175: Amazon Attribution, List Cleaning, FB Ads ROI?: Our Top Questions Answered
Posted by December 4, 2018on
Today’s episode is a little bit different, in a good way. We’ve had some great questions recently in the Brand Growth Experts Membership.
They were so good we decided to answer some of them for our listeners.
We answer questions including how to attribute Amazon sales, how to weed out inactive email subscribers, what type of an ROI you should expect on your Facebook Ads, and Andrew explains the best way to set up a budget and test Facebook ads. This is a meaty episode. Enjoy!
- 7:33 Member Question #1: How do I identify inactive subscribers and re-engage them?
- 10:40 Optimal open rates and other key metrics you should look for when you’re trying to reactivate your email list.
- 12:25 Member Question #2: What is the best approach for setting up a budget and testing Facebook ads?
- 12:49 The first thing you need to do before setting up a Facebook ad test.
- 13:13 Andrew shares three approaches to setting up creative testing for your Facebook ads.
- 14:50 Testing your Facebook ad creative with campaign budget optimization.
- 15:33 How to test the performance of your Facebook ads: the number of ads you should have in an ad set and how much audience overlap to look for.
- 16:30 Determining your daily budget for each ad.
- 18:11 Member Question #3: Facebook advertising and tracking spending off Amazon.
- 19:18 Techniques for determining the value of your off-Amazon campaigns.
- 22:48 Member Question #4: How to determine whether you’re achieving a good return on ad spend and the right metrics to look for.
- 25:00 Why your return on ad spend goes down as you spend more.
Links and Resources
Become a Member
If you liked this episode, you’re going to love the Brand Growth Experts Membership. It’s a community of top ecommerce business owners and marketers who I coach one-on-one to help scale up their businesses. Together we’ll create a plan that will help you scale up your business, and then I’ll help you execute it.
If you want to make sure you’re growing as quickly and sustainably as possible, click here to learn more. Hope to see you on the inside!
Today’s episode is brought to you by the always wonderful Hotjar. Hotjar is an incredible tool that allows you to see what’s happening on your site with potential customers with recorded sessions. You get to basically go back and look at what’s happening, what people are doing, where they’re dropping off, why they’re not adding something to cart on mobile, on the desktop, and you can actually understand that.
Hotjar allows you to get in-the-moment visual feedback, see how people are using your website, and uncover insights to make the right changes. Hotjar is absolutely one of my favorite tools out there and if you’ve not used it, we definitely recommend checking it out.
You can go to EcommerceInfluence.com/hotjar to learn a little bit more.
This episode is also brought to you by Klaviyo. If you’re running an e-commerce business and sending emails to your customers, you should be using Klaviyo. It will help you find out who your best customers are and target them one-to-one to make more money.
I’ve been using Klaviyo since they were just two employees. Now they have a team of 150 and are rolling out new features almost weekly. If you aren’t already a customer, head over to www.ecommerceinfluence.com/klaviyo and you’ll get a free trial + priority on-boarding.
Austin Brawner: What's up everybody? Welcome to another episode of the Ecommerce Influence podcast. My name is Austin Brawner.
Andrew Foxwell: I'm Andrew Foxwell. Hey, man, how are you?
Austin Brawner: I am doing very well here. We are into December, into winter weather and it feels, feels pretty good. How about yourself? What have you been up to?
Andrew Foxwell: Yeah.
Austin Brawner: What have you been up to? I haven't talked to you that much recently.
Andrew Foxwell: Yeah, I'll be honest, I have been throwing pots and getting back to the old pottery wheel, so haven't done it since high school and I needed another, wanted another creative outlet. I needed something else to get my brain out work. I think a lot of us can identify with that. I walked into the local studio, pottery studio, which just opened down the street from my house and this dude's like 26 and this is his side hustle. Started this studio and took a class, got back into it and I got a membership now. Yeah, I'm producing quite a bit of work. I'm not going to say it's good, but I mean nice bowls, cups, plates. Not making any vases yet, but it's possible I could see that becoming thing a thing.
Austin Brawner: It's possible. I mean that's what, I got a funny story about pottery class. I actually took a pottery class in college because I thought I need an easy class to take. It was, I think it was the junior year, something like that. I had a lot of demanding classes. I need an easy class, I'm gonna take pottery. Wrong choice.
Andrew Foxwell: Turns out it's not easy. Yeah.
Austin Brawner: Not easy, not easy. The professor decided to stick it to everybody who thought they were in this for an easy credit and I've spent more late nights on the pottery wheel that semester. It'd be 2:00 AM and it'd be full of all the students trying to get everything done. This guy was crazy in the best way. He had all these crystals. He was in his hand at all times. He was just jiggling and he would talk about his crystal energy and he would have these, he would talk about lucid dreaming all the time and he would wake up in this state of his right, as he woke up, he'd write these really long journals and he said he trained himself to be in a lucid dreaming state for minutes, so it was like 10, 15 minutes. He'd be just kind of just going right on the page and that was an inspiration for artwork. Super interesting guy, but yeah man, good to see you're back into the pottery game. It's a fun, stress release activity.
Andrew Foxwell: Yeah, yeah, yeah. What about you? What have you been up to?
Austin Brawner: Oh, I right now I am trying, I'm just back in Austin and getting back. I've been on the road so much in the last month that I've just been being consistent. Getting to the gym, working, trying to get kind of feeling healthier again. It's so hard to stay healthy when you're on the road and feel good after basically being on the road for almost a month. I got back and I was like I need to just eat some normal food and chill out before the holidays because I wanted to go feeling good into Christmas, which is something that a lot of people can relate to. It's hard around this time of year when you're quite busy to feel good about yourself and not go into Christmas so tired, so that's what I've been doing. It's been good.
Andrew Foxwell: Yeah, that's-
Austin Brawner: I'm feeling great.
Andrew Foxwell: That's good. Good time to be consistent on those habits I think for sure but-
Austin Brawner: Also been in the membership, which is one of the things that I'm trying to be consistent in there working on the membership answering questions, which is the whole summary of the whole topic we're going to be talking about today is we've had some really good questions in the brand growth experts membership recently and we want to take some time to answer some of those really good questions for you guys who are the listeners.
Andrew Foxwell: Yeah, yeah, definitely, it's interesting. I mean the conversations in there are thought-provoking and Rose always sends a nice weekly recap too, which is really helpful of here's what people have been talking about and the conversation is, it's deep. I mean, it's things that are, I think on a lot of people's minds and I don't know, I really, really like being a part of it. I learn a lot in terms of everything from Shopify to email stuff that I didn't know, and a lot of good Facebook questions to of it's really a safe space. It's the tree of trust I think is a good way to describe it.
Austin Brawner: Yeah. No, it's nice because it's kind of a combo. You can get some of the bigger picture questions answered and then also go in and get some questions that have been on your mind for a while that might not be that advanced, but something that you've been thinking about that chances are a lot of other people have been thinking about as well. That kind of leads us into a question we've been getting a lot recently and we've gone into in multiple different threads because it's a question that while initially, it doesn't seem like that advanced of a question, it has elicited a lot of follow up and feedback. The general question is how do I decide when to weed out inactive email subscribers? Right. This is a question that has been asked in three or four different ways. How do I weed out inactive subscribers, how do I find inactive subscribers? How do I re-engage or reactivate a cold email list?
That's been a big theme recently because it's quite confusing, right? You've got this large email list and for your entire life as a marketer you've been told grow, grow, grow, grow your email list. As that list continues to grow there's gonna be parts of it that are inactive, that are no longer engaged. When you start sending emails to those people it knocks down your open rate, it knocks down your click-through rate, and overall your email metrics are kind of dropped down to a level that is not really where you want them to be at. The big question is, which I'm going start with, how do you figure out, you got a large email list, but a lot of them are not engaged according to a Klaviyo's unengaged segment, which means they haven't opened in 180 days. What should we do?
Really what I recommend people do is not to suppress or delete anybody right off the bat unless they're, have an invalid address that is hard bounced, or maybe they've never received the email. Those people are fine to get rid of, but the place, if you've got a bunch of people that are unengaged you might want to run a reactivation campaign to those people first and if you're interested, if you want to, if you haven't, if you've got an email list or a group of people that haven't really received the emails from you for a while, now what you can do is you can go and start reactivating those people. One thing to look at is you might want to validate your emails. You can use a tool like Zero Bounce or some Zero Bounce alternative that you upload your list to and you can scrub it for email bounces and spam traps.
If you ever have a cold email list and you haven't emailed it for a long time, and you're thinking about emailing it, first make sure you validate those people and use a tool like that.
Andrew Foxwell: Okay.
Austin Brawner: Once you have that you're going to want to get rid of some of the questionable ones and if you're going to want to try to reactivate an email list of people that have been cold, start with a small sample, maybe 1,000 or 2,000 people and send them a test email. Then you're gonna want to do is you want to see from that small sample the results of that campaign. If you've got good solid metrics you can prepare to send to the full list. Now, what are the metrics you're looking for? You're looking for unique open rates, unique clicks, bounce rate, unsubscribe rate, and spam rate. Now, if you will, people always ask what's a good open rate? A good open rate for me, I like to keep it around 16% or more. Most of the time I like to keep it above 20%, but if you're at 16% or more you're in a good in a good spot. That's also coming from Klaviyo as well. They've looked at it. If you're at 16% or above that's healthy.
Unique clicks around 3% or more, bounce rate less than .5%, and an unsubscribe rate of .3%. If you hit and you're, you send this test email and your metrics are good then you can go ahead and email the rest of those people who are less active or unengaged. If the sample, random sample hits and it's, the results are good then I'll feel comfortable emailing the rest of them on a reactivation campaign. If they go through, don't reactivate, then at that point, I will clean the list.
Andrew Foxwell: Okay. Okay. Interesting. I never knew that to be honest. I saw a couple of your responses about that within the membership area, but I was curious to how it all worked so that's really helpful. Thank you for enlightening me.
Austin Brawner: Well, it's a complicated little, it's complicated. There are more resources in the membership but yeah, it's something that you definitely, definitely want to do. Andrew, why don't you go through a question that you answered in membership recently?
Andrew Foxwell: Yeah. It's a question that a lot of people have asked and I'll read this specific question and then kind of give you an idea. The specific question is, tell me the approach that you use when setting up a budget and testing of ads. I have 12 ad groups I want to test and try to separate them. We're also running dynamic product ads. Generally, how do you test, and then how do you set up the budget for a test? I could answer this a lot of different ways. The first thing I'll say is when you're setting up a test on Facebook the first thing you have to understand is, and establish is what do I want to do from this? What's the goal of what I'm trying to set up? If you're trying to set up and understand creative, then you should set it up in a different way than if you're trying to set it up and test for performance, and creative being less optimal or less as part of the testing stream.
An example of what I mean is if, let's say you're testing creative and you say, Andrew, my goal is to understand what people really are interested in, then what you want to do and what I do is let's say I have five ads, I launch those ads into my prospecting ad set. I'll launch them and I will turn one on for a period of 48 to 72 hours while the others are off and then I will turn that first one off, and I'll turn the second one on. I basically repeat that process over a week and a half. That's the way that I do creative testing is I essentially turn it on or off every 48 hours or 72 hours. It depends on if I feel like I've had good data on it. Then I compare the data at the end.
This method I can, you can also launch, you could do, let's say you had two different objectives. You had one in a conversion objective to a prospecting audience and one in a page post engagement objective and they're the same ad, you could aggregate the social proof on them and that's a good way to double check, which makes sure that the objective doesn't have all to do with it, right? That it's like, this is actually an interesting ad. Facebook thinks it's interesting. A cross objective too. If you're trying to do creative testing, that's the way that I go about it. Now, the scientific way of going about creative testing is there's actually the split test feature, right? You could basically take and say to Facebook, I want to actually statistically separate these two groups and it'll show one the ad in the other, it'll hold back and not that, that group wouldn't see the ad. That's another option that you have.
I like it the first way because to me it's a little bit more controlled, not giving into Facebook for too much control. The third option for creative testing that you have is you have campaign budget optimization you could use and you could also use dynamic creative. You could set up campaign budget optimization, which is if you've not used it an interesting methodology to try right now. You set the budget at the campaign level, you build out your ad sets, it distributes that spend across the ad sets that are working, and then at the ad set level you upload multiple creatives and multiple lines of copy, and Facebook combines them to find the ones that it thinks are the best. There are three ways that you can do creative testing to find out. Now if you say, look, my goal is to get, just to test for performance. I've done everything that I can. How do I set up a test?
If I'm doing this, then you're going to say, okay, look, I never put more than two ads in an ad set. Unless I'm spending more than $500 a day on that ad set two is about enough because Facebook will always choose one quickly. Then I will, to have two in the ad set and I will determine how many different interest targeting groups I'm trying to go for while combining the things that are super similar. An example of this is I want to make sure that I don't have more than 30% audience overlap on my audiences. If I don't I'm going to separate those two things out. Justin Marshall's talked about this on his first episode. We've talked about this before, where you want to make sure that if there is more than 30% overlap you're putting those two audiences in an ad set together.
You're generally trying to think about consolidation and you're trying to think about what are the core things I'm doing, and I'm taking my budget of let's say it's $500 a day. Let's say there's 30 days and, let's say the budget's $10,000 for the month. You divide that by 30 days. Okay? $10,000 divided by 30 days is $333. You divide that by, let's say you got five ad sets. Each ad set is going to get $66 a day. That's the way that you can budget and start to set it up. It's just simply dividing and splitting it out, and then you can optimize from there. Those are some ways that you can think about how you want to. A lot of times when people say they want to test they try to set out multiple different interest groupings or multiple different lookalikes, and they don't think about the combinations that you need to be able to kind of bring things together. If you want, the way that I like to do too is you can clump in interests, you can clump together lookalike audiences, you can clump together previous engager audiences and then you can clump together remarketing.
You kind of have different segments and four different core segments that aren't going to necessarily be competing with one another.
Austin Brawner: Makes sense. I think that's definitely on the minds of a lot of people, especially as we're going into, well, I mean right now hopefully if you're spending right now, you're spending on tests that have won, but again, always something that you're going into. Thanks for kind of breaking that down. I have another question that we got earlier this, I guess it was earlier this month. I'm actually going to read the question because I think that they wrote it out in a very clear, clear way and it's going to be relevant to a lot of people. It says for those of you selling on Amazon, hoping to get some input on an issue we're facing with a new project. Since we sell a single product which is priced at under $10 and are doing FBA. It's a product that should generate replenishment orders in the futures so we need to carefully monitor costs of acquisition and lifetime value.
As part of the launch strategy we had been running Facebook ads on Instagram, on Facebook and Instagram, and driving traffic. First two are non-ecommerce enabled website and then encouraging users to click through to our Amazon listing to complete their transactions. While we're able to re-target traffic due to the interim step of having them visit our site, there doesn't seem to be any way to apportion the Amazon sales to the paid off Amazon traffic. Has anyone got techniques for determining the value of our paid off Amazon campaigns? Interesting. Interesting question. Right? And we've got a couple of different answers in here from members talking about what they're doing.
The first, there's a couple of ways to go about it, right? The real question just to sum it up is where they're driving traffic through their site, it's not e-commerce enabled, and then people are clicking through and going to the actual Amazon store to complete the transactions. How do we actually determine the value of what we're spending off Amazon? A couple of suggestions for people that are interesting, right? One of them says, hey, we currently track off Amazon by having a specific promo code. What they're doing is they are taking the ad, they're directing it to a lead page which is non-ecommerce enable, and they get an email opt-in for a discount, and then drive them to Amazon. That way they can then pull the information and figure out how many people bought from the paid traffic off Amazon.
Also, the other thing that you can do is use Amazon affiliate links to track if you're not using discounts. Basically, you can put a unique Amazon affiliate URL and figure out that somebody clicked on that link and made a purchase from there. One key thing with that is to make sure that you have another affiliate commission, another affiliate account so that you're not showing Amazon that you are getting affiliate commissions on your own products. Those are two suggestions. The third suggestion that we got was that you can create your own Amazon store in FBA or vendor central, and then change the URLs from there. Three different ways to solve that issue. If you're running into the same type of issue there's different ways to track it and figure out how to track some of the off Amazon paid and get a return if you are then driving them to Amazon.
Hopefully, that's helpful if you're in that same situation.
Andrew Foxwell: Fascinating. Absolutely fascinating. Amazon is a beast. Oh, my gosh.
Austin Brawner: So man, It's still a wild west, man. Hacking and there's so much stuff going on in Amazon that you're still just figuring out. It's black, to a certain extent it's a black box.
Andrew Foxwell: Yeah. It's interesting the ways that people have figured out the tracking side of it. Yeah, it's interesting the way one of my best friends was just actually on Shark Tank recently with her business, the KombuchaShop.com.
Austin Brawner: Oh, yeah.
Andrew Foxwell: It got funded, it was awesome, but she is a big Amazon seller and she sees through that the loyalty of the people that come back after buying on Amazon's insane. People that come back to her site to buy different teas and blends and stuff. It's really cool. She's been able to close that gap with different technologies of a kind of closing the loop. Anyway, it's just really fascinating as its own space.
Austin Brawner: She got funded?
Andrew Foxwell: She got funded, yeah. She got funded by Barbara and, Barbara Corcoran and the founder of Spanx.
Austin Brawner: Okay. She didn't get sharked then? I haven't seen the episode.
Andrew Foxwell: Yeah, she was a guest shark, so pretty, pretty amazing. Anyway, it was cool. You should check it out and she gives a shout out to Madison, Wisconsin which was awesome.
Austin Brawner: Nice. The Midwest pride there.
Andrew Foxwell: Yeah, yeah, yeah, yeah. Another common question that we get in the forum in the membership is about return on ad spend. A lot of people, this is like a black box because so many people give incredibly inflated results and a lot of the advice around this particular topic is what's a good return on ad spend, is built on people that are Facebook marketers giving advice. I heard a Facebook marketer who I won't mention on a podcast I just listened to, very well known podcast, say that he was getting 17 cent email leads. I'm just like, that's awesome but you're super famous. Everybody knows your name and you're targeting Facebook marketers. I'm just saying for the record that's not super helpful. It's always good to have an idea of a metric of what return on ad spend is actually good. There's a new Facebook attribution tool which we're going to talk about and do another flash episode about soon that's coming out, but we can help to kind of answer more of these questions.
Here's the way that I think about this. For me, if you are under $10,000 in spend, if you are at a three to four x, that's where you'd want to be ideally. Now there are different parts of this that are going to be challenging. There's a client that we work with that's at $15,000 a month and their average order value though is only $21. It's a very low-cost deodorant product. Okay? They're at a two x, which is huge. That's great. That's a fantastic place to be, right, because they have such low average order value, but their lifetime values are going to be, they are a lot greater. That's kind of the metric. If you're at that under $10,000 a month spend being at a three x is really good. Now if you're at something over that, you're at like a 10 to 20 x or excuse me, $10,000 to $20,000. If you're scaling then looking at something like a two and a half x is going to be fantastic.
There are accounts that I work with and am close to that are spending $300,000 to $400,000 a month and their role as that they're looking at is, and that's really great, is a 1.9, 1.6, 1.7. It depends on what you're doing. Couple of things to mention here, one is your return on ad spend would go down as you spend more because you're hitting people that are less likely to be interested in your products. Okay? That's, you can't just assume that another 100,000 people care as you scale up. That's one thing that as you scale return on ad spend will go down, which is why making sure that you're targeting previous purchasers, you're targeting your email list on Facebook, you're hitting people to raise their average order value and their lifetime value is going to be really helpful.
The other thing is that time of year has tons to do with this, right? You're listening to this in December. I mean, you listen to this in December obviously Q4 is nuts. If you're in Q4 there are some people that it has to be five x or greater, right? I just worked with one client in a pre-black Friday thing and that was a, they were at a 15x, which is absurd. It has a lot to do with that and their competition's high, et cetera, et cetera, et cetera. People are in the mood to buy. As we're moving into January and into Q1 it's just not, people aren't gonna be that interested as much as they would have been as earlier because they just got a bunch of stuff, they probably don't need to be shopping right away. Maybe they will, but-
Austin Brawner: Unless you're in the fitness industry and then that's back to what you're saying-
Andrew Foxwell: Unless you're in the, and that's true. Unless you're in the fitness industry. So true.
Austin Brawner: It all depends a little bit on the product because where you are, what type of product, your time of year. It totally makes sense.
Andrew Foxwell: Right, yeah, so that's an important thing to do. Those are kind of the metrics that I think about and I'm looking at in terms of return on ad spend, and some of you may listen to this and say, Foxwell, that's nuts. I need to make at least a five X. I've been killing it on a five x, at a four x for years. You're a joke, which this is feedback I've received. Here's the reason I say this is because it's hard to assign a return on ad spend across all these verticals, but I want you to understand that if you're getting a three x or you're getting even a two x at scale, that's good. That's something to be proud of. Competition is high. It's a challenge. Know that that's okay. If you're getting a four or five x, then fantastic. You're doing something right and you've really found a good audience. Yeah, I think that's just important to mention and it's important to think about kind of a setting your expectations around returns.
Austin Brawner: No question. It also is so important too when you start thinking about these numbers and if you're talking about two or three x return on ad spend being really, really solid, well you got to make sure that the product that you have and the business that you have lends itself to repeat purchases so that you can continually grow with a business that hit two to three x return on ad spend. If you can-
Andrew Foxwell: Right, absolutely.
Austin Brawner: -that's, if your product is a one and done and you're just, it's a money grab type product, well understand you probably needed to be at four to five x, but I'd really recommend you start thinking about what can we do to fundamentally change the business to make it so that you have repeat purchases. What can you offer?
Andrew Foxwell: Right and here ... totally. I completely agree with you, and here's the final point on this. A lot of you, we have a lot of listeners. Now we're over 19,000 listeners a month. Thank you very much. Seriously, thank you for listening. We have a lot of listeners in other parts of the world, so this is a way that if you're in the US you can horizontally scale into these other countries is a good idea because there's, because it's not as competitive, but if you're in one of these countries specifically, let's talk about Australia.
A lot of Australians listen to the show. If you're in Australia then a three x is actually pretty normal at this point, right, but you're seeing it declining. We're sort of at the front end of this with competition in the US and Instagram, I was talking to an advertiser in Australia the other day and he said Instagram is just still not 100% there, which is fine. Thinking about if it's not 100% there what are the ways that you can be smart based on what you've heard from us talking about moving into that to kind of be an early adopter to keep that row as high as it starts to decline and as competition rises?
Austin Brawner: 100% great, great advice. Hope this was helpful, guys. This was just some of the questions we pulled out of the membership. If you are interested in joining head over to BrandGrowthExperts.com. It's a big button, you click start here. You can go, it'll ask you some questions and see if you're a good fit for the membership. There are 125 people in there right now. Business owners, marketers, smart questions being asked all the time. If you feel like if you liked the podcast, you should join and we'd love to have you in there, so hope that was helpful. We love chatting with you guys and sharing some of the stuff and have a good December. Get some sleep. If you haven't gotten sleep through the, after the crazy weeks of Black Friday, Cyber Monday, just get some rest. You will feel better about it, about yourself, and everything.
Andrew Foxwell: Yeah, I agree. Just get some sleep. It's going to be okay. We're in this together.
Austin Brawner: We'll talk to you guys soon.
The “Perfect Popup” Template
Discover 7 Ways to Capture More Emails Without Giving Away Discounts