lot of those purchase events aren't getting reported back into into Facebook.

Greg Marshall 16:10 So here's, here's a question. So like, I've seen with a lot of my clients that have, I guess, you know, seasoned pixels, or you know, they've got a lot of data, the end goal of purchase or lead works perfectly. Yeah. And that's because they have 1000s and 1000s of those events that have happened previously. And even happened now. My question would be, what if someone has a brand new ad account? Yeah, I have heard opposing positions on this over and over again. And I'm curious on what your thought is, because my thought is, logically, if I have no data, and AD account, I should be starting at the highest level, and then work your way down. Yeah. But then you do I hear at least I don't know about you. But I hear from people that just say, always optimize for the bottom of the funnel, even on a brand new account. Yeah. And I don't know if I always agree with that. And I've heard this from like, legitimate people that I would consider good marketers. Yeah, I'm not saying they're completely wrong. Yeah. But it's an it's an idea that I don't know if I 100% am on board right with that.

Blake Beus 17:27 So here's what I would say. I would say there's a lot of seasoned marketers out there media buyers that are talented and really good at what they do. But they can still not have a great understanding of kind of inner workings. The reality is, is Facebook and Google, they don't publish. They don't publish the details about how everything specifically works. They will publish some best practices and give you some concepts, that data and information tends to always be skewed towards making them money. So keep that in mind. So sometimes their best practices. Their best, right? are best practices for them. And sidenote, how many times have you talked with like a Facebook? Like, they'll reach out to you if you spend enough money, they'll reach out, Hey, have a free consultation call. I do it every time. And every time I've had a call with them. I'm like, I'm pretty sure the information you're giving me is lining Facebook's pockets and not mine.

Greg Marshall 18:20 Well, I don't Okay. To go. I don't know how many times I actually instruct my clients. If you get reached out by a REP. RON. I remember thinking of this one client, where she was her Azur apps. I mean, I'm talking absolutely crushing. Yeah, cold trial. Spending. I want to say she was around $400 a day. Okay, three or 400. So not huge amounts. But also that's, that's a significant amount for a lot of businesses. Yeah. And so she was spending that and of course, Facebook calls, right? They reach out to her, they get on the phone call, and they tell her to change all of her bidding. And so she changed that. And the the conversions just like stopped, because of how she changed, you know, they made they're optimized for different things. And that was a big eye opener for me. This happened a few years ago, where I was like, This is not good. Because several times when I talk to him, some of the advice that they give you doesn't feel like they know they're talking it feels like they're just randomly telling you anything. I whatever the sales sales manager told him to say when you call

Blake Beus 19:35 everyone well, I think I think the reality is, is if you're not spinning serious money with Facebook, and when I say serious money, I'm talking like $10,000 A DAY PLUS, you're not getting someone who's on the phone with Facebook, you're not getting someone who's actually run serious as you're getting someone that's been through a six weeks training and they're regurgitating what they've heard and That's fine. Yeah, but in a test Yeah. And it's okay. Usually I take those calls, because I want to pick their brain about something and see what they say. And I just have this morbid curiosity, like every time, but we kind of got off topic that what were we talking about? So optimizing? Well use you said you have you've seen serious advertisers say you always optimize for the end result. Yeah. So their line of thinking is essentially, it's like, you always want to tell Facebook or Google exactly what you want. And that makes sense at a high level saying, I actually want purchases, like, I really do want that. So I want you to use your algorithm magic to find purchases. And I don't think that's entirely wrong. But if it's not working, it's not working. Yeah. The when, when the rubber meets the road, and you launch that campaign, and you're spending $100 a day and you're getting $20 in sales, that's not working. You can't keep up with that. And a lot of them, and I've talked not, not a lot of them. I've heard some say, Well, you just got to spend more spent, you got to lean in and that spending and I've tried that. And and sometimes that works, but not always. And then you're and then you you lose several $1,000. And you're thinking okay, this, this is not working. So it's all about testing what's going to work for you, your ad creative, your audience, all of these things are completely unique. Yep. Even if you have a brand new account, your picture that you're using is probably unique. The ad copy you're using is unique. The landing page or on your cart software that you're using. All of those things are unique, and means that your situation is unique. And so if you want to try testing something bottom of the funnel right here on a brand new account, go for it. And if it works run with that. But if it doesn't work, you need to turn that off. Don't spend more money. And then and then work on kind of building engagement audiences and then retargeting retargeting those engagement audiences? Well,

Greg Marshall 22:03 here's, and this is where the stats from? So you hear well, ad accounts are unique to your business. And I believe that because it does learn who your customers are. So what did you say if if I'm going into a new ad account, and it has nothing? And I tell it to get me a customer, but it literally has never had a purchase? Right, then? Are we shooting blind? Because the the flip side of that was I hear? They say Yeah, well, the algorithms have so much data on everyone else's ad account that they know how to put in front of someone. Yeah. But my question is, but do they really like so let's say I sell an $800 washer and dryer? How many $800 washer and dryer purchase Do you think Facebook is really getting throughout the world on ads can't be can't be that many,

Blake Beus 23:06 that's a pretty unique offer. And so I would say I would say that, if your offer is in line with a lot of other people doing the same thing. So let's pick on Realtors for a second. Like a lot of there's a lot of realtors out there. A lot of them use Facebook ads to get leads, it's a very short funnel, right? Like you can say, here's this house, check it out, give me a call, you know, enter email in so you could even use the on the on, you know, in paid Facebook lead little tool or whatever. And that works for realtors, partly because whenever they make a sale, their Commission's are so big compared to their ad spend that it totally works. And if you have an offer like that, that's very similar to lots of other things out there. But unique because it's you and you're in this geographical area, I can see putting a purchase or putting that lead at that event. That objective. At the end of the funnel, I can see that totally working on a brand new account right off the bat and because of historical data from from everybody, like that's a very easy thing for the algorithm to say to figure out like, it's easy for them to for me, it's easy for the algorithm to see that I've been looking at every house, someone posts on Facebook, and I'm definitely in the market for a house like that is an easy thing for them to determine. And it's an easy funnel, and it's an easy event for them to track. And there's lots of people doing that kind of strategy that could totally work. But if you're selling something unique, like a course that covers something or a membership, like the SM3 membership that we do. Those are those are definitely more unique things and there's not there's probably no historical data on other ad accounts relevant to that. So when I think

Greg Marshall 24:54 and here's here's a thought that I'm curious about, which is cool. Did you almost customize your ad account? Meaning? What if? Okay, let me go backwards. When you talk about ad accounts, right? Does an ad account, even if they don't get to your website, learn from video views and engagement? And all of that? Does that store somewhere? Yeah,

Blake Beus 25:23 yeah. Yeah. So Facebook, I know for a fact that I've actually read this, they can track if you just stop scrolling and look at a post, even if it's not a video or whatever, they don't give marketers access to that data, they do give marketers access to video view data, like, you know, 10 seconds or 25% figures that they give they give access to that they don't give access to the data have always stopped and read this post, right, they do give access to the data, have someone hit the like button, or they made a comment or engaged in some way, shape, or form, or on Instagram, if they viewed my my profile page, they give us access to that data. Oftentimes, it's high level access, but that they let us target those people. But they know if I stop scrolling and read something they know, if I opened the comment section and read some comments. They know if I know those people, and they use all of those data points to put something together. So they do learn a lot on your ad account. Even if some people aren't clicking on the ads and going to some

Greg Marshall 26:26 and they and they do they store this information for their own learning? Like, see, what I'm saying is do they tie the info? Kind of like if you had 1000 purchases last month? You know, it's like kind of in the ad account? Does that ad account also store that information? Or does it only store the events? Like you tell it to based off of view content Add to Cart purchase? Yeah, so

Blake Beus 26:50 they don't publish this? But I I'm very confident saying that? Yes, they do store Okay, all of that stuff. I mean, if you think about it, Facebook's specifically Google to their they are data brokers, like their data collectors, data warehouse platforms. And they definitely store that they may

Greg Marshall 27:17 aggregate that data a little bit. So they're not still storing every single little data point. But they may kind of aggregate that into some collective internal data they can use for decisions, but they do they do store it. Again, they don't publish this. Yeah. But I would be very surprised if they didn't store that and relate that to your ad account. Got it. Okay, because because in my head, I'm structuring a strategy of like, so you could take an ad account, and almost customize that ad account early on, like a brand new one, you could train it up, like an employee, and show it and start with here some, you know, engagement, video views, clicks, blah, blah, blah, then landing page, and then view content all the way to the purchase? And if you're patient enough, would it? In my opinion, wouldn't that be the most cost effective and more customized approach than because here's what I've seen before with new ad accounts. I'm thinking of this one particular ad account, I remember when I first launched, we were going after purchases, and I was almost getting no clicks on the ad saying, Okay, same ad. So remember, anytime I testing, it's with the same exact thing, same audience, you will always get no clicks, no engagements, no anything. I took the same ad post, Id moved it into a view content, suddenly started getting a lot more clicks, engagements, and people even moved down closer to purchase, versus starting out the purchase. That was like one of the first times I was like, that's interesting. It's the same audience same ad, nothing's changed. But when I told it to get a purchase, nothing happened.

Blake Beus 29:04 And I'm, I'm gonna guess you didn't say this. But I'm guessing the CPMs were way lower on the view contents, way, way, way, way, way, way lower. And that's the other thing, right? Like, a lot of companies out there with very, with a lot of data in their ad account, are optimizing for purchase because it's working for them. And they've put that time in. But those are more expensive eyeballs. Yeah, because it's harder for Facebook to determine who clearly fits in that group number one, and number two, there are more valuable set of eyeballs. Yes. Right. And so you're going to be bidding against people with bigger budgets when you optimize for purchase, which again, if you're optimizing for purchase, and you're getting a 2x or 3x Return On cold traffic, like keep doing that, yep. But if you're optimizing for purchase and you're not Yeah, then we got to read. We got to rethink that. The eyeballs for view content or whatever, those are going to be cheaper eyeballs to get to take action. And some people interpret that as a lower intent audience, which you'll hear, you'll hear a lot of advertisers talk about their lower intent intent audiences. But the reality is, is oftentimes, they're lower intent, simply because Facebook can't determine what their attempt intent is. They haven't done the things to give Facebook a clearer idea of what their intent is. And with the right message, the right the right hook, all these things we've talked about, we're still talking to people, even though we're using algorithm with the right hook and the right message and everything you can make that make that work and determine that intent for yourself. Through actual version.

Greg Marshall 30:55 Remember, over the summer, I share with you how I had a client that I think it was over last summer where I said, Yeah, we're actually optimizing. Well, we had multiple objectives that we're going after, right? But I had shared with you that one of the campaign's was the objective was view content. And it was actually getting purchases for the same cost as my purchase, right? Which that threw another like, like, that's interesting, because the view content, which is considered and the CPMs are way lower, is always, not always but mentioned many times from other marketers saying, Yeah, you don't ever really want to optimize for that, because they're not. They're lower quality eyeballs, right? That's essentially what they're implying. But the data in this account was showing otherwise. And so I was like, I don't know, in my mind, I've always interpreted Well, if an account was maybe early on, or whatever, it's better to go after that anyways. Because it's going to be cheaper. Because the other thing that you want, I think, for you, is you want to keep in mind, of what stage is the person giving you that advice? And definitely don't bring up that too. Yeah, because one thing that I have started to challenge a little bit is the people that give a lot of the advice, there are by no means purposely trying to misdirect you. There are only speaking from their current spent, right. Some of these companies are spending 1000 2000 3000 10,000 The horse a day, you're gonna do things a lot differently. It's kind of like, if the CEO of Coca Cola gave you advice on how to grow a business. It's almost like not applicable to you if you're started yesterday.

Blake Beus 32:50 Right? Yeah. CEO of Coca Cola could come in and be like, Okay, we've got this. I don't know craft store yet where you can buy like rubber stamps and things. And we're going to take this multinational, yeah, they're probably not going to do a very good job at data like, and that's the same thing. So you get a lot of these, these gurus and many of them have good information, right, like, and good intent, like they're not trying to screw you over. But many, many of them are will say things like, you know, this is what I did, and spending over $2 million in ads, yes. But if you're only spending 10 hours a day, that your strategies are totally different data different. Let me ask you this, though, I want to circle back to what you're saying you had the purchase event over the summer, and the view content event, and they were getting you similar purchase results, cost per purchase cost per purchase results. Let me ask you, in your opinion, which of those campaigns would would be more valuable in long run? And your Yeah,

Greg Marshall 33:50 and my opinion, it's just gonna go against everyone's opinion, I believe the view content on to me? Yeah, because it's gonna be a bigger audience. It's gonna be a bigger audience. I could feed it more data faster. Yep. It can optimize quicker. That's just my opinion, though.

Blake Beus 34:07 No, I would agree with you on that. And that's why I brought that up. Because a lot of people say cool, well, these are both working. So we'll shut off the view content, cuz we're gonna go just the purchase, because that's what these guys are saying. And it's working. But I would say the view content is probably going to get you more consistent results, like you. And I've run into this with purchase conversions. Sometimes it's all over the place. Whereas if you're optimizing for something else, sometimes it's way, way, way more consistent. Because you're using bigger, your CPMs are lower, so you're in front of way more people. So statistically, it should smooth a lot of these things out. And you're creating retargeting audiences that are much larger, that you can then retarget and you and I both know this, oftentimes, your initial cold traffic audiences, you'll get maybe maybe a 1x return on your row and sometimes a little less, whatever. Sometimes to x like if you've got a cool offer whatever, but then your retargeting ads will get 567 10x return on your row as row as. And that's where that's where all the magic happens. So if you can create that retargeting audience for cheaper, yes, and you're getting sales, I think that audience is bigger. And then you just launched a retargeting

Greg Marshall 35:18 campaign. Funny, funny follow up to that, that client forced me to turn off the view content. Like maybe back in September or whatever, okay, because they heard some other gurus say should only do purchases, right? And so we turned it off, and their sales tanked. Now we have to rebuild it back up. And the database is the data right there. We were getting that cost per purchase. It was working fine having the view content was in there. But once again, maybe a well in tenant guru pension for that to only optimized person get rid of the view content. They basically strong army to turn it off, even though I did. And then they saw an immediate drop and say,

Blake Beus 36:04 oh, so let me ask you this. You have the you were used to running both our purchase and the view contest, you show off the view content. And you saw a drop in so did you see a decrease in performance on the purchases? So so it's almost like the purchase campaign was being propped up? Yes. By the view content campaign because Facebook was basically stealing data from the view content campaign to determine who had the proper purchase intent. Correct. So it was essentially an an unintentional maybe retargeting. Yep, campaign is what might have been added. We

Greg Marshall 36:39 talked about this, where we run side by side campaigns with different objectives, and that they seem to give more signals and feed off each other. Yeah. And I don't think that's a bad thing. I think that's a good thing. Because you're simultaneously growing those retargeting audiences while giving the machine enough data consistently, to actually optimize versus when you do the purchase. And you don't run into this as much as into lead generation, lead generations easier. It's a pretty easy conversion, when you're asking people to pay money right away, you know, you get a lot more fluctuations and your results with the cost versus where one day it will be $15 A purchase for three or four days. And then another day, it's 68. And then another day, it's 30 minutes

Blake Beus 37:31 before and I've been there before. And that is frustrating, right? That's the stuff. That's the stuff that keeps you up at night. Yep. And that's the software. As soon as you wake up in the morning, you're checking your ad accounts, see where you're at that day. And that like that is that is frustrating

Greg Marshall 37:47 goes up and down. And I think one of the one of the thoughts when I set up that side by side campaign was what would happen if I feed enough data constantly to stabilize the account? And that's one of my theories on why, while those were running side by side, we almost never had a jump. And I think it's because of what you're saying it's stealing from the other campaign. And it's able to maintain that level of performance, right. And every day

Blake Beus 38:18 will end in traditional advertising wisdom, with with Facebook specifically was you don't want overlapping audiences. And every Facebook ad rep will tell you that and they'll say, Oh, you're

Greg Marshall 38:32 you're stealing from currently running? Yeah. Overlap. Right now. It's working very well.

Blake Beus 38:38 I personally do. Yeah, I think the only time that matters, is when you're using big budgets, right? Like, if you're spending maybe $1,000 a day, $2,000 a day. You you might run into where you need to stop overlapping audiences, especially if your audience isn't super big. But you can actually a lot of people don't know this, you can actually hop into Facebook, I can't remember where it's at. But you can see how much you're overlapping on those audiences and make that determination. The reality is on smaller, much smaller spends a few $100 a day or whatever. Your audience size is so big. Yeah, that that even if it's like a million, you know, million person audience, oftentimes we're targeting 2 million 3 million person audience. You're you're not you're not gonna overlap in a way that's going to make you a make it make an impact or waste your money, right.

Greg Marshall 39:31 It's overly I think it's like, what is it like an over exaggeration, right of the overlap? Because I have found even with clients spending just two or $300 a day I don't think overlap becomes a problem until you even hit 1000. Right? Right. So that doesn't apply to most average

Blake Beus 39:53 law. A lot of businesses will never never go past $1,000 a day because they simply just don't need to Like, you think about, I saw some ads for a roofing company, right? We've talked about roofing companies before, I had my roof redone a few years ago, because it was damaged in a hailstorm. And the bill was $25,000, or something like that, right? They don't need to spend $2,000 a day to get roofing clients, right? They can spend $200 a day and be profitable all day long, all day long, every day, at $200 a day and no need to spend more than that, because they can get more work than their crews can accomplish in any given amount of

Greg Marshall 40:35 time. And you bring up a good point to strategy because what I do think when we make these podcasts, my intent, and we've actually never really had this conversation gold, nice. The intent is to help more of the early stage to mid stage advertise. Right, right. So I want to speak almost on those terminologies to help that person either start and get going, or if they're getting going where to go. I don't necessarily intend to try to reach the $10,000 day spent.

Blake Beus 41:12 Right. You don't expect McDonald's to reach out to you and say, Hey, Greg, Greg, I was listening to your podcasts about we're spending $100,000 a day, how can we optimize that?

Greg Marshall 41:22 Exactly. I'm not intending for for that particular audience. Because at that point, they're typically not the audience that needs help. Right? The person who needs help, is the person to starch events here, and how to grow that to be a profitable

Blake Beus 41:39 business, right. And I wouldn't even say at that point. They kind of maybe don't even care about results, like exact, there's so big, and they're so big. And it's like that we have a marketing degree, have a marketing department and we have a marketing budget, let's spend it on ads. And we'll write we'll run some reports to show that we've done some stuff. But the reality is, they're making money regardless. And so they kind of they they don't need to know the nitty gritty details, because they're all getting their paycheck and going home at the end of the day anyway.

Greg Marshall 42:09 Yeah, that's a great point. Because they have an ad budget. They just need to spend it. Yep. So they're not really worried about

Blake Beus 42:14 now their job is to spend that much budget and make it look like they're doing something I've worked in corporate. That's really what that's really what most corporate jobs are, is to make it seem like to whoever it matters above you, that you're doing your job.

Greg Marshall 42:25 Yep. So I think that's what our podcast is all about is helping that person that's from the start to mid level tier that, you know, I would, we could almost put a number two, as far as running paid traffic. We could almost say the person that wants to get up to 1000 hours a day. profitably, yeah, yeah, up to 1000 hours a day profitably making that work. Pass that I think that's a different audience. That, you know, they're obviously they know, these standards, they wouldn't get to that.

Blake Beus 42:57 And that's something we could speak to pass to that Absolutely. A bit. Right. But you know, a lot of the principles, you if you're spending $1,000 A day profitably, you've already kind of worked out what's working for your ad account.

Greg Marshall 43:09 So that's kind of the target person. Yeah, that we really want to serve?

Blake Beus 43:14 Well, Greg, we got to wrap this up. This is one of our longer episodes, I'm just looking at the timer right here. And but I think this has been a really interesting one to go over. And it's something I feel like a lot of media buyers need to have a better understanding of, because it's very easy to just kind of throw stuff out there and see what sticks without having an understanding of sending the right signals to the ad platform, how to what the differences are and the nuances for a brand new account versus one that has conversions. We didn't talk about this a whole lot, but versus one that has like messy data. Yes. Right. They've run all these different offers that are not related at all. And now we've got some messy data, right? Yeah, definitely. Since we didn't really talk about that, I would kind of treat that like a new ad account. Yep. Is what I would do. And I would kind of start from the beginning and go from there. And then try to keep all the offers in line. And go from there. But yeah, so let's let's wrap this up. Greg.

Greg Marshall 44:12 Well, if you want to reach out to me book a free strategy session, go to Greg marshall.co. And Blake, Blake beus.com.

Blake Beus 44:19 There's the SM three group in there on social media marketing, organic and paid. Hop in there. Check that out.

Greg Marshall 44:27 Thanks for joining us. Talk to you later.

Blake Beus 44:30 All right, bye. That was that was a good one.

  continue reading

55 ตอน

ทุกตอน

×
 
Loading …

ขอต้อนรับสู่ Player FM!

Player FM กำลังหาเว็บ

 

คู่มืออ้างอิงด่วน