Sometimes you can be doing everything right and yet you see your CPMs and CPAs rising. What gives? Hear from Josh Henry-Hicks, Senior Campaign Manager at Facebook ad agency MuteSix, on why this happens and how to stop it. From the Facebook data breach to the end of the quarter - find out what's hurting your bottom line and how to correct it.
Episode Transcript — It's Not You, It's Facebook - How to Combat Increasing CPMs
Speaker 1: You're listening to the Spend 10K a Day podcast brought to you by the performance marketing experts at Mutesix. This is your source for cutting edge insight into the world of online advertising from the team with more Facebook case studies than any other agency on the planet. Here are your hosts, Steve Wise and Stuart Anderson.
Steve Wise: Hello everyone, welcome to another awesome episode of the Spend 10K a Day podcast. If you're in Los Angeles over the next couple of days, hope you're going to stay warm, and stay toasty because it's raining hard outside. I think in a past career, I was a weather man, but thank God I am not in a past career. Today we have an awesome topic. We're going to be talking about trends inside of Facebook ads, and joining me today, I have one of our senior awesome campaign managers here at Mutesix, as well as one of our full funnel analysis, came all the way from Australia to be on this podcast today. Actually, he lives in LA now, but he's from Australia and he's got a much better accent than all us other Americans in here. Josh Henry Hicks.
Josh Henry H: What's up Steve? Good to be here.
Steve Wise: Yeah, definitely man. And before we jump in, I always talk about people with two last names, you know. You got to be careful of those two last name type people, you know. If two last names, just look at them a little more closer.
Josh Henry H: It's the Henry Hicks. It's the Henry and the Hicks together.
Steve Wise: Cool. So, in our team meetings, we always talk about understanding trends inside of Facebook. It's really important to really understand trends. Obviously, there's only so much you could understand from a trend analysis perspective because it evolves so rapidly. But what we see is, trends that always ... nine times out of 10, always come correct.
Obviously number one, when the quarter is close to ending, we see that the gross CPMs go up. What that means is even deeper, it's costing more and more money to place an ad on the newsfeed. So, the amount of money it takes to reach a thousand people on average goes up. And that's not even counting in, how relevant your ad is or how much virality. That's just across the board. It's going up because there's a lot of budget coming in at the end of the month, and obviously number two, Facebook actually places ... I've been told this, I'm not 100% confident, but toward the end of the quarter, they place less ads to make it more volatile on the newsfeed. Obviously it's not confirmed 100%, but that's a theory that I have. What do you think about that Josh?
Josh Henry H: Yeah man. End of quarter is always crazy for CPMs. You know, we see it across the board. It happens every quarter. Some of the big companies with big budgets tend to clear the decks, and empty out their budgets, and we see that across multiple accounts. Especially even if you have like a medium sized ad budget, you see that effect and that flows through to CPAs. And so for me I think, one of the big talking points is this. Is like how do you combat that? How do you, as a strategist come up with ways to combat that increase in CPMs. Because it's harder to make the same amount of money as you were making even a couple of weeks ago.
Steve Wise: Yeah. I think, it's very important to understand why CPMs go up, why there is an uptick. Obviously there's two different types of advertisers on the newsfeed. There's people who buy eyeballs, and there's people who buy performance marketers like us. The people who buy eyeballs or the brand [inaudible 00:03:50] brand marketers, they have budgets they need spend. If they don't spend those budgets, they won't get similar budgets the next quarter.
So they're heavily focused on figuring out how to allocate all those brand budgets toward the end of the quarters, so they could allocate just so they get more budget the next quarter. So they have to spend that money. And that's what brings a lot of ivolatility to or volatility should I say to the newsfeed. Also top of that Facebook, if you'd been watching the news, they're obviously being investigated for data breach, which I think also has a negative impact on the newsfeed.
I personally ... now this is a podcast around conspiracy theories. So none of this is verified. I want to be very clear that none of this is verified. But I have a conspiracy theory that when there is ... and I've seen this multiple times, because Facebook has been the news and you know, for being investigated for this, for that. But I feel that when there is any type of investigation going on, Facebook serves less ads. So in essence the price gets commoditized up and there actually is more people bidding for less amount of inventory.
I think there's a reality around that because of the way ... now Facebook's being asked to show their faces when it comes to what they do with user data. And that's a very scary thing for Facebook because if they're have to show exactly how they use user data when it comes to marketing, when it comes to privacy, I think that there's going to be a lot of stuff coming out that's going to scare people, might scare people away from actually spending time and money and effort on Facebook. What do you think about that Josh?
Josh Henry H: Yeah man. You're 100% right. Coming from my perspective and I'm in the weeds, I'm in the campaigns every day, and we see that. We see it now and there's no explanation for it. But when there's troubled waters, you hear that there's the problems with Facebook. And whether it's the privacy issue that's going on right now at the moment, or something else. I feel like we see the effect and if the CPM is one trigger, but you know, there are other triggers that we see, which are telling us that, our ads are being less effective for a period of time.
So it's very, very interesting. And as a marketer it just puts pressure on your strategy and you need to have an effective strategy. I feel like the people that that really don't have that full funnel approach at this time this is when they can really see the effect on their campaigns. So really just highlights you need to be smart and you need to have that effective full funnel strategy on Facebook.
Steve Wise: So the two questions is, how do you ... number one, Facebook doesn't do a good job of telling you, Mr Facebook marketer or Mrs Facebook marketer that there's volatility in the market. There is no way of knowing. The only way you actually know is when you look at your campaigns and you see CPM costs. So we'll talk a little more about how do you know when there's volatility coming inside of your campaigns. How do you know when it's either you messing up or Facebook messing up? This is a question I get a lot. Well, is it the CPMs going crazy? Is it the ... our ads are just aren't performing? And this is why there's a lot of tall tale signs of showing you know, volatility in the market. Number one, you're looking at longterm attribution over 28 days.
Obviously if you're showing less ads to less people than your actual CPA price over the 28 day attribution period from a view and the order click is obviously going to go higher. Your [inaudible 00:07:33] because you're showing less and less and less ads. So I'm always looking at the longterm attribution when analyzing data inside of Facebook when I see a bump up, because that means is that this problem didn't actually start today, maybe it started two weeks ago when I saw it, maybe that's when increased volatility happened. Because remember there's a life cycle to how long it takes for Facebook to actually record conversions.
Number two, auction insights, there's a great tool, Josh am sure you ... and we talk about this in our meetings about using the auction insights tool to see the amount of first impression users that you're reaching. That's a big key because the cheapest users are going to be people who have shown prior purchase intent and Facebook really wants to show your ad to most relevant people. So using the auction insights tool and really understand the percentage first impressions and a plethora of other interesting information is also a way you could really tell volatility. And Josh you may think they [inaudible 00:08:27] when it comes to measuring volatility?
Josh Henry H: No, not really. I think you bang on like the longer attribution window is key. I think obviously Facebook insights is a plethora of information. CPM is certainly your main game when looking at competition and trying to understand what's happening with the Facebook algorithm. For me, I like to look across multiple accounts. Obviously not everyone can do that, comparing different types of industries, seeing is this isolated to a particular sector or is this kind of across the board? That helps me form up an understanding. Okay well, Don't change your strategy too drastically because it is something that we are seeing across accounts and across sectors or if it is isolated then we need to dig a little deeper and start coming up with some strategies that can try and counter the Data we're saying.
Steve Wise: Yes. That's interesting. I mean I assume that not everyone has the ability to really see if we manage so much advertising spent here at Mutesix and there's so much resources that we have from looking at all the users ... from looking all the data that we have coming into our back end, we're able to really use predictive analysis to help our partners predict trends.
But, if you don't have that data, you know, what can you do to help predict trends? I think that's a question that we're trying to answer on this podcast. And secondly, the bigger question I get is like alright "I know something's up, I know there's an issue, I know there's volatility, I'm seeing having delivery issues, I still have the hit numbers, I still have quarterly goals, I still have stuff I need to hit. What can I do beyond just Facebook, beyond just Instagram to really reach my goals? Even if delivery is not doing too well." I think that's a question that everyone is asking on a daily basis. Josh, I have a couple of thoughts on that I want to share. Do you have any idea? Do you have any thoughts you want to share on that topic?
Josh Henry H: Yeah. I think for me, I think you needed do some controlled testing and what I mean by that is. You're seeing this data, behaving in a certain way and say if it's specifically CPM data, which then will flow across through to your cost per clicks, your ad to cart, your purchase, return on ad spend, CPA et Cetera. For me, it's about identifying is this in a controlled area. So for example, say you've had your four best audiences that have been running for 30 days plus or even 14 days plus and they're your best performers and these things just churn out profit and normally let's say this is in a prospecting type scenario and these audiences start to ... we see that CPM increase. My thought is you really need to do some controlled testing to say, "okay is this the audience type?" Whether you're looking at look a likes or whether you're looking at interest targeting. Is this, selected to the particular audiences that I have live. So whether it's ...
Steve Wise: Josh, can you give me an example of a controlled test that you've done in the past?
Josh Henry H: Yeah.
Steve Wise: Just so the audience of people listen to podcasts can really understand the way you think about testing.
Josh Henry H: Yeah. So basically our strategy here at Mutesix is we'll get a split test within Facebook. Often we'll use the actual split testing functionality that is within power editor or ads manager with whichever one you're using. And essentially we'll set up the same ad, basically the same budgets, the same timeframe. We'll try to reduce as many variables as possible, same creative and really just use, in this case it would be our audiences. And so we would use the audiences being the variable and just really looking at that CPM and doing that on multiple occasions with multiple different audiences.
So basically you're just getting an understanding and a feeling for okay, this audience segment has a higher CPM than this one. Is there any learnings that can be made from that? Whether that's the type of targeting? How big the audiences is obviously a very interesting factor. For example, maybe you did a small audience versus a larger audience. So I think just being able to have those two strategies or audiences played off against each other, really is going to give you an understanding of where are these CPMs going? What audiences are actually showing a lower CPM especially during this turbulent times? I think, it's a little easier when things have gone back to normal, but I think when push comes to shove you want to be able to, kind of clearly identify which audiences are not as affected as most.
Steve Wise: Yup. And I'm going to turn the topic a little bit of what can you do specifically on the question of during turbulent times to really show performance. What can I do? And I'm going to turn this focus more or less toward retention, toward figuring out how do you get more out of your retention based audiences? Obviously, we talked earlier on previous podcasts about looking at your email list and trying to figure how do you incentivize customers to rep your brand? How do you incentivize them to share your brand? Talk about your brand? Whether it's creating video content where there's creating social buzz, how do you turn your best customers into your, into kind of like salespeople?
I think we've had a lot of success from our email side and our messenger side of really incentivizing people to really talk about the brands that we partner with. And I think that's really important because I think not a lot of people think about retention. You don't think about this thing called the K factor where if you're able to ... what the K factor means is like, you know, your viral factor. If you have one person and you're really good on retention, that one user could equals six new users. So I want to turn the focus toward figuring out retention. Like how could retention help during tumultuous times?
Josh Henry H: Well, to bring it full circle if your CPMs are high, obviously this is normally going to be affecting your acquisition or your prospecting strategy the most. And so really if you can kind of make it up from that remarketing or retention areas of the funnel this can really help outweigh what you're seeing in the prospecting side. So I think from what I've seen from some of that clients that have done some great things is a really solid referral program where uses are incentivized to share whether it's a post purchase or some sort of value where the user can extract value by promoting your brand. And exactly like you said Steve, then they're marketing you're your brand on your behalf and again it just comes full funnel. Like once you got everyone investing to acquire someone. If you can acquire someone and then they become your advocate and promote more sales, you're getting more bang for buck on that initial investment, which is the main end of the game.
Steve Wise: Cool. That's interesting. I'm just thinking in my head about different areas that you could really start focusing on as you start seeing a volatility on the newsfeeds. So number one, really looking into your email audiences trying to figure out which specific email triggers key sent people to incentivize them to bring more and more people to the funnel, turning them into virtual salespeople. Obviously you're looking into doing different split tests from an audience side on Facebook. Really trying to figure out what can you do to increase the virality of your Facebook posts.
Josh Henry H: That's a good point.
Steve Wise: That's something I'm always thinking about when I look at volatility in the market. What that tells me is that if I go back into Facebook and I'm trying to find how to get my ads to get more impression share, I have to do a better job of creating posts that generate virality.
Josh Henry H: Yeah. High relevancy scores is bringing down your advertising costs. It's just serving your ads further. So again audiences is a massive pot, testing in that area, but also creative like, vitality and creative. What can we do to get this ad to be served more, to be served further will it needs to be more engaging to the user groups. So again, creative is probably the next most important thing. How are we getting the vitality? But also how are we creating that direct response nature where we're actually promoting a purchase.
Steve Wise: Or actually go look at past post IDs and start really ... from having difficulty in the marketplace, I'll actually go back and look at old post IDs and try to set some of those live posts, ids that have a ton of virality. That are still relevant to the time of day. So I'm always trying to figure out what's going to get the edge. I'm also going to be split testing, you know, conversion objectives. I think that if what you're currently doing is stale, the thing you have to be doing is you have to be testing out different ways to lower the gross CPM and bring people into the funnel.
Obviously, I'm not always looking at CPM, but CPM is a tall tale sign of the volatility in the Facebook marketplace. It might not always be that, okay, I'm doing something wrong that means I have a high CPM. Sometimes it's the reality is in the quarter that it is very volatile. I think we're always learning a lot. I think we're always going to be sharing the intel that we have from managing over $100 million a year in spend. That's the goal of this podcast, that's a mission, well the core missions and myself and my team is to really, knowledge share as much as humanly possible. Josh, before we wrap up, do you have any other topics you want to discuss or the is anything else you want to tell our amazing audience?
Josh Henry H: I think raised a good point like just as you were summering there you talked about the objective right and obviously the Facebook objective basically tells the Facebook algorithm find these certain types of people. Whether it's, someone that's most likely going to convert, someone that's most likely are going to be clicking an ad. Someone that's most likely going to view a video. And obviously with different conversion objectives comes different CPMs.
As the algorithm is trying to find different types of people. So that's probably something that we could talk in a little more detail, which is very important is to make sure, yes, we're testing creative. Yes we're testing audiences, but also from an objective point of view, a lot of advertisers can get very conversion heavy and I would say, you got to be careful with that. You still want to be testing that. At the end of the day, the end goals is obviously sales, but we need to pass them through that funnel. So we'll definitely be testing, different types of conversion objectives, whether that's clicks, whether that's video views reach, especially when we're seeing this pressure on CPM, I think it kind of opens up more avenues than we perhaps would consider on and everyday.
Steve Wise: Cool. All great points. I think that you always be testing and I think that goes without saying. You hear that a lot. I think structured testing, everyone has their own theory on testing. Having a theory or having a hypothesis when you test it is really important. I don't like the thought of testing just because it's the right thing to do with testing. I always train the team here and I always have a hypothesis. Really try and figure out what, Maybe if there is volatility, maybe there's a way to hack volatility, you know? Maybe there's a way to use volatility to your own advantage.
Josh Henry H: Yeah.
Steve Wise: [inaudible 00:20:54] and I'm always trying to ... it's really important to think as a marketer, always think about how do you get that unfair advantage over your competitors.
I'm going to wrap things up, Josh, before we say ... do you want to say goodbye? You want to say is your first podcast top [inaudible 00:21:09]
Josh Henry H: It's good to be here podcast cherry.
Steve Wise: Podcast cherry. There you go.
Josh Henry H: Thanks for having me. Yeah, you're welcome back in Ozzy land any anytime soon.
Steve Wise: Yeah, he's just referring to the last time I spoke at our conference in Australia, it was an amazing experience. For those of you that have never been to this island off the coast of Antarctica. It's an amazing adventure. Great people, very good breakfast. They know how to do breakfast over there.
Josh Henry H: Brekkie.
Steve Wise: They're commonly referred to as Brekkie. I appreciate everyone, can wait for the next podcast. It'll be a surprise. We have an amazing guest. Josh, thanks for joining. [crosstalk 00:21:51] have an awesome Wednesday, actually Thursday, should I say take care of everyone.