Savannah Sanchez, Michael Taylor
How to acquire high-quality customers for the lowest cost
[00:00:00] Ian Martins: Hello everyone and welcome to our Paid Growth session. Today, we have joined by Michael Taylor and Savannah Sanchez. I'll do some introductions here and then we'll get right into it. Michael Taylor founded Ladder, a 50 person growth marketing agency that worked with Booking.com, Monzo Bank and Time Out.
He's writing Marketing Memetics, a book applying concepts from evolutionary biology to creative strategy and testing. And then we got Savannah Sanchez here who's a TikTok Media Buying and Ad Creative Expert. She works directly with a handful of eCommerce brands to give them top-tier ad creatives and manages campaigns across various social channels.
So great folks to have in the room today to talk about Paid Growth. In terms of kicking things off, what I'd love to get is both of your perspectives, and maybe we'll start with Savannah on How you feel paid growth has perhaps changed over the past couple of years? I think there's been a lot of changes and a lot of things to dive into, but it's definitely a very different game than it was, let's say, back in, 2016, 17 when I was hands on keyboard, running campaigns myself.
So I'd love to talk to about, get both of your perspectives on how the games changed over the past year or two. And it'll start with you, Savannah.
[00:01:19] Savannah Sanchez: Hey, thanks for having me. Yeah, tons of changes. I'd say the first and biggest one being iOS 14. Changing the way that media buyers approach ad accounts, like you were saying back in 2016, 2017 and 2018, it was a lot more about manual media buying tactics like changing bidding strategies, testing a lot of audiences, trying to find EF that way.
Now, especially after iOS 14, it's all about how can we keep the accountant structure simple? How can we focus on creative testing? And brands are really trying to understand that the most important thing when seeing success on meta or TikTok ads is the quality of the creatives in getting like that testimonial.
Like authentic UGC style ad. That's really been a major shift in the industry over the last year, two, where that's become a larger priority than it used to be.
[00:02:16] Michael Taylor: Yeah, for sure. Like I've seen the same thing and yeah, good to be here as well. Yeah, so I remember when I started ad buying my special power was that I was like one of the few people in the room that knew how to use Excel.
And , that's not really a special power anymore. Unfortunately, , I had to learn how to do creative testing and actually pretty early on, I think Savannah you did a ton of content on the importance of creative testing that we read in our agency. And we just saw that was where the alpha was.
That's where over the past few years anyone who got quick learnings from the creative were ahead of the game and now I think a lot of the kind of optimization stuff that we used to be able to get an edge from just has gone away.
[00:03:07] Ian Martins: Then media buyers need to also be creative marketers as well.
So a bit of a shift there. Creative testing is a very interesting topic and there's a lot of different ways to test creative and a lot of different approaches to it. The audience here is mostly earlier stage startups. Love to get both of your perspectives on how to think about starting
your creative testing, right? Like how do you think about I'm gonna turn on TikTok ads or meta ads for the first time. What should I be thinking about to approach creative testing on these platforms?
[00:03:46] Savannah Sanchez: Sure I'll tackle that one first. So I do lots of video ads for new brands, and typically what I like to do because I'm not a risker at heart, and I think that's part be part of being a good media buyer is like taking calculated risks, but knowing like which risks to take.
And when it comes to creatives, especially for a new brand, I always like to go for the tried and true proven formats. Like it's not really the time to be I would say like super creative, innovative, but rather really study like what the larger brands are on Meta and TikTok are doing in terms of what hooks are they using, what formats are they using, what's the length videos style are they using?
Really become a student of what the top spenders on these platforms are doing and try to recreate that for your own brand. So there's no need to reinvent the wheel and really try to think something creative or out there. Of course, like as you keep testing and you wanna expand beyond your core videos.
It's great to do that when you first getting started. Really study what your competitors are doing and what the large brands in the space are doing and try to copy some of those formats so that you can take a calculated risk of, rely off of what's already been on the platform as opposed trying something completely brand.
[00:05:03] Michael Taylor: Yeah, I agree. Like I think you need to get to average first, right? You can't get to excellent creative straight off the bat. You don't even know how many thousands of creatives of like we tested and then you just can't predict which ones are actually gonna work half the time.
So I think best thing is to take a look at what other people are doing, what's working for them. If they're in your category, that's great because that's what the consumers expect. You wanna see what memes come up essentially. By memes, are they grabbing the attention in the first few seconds of the video?
Are they showing a product shot first or, these sorts of things that these patterns that you can spot that repeat throughout the different creatives that you've looked at. So build a swipe file up of from the ad library what have you seen, what is interesting, what's being used and then just try and emulate that as much as possible.
And then once you start to get at least average performance, then I think it's time for taking that leap above.
[00:06:12] Ian Martins: So sticking with the the earlier stage startups for a moment and getting into creative testing, creative production budgets tend to be a little lower when you're starting off.
And I think if you're..
[00:06:26] Michael Taylor: There usually isn't a budget, right?
[00:06:28] Ian Martins: Or there isn't a budget.
[00:06:29] Michael Taylor: It's usually just like me, like someone who's not creative, trying to mess around in Canva.
[00:06:33] Ian Martins: So what are some of the approaches you've taken to get to video ads? Cause I've heard you both say testing videos and where in the video does this happen or that happen?
And so video is obviously an important format. How do you start putting together video ads when your superpower used to be a spreadsheet?
[00:06:52] Savannah Sanchez: With this
[00:06:54] Michael Taylor: The mighty iPhone.
[00:06:57] Savannah Sanchez: Mighty iPhone is never easier to create great video ads with zero budget. If you have an iPhone, you can create great video ads. I would say that how I got started was making UGC brands like they would send me products, I would film at my house testimonial videos of me using it, showing it in good lighting.
Of course, if you don't wanna be on camera, maybe enlisting some friends or coworkers to start this process. You don't have to hire influencers or monitor off the bat, but just yourself and using the iPhone apps to edit videos. So one of my favorite apps that I still to this day is called Cap Cut. It's made by the makers of TikTok Bite Dance.
So as all the native TikTok, it's super easy to edit, and you can even get Cap Cut on your laptop now if you wanna edit it there. Filming on your iPhone plus Cap Cut is the easiest way to get started. Get quality ads that even the top brands are doing. They're making stuff on their phone and using these type of apps to get their video ads.
[00:08:00] Michael Taylor: Yeah we actually, when we were doing the ads for Monzo Bank, like this was at a time when they were growing really fast. They were just about to do TV advertising. They actually had a whole team of really clever talented designers. But that team was busy doing the TV campaign.
And we literally just rounded up a few people in the office that looked like they would be Monzo Bank users. And and they were actually, so that was helpful . And then and then we went to the local pizza joint, bought a pizza for them. And then we filmed loads of times.
Like one, one of the big features of the phone was of the app was that you could split the bill between all of your friends automatically. It would work out the differences of who owed what. So we just filmed them like splitting the bill after the pizza.
And and we actually got some really good creative that outperformed some of the TV stuff. And we didn't cost, hundreds of thousands of dollars to do that.
[00:09:03] Ian Martins: And let's go forward a bit of a later stage company. Maybe they've, it's Series A, Series B, a lot more reliance on paid media as a growth channel.
How do you approach scaling creative production? I remember one of my biggest challenges when I was a media buyer was just like not having enough access to volumes of creative and as brands grow, they tend to be a little bit more protective of the brand. Where you used to be able to just run around with your iPhone and make stuff with folks in the office.
Now all of a sudden they're a bit more risk averse to that sort of approach. Things got to get approvals and what have you. How have you both approached scaling creative as you've required it in scaling these channels.
[00:09:48] Savannah Sanchez: Sure. I think one thing though is that these bigger brands, I think especially over last year, are realizing like how big of a force TikTok is and that people like this type of short form, raw style video.
So more and more are they becoming accepting that this type of format that we need to have in media tool. So I would say for the larger brands, it's not just relying off the UGC TikTok style video, which we all know to perform really well. Also diversifying it with higher quality video production and studio shots.
Still videos founder or still images like high quality founder videos. So having a larger toolkit of the variety of the type of ads that they're doing. They're not just relying off TikTok style UGC, but I have noticed in the last year, brands are in the bigger ones that previously have very brand guidelines and want things to look very polished.
Are now realizing that if we want to reach the younger generation and be on TikTok and get on these new channels that we have to content in with what people are used to seeing on these channels. No one wants to see a quickly polished commercial on TikTok. They're gonna realize this is an ad and they're gonna swipe away instantly.
So you have to make it look raw and native and look like it's just a regular TikTok if you wanna succeed on that. That's something that I've seen with my clients last year.
[00:11:16] Michael Taylor: Yeah, like one of the things we found useful at the agency was building like what we call the creative lake. So basically a lot of preapproved idea.
And concepts and even specific executions of those concepts. And and then being able to like piece those things together. So like the biggest level of approval you need to get from the client, I think or your boss if you're in house is to just get them to approve concepts and not executions.
Because if you're approving every single execution you're gonna be like waiting for weeks to put new creative live, and legal will get involved. And, actually I had one client where it would take three months to get approval because they're in the financial services team and and every time we did approval, it would like the call with the lawyers would cost them like a couple of grand, at least
So the only way you can manage that and retain your sanity is to get them to approve, like not the final ad, like how will everything is cut together. But are you okay with this shot? Are you okay with this stock photo? Are you okay with this line of copy?
And you have to retain ownership over how you're gonna piece those things together. And if you can do that, then you can create as many variations as you like. And even with just like a small team, you can create thousands of variations and and that's way more than you'd even get to test unless you're spending a huge amount of money.
[00:12:50] Ian Martins: How do you so speaking about testing thousands of variations, how do you think about your testing strategy and getting to that point where you're testing thousands of variations and trying to quantify all of that? I'd love to, to dive into that, would things they progress and they get increasingly more and more complicated.
How do you start to think about that, especially considering you both alluded to the fact that, where paid media has gone now is the opportunity for improvement is gonna really rely heavily on the creative side. So as you continue to get more and more sophisticated. How do you think about your testing strategy in a world where you're testing thousands of iterations?
[00:13:33] Savannah Sanchez: I'm not someone that likes to test thousands of iterations. Even for clients that I'm working on, where they're spending over a million dollars on paid media a month. We're testing maybe two to four new concept week. So I know some companies will do the of iterations methodology, like the more we put out, the more chances of something's sticking.
But I always like to think about it in terms of what is feasible in the budget. Of course, like creating thousands of business, oh, a month is going to be expensive. So what we fit in the budget and if we can really strategize four grads a week and rely on past data in terms of what hooks are working best, what's working well for other clients, and really four concepts that we think are really strong.
I always, that's a better approach rather than trying to do thousands of iterations per month. Two to four is what I've found is a sweet spot for what I like to do with my clients in terms of new ads, just prioritizing quality over quantity, because if you do get that one really great ad, there's ads that we've been running for over a year on TikTok and Facebook that's still this performer or one of the top performers.
So if you can get those ads with a really long lifespan, In my opinion is worth pursuing more than I want a thousand different variables in the ad account. And then also the manpower, even if it's even possible to analyze a thousand different variables of these are slight changes we've made over all these hundreds of iterations, and this is
why it's important. I think just analyzing that becomes too much of a task and perhaps not even meaningful if you can't get significant spend on every variable. Like saying, Oh, this ad had a $1 CPA could only spend $10 doesn't really mean anything. I'd rather spend $5,000 on one variation and be able to say statistical significance.
Okay, I see this data name, this, click through rate the thumbs rate the CPA, rather than trying to have way too big a sample size for analyzing.
[00:15:38] Michael Taylor: Yeah, I think it's definitely moved more that way as well. Like we used to be able to get away with testing thousands of variations, even for like relatively low spend.
And we'd like with statistical significance, you can. It's a function of like how much data you have, but also how big the effect is, right? What we would do is we'd trash like lots of variations quite early. If it didn't move the needle in the..If it didn't even get a click for the first $10 and it's probably not gonna get like a $1 CPC, right?
And if it's not gonna get a $1 CPC, then the conversion rate would have to be enormous for us to hit our CIC target. So you can trash the variations pretty quickly and move aggressively through them. I usually only recommend doing a lot of variations if you have no idea what's working.
Or if you're like, really far away from your target, like if you're like if this is like your last month with the client before they fire you because you, you're lighting numbers or actually with us, like we had over 200 clients and a lot over the six years and so many of those clients had never done paid ads before and like they weren't even sure if it would work for their business.
And so a lot of what we do is try and quickly figure out if it was worth pursuing. I think you can make paid ads work for any business eventually, but there is an opportunity cost, right? Like it's a distraction if like you could grow better through SEO or through something else, right?
So that's only scenario where I think you should be that aggressive. I think and even when you are testing lots of variations you're not really like AB testing them necessarily. What we tend to do is like only really AB test the concept level. If it's like a really big change, so going back to the Monzo Bank example, like a lot of the users wanted them cause they had like really cheap foreign exchange within the app.
And then like commission free. And then a lot of users wanted to use them for their bill splitting function, which I talked about a minute ago. And then there were like two or three other concepts and we just weren't sure what was most important and they were important to different groups of people in some cases, right?
So that's when, like a AB test makes sense. And you then you want to like, run the test for two weeks to make sure that there's not seasonality and like you wanna figure out like the statistical significance and actually make sure there's a big difference between them.
And, but then once you have the concept, then when we're adding more variations, it's usually just we're just like putting them in the same adset and letting Facebook decide specifically. A lot of people don't like that approach because then like the problem that you run into is Facebook decides too soon and they don't give the new variations a chance.
But if those variations are like relatively cheap to produce and they're all like, very similar to the main concept that you know is working then a cost of additional variation production is very cheap. So you're only making small differences and we found that really extends the the life of a good concept in some cases.
They might get tired of seeing exactly the same ad again and again, and maybe there's some algorithmic bias on Facebook's behalf of showing the same ad for a long time. But if you're making small variations, then Facebook doesn't necessarily know that's like basically the same ad, right?
And like people will see more differences and in their heads it'll be like slightly different, even though the punchline is very similar. So I think that's this is like you really don't need to worry about this at all unless you're spending like at least hundreds of thousands a month.
You could basically just one or two new ads live a week and you'd be fine from most accounts, like most advertising accounts don't spend that much. So this is only really when you're like scaling aggressively. And, we were doubling budgets every two weeks at the time with Monzo.
So it's like then we had to go really aggressive, but that's like war time, that's not normal. Peace time.
[00:20:10] Ian Martins: I've been noticing a few questions in the Q&A section around platforms, right? What platforms best for this? When should I think about going all in on one platform versus diversifying?
Would love to get perhaps your thoughts on how do you think about different platforms, perhaps for different use cases. So maybe let's start off with something TikTok, right? It's an emerging platform. I think it's for even for a lot of experienced performance media buyers.
It's still fairly new in terms of what does performance look like in TikTok? We'd love to get your two both of your perspectives on like how to TikTok approach as a platform effectively when to use it. Who should use it? Would love to selfishly pick your brains on that.
[00:20:59] Michael Taylor: Actually Savannah, I wanted to ask on that are you all in on TikTok now?
Cause I know you were really early on Snapchat ads and and I think maybe you start with Facebook, Instagram ads as well. Have you like focused purely on TikTok cause it's working so well? Or is it you do still do a little bit of everything?
[00:21:19] Savannah Sanchez: I would say Snapchat. I used to love. iOS 14 really killed it, at least for my clients.
Like we really stopped seeing efficiencies on that platform after. And that's just how this is with paid social like you have like always changing, the algorithms are changing. People's behavior is changing. Like TikTok straights just spiking over the last two where you have to adapt. I would say my clients, their channel distribution has changed so much over the last couple years where it used to be more heavily on Snapchat interest.
Now they're maybe 50% on Meta, 50% on TikTok. I still don't have any clients that are primarily on TikTok says most people, the split is either 50 50, Facebook, Instagram versus TikTok, or there's even more a percent in Facebook. So I think it's still important to focus on Meta ads, and that's still like the largest distribution channel.
For TikTok, I would say if your product is under a hundred dollars, then that's really the sweet spot for TikTok. From my clients, we've seen that it's very much like impulse buys. They TikTokers sales, promotions like accessories, beauty, home. I think under a hundred dollars, it's really the sweet spot and it's definitely not just for teenagers like we're seeing.
Really good TikTok results for like women 50 plus on home goods and and different stuff like that. So it has a really large demographic on TikTok now. I think price is really the most important thing when it comes to stock because they want impulse buys and they're really price sense. That's what I've been noticing over the past particular.
[00:23:00] Ian Martins: No six figure SaaS company is then probably not the best.
[00:23:03] Michael Taylor: Not yet. It'll happen though. I mean like people used to say the same. I remember I was like doing like Google ads. That was my first job was doing Google ads for clients and Facebook came out and the number one thing every client said was, Oh, my customers aren't on Facebook.
And eventually that became not true. But there was an arbitrage opportunity for a while where they actually were on Facebook, but nobody was targeting them. And yeah, we even had some success with a really large B2B client. Their target customer had were in the procurement department of like a corporation that had at least 5,000 employees that was ICP.
And you wouldn't expect necessarily to be able to target them on Facebook and definitely the quality, the average percentage of people who are qualified leads was much lower than LinkedIn. We're mostly advertising but like the cost was much cheaper than LinkedIn.
On average, like if you throw away those low quality leads then the cost per quality lead was great. So I think this probably it feels similar now to me, like I'm on Tiktok and I could buy a SaaS software, right? There's tons of people on TikTok that don't necessarily even tell you that they're, that they like TikTok, right?
Because there's maybe a little bit embarrassing. Especially with I think initially it was all like people dancing, right? And and it was just like, not really like the vibe that me is like a, a 36 year old dad like, wants to give off, right? But now I think I know there's Excel tutorials on TikTok, right?
There's jokes and memes that like appeal to everyone, right? There's like news on TikTok like I've seen a bunch of breaking news I've seen like on TikTok first, even before Twitter, which is for the past five years it's been Twitter first.
Yeah, I think it's, I think it's changing pretty quickly and there's probably already an arbitrage opportunity. .
[00:25:16] Ian Martins: So when do you think it's a good time for advertisers to start diversifying channels? Let's say you're on Facebook and Instagram and that ecosystem's working quite well for you.
Like what would be the appropriate signal or time for you to diversify and start trying other channels versus trying to just do more where you are right now?
[00:25:42] Michael Taylor: Yeah I could take that. It's I think it's about like how desperation in a way, . If you're like, if you're a new brand and you can't really afford if you try Facebook ads, and it's not, it's not a Meta ads, I should say. I keep you calling it Facebook, if you try it and it's working okay, but it's not gonna get you to series A and you need something like, you need some edge, you need some leverage.
There is like at least an opportunity to try a new channel and dominate that channel before it gets saturated. Because once everyone figures out the channel, then you know it gets expensive, right? And it's very hard. Unless you have something that's very unique about your product that matches that channel.
It's like very hard to compete. Or if you know you have a very good media buyer, right? or a creative team. But yeah, so I think there's an opportunity when like you're a small player and like you can't quite make the economics work on the main kind of saturated channels.
And if you're in that situation, you also don't need that much scale, right? If you get a hundred new customers this week, that's gonna be great, whereas if you're a Unilever, right? Like you need millions of customers before you even, get out of bed. So I think that's, you saw, I think the other situation is like when you're.
You do have a channel that's working really well, but it's starting to mature and you're reaching the ceiling like you're reaching diminishing returns with that channel. And then you're looking for your act too. And you're not gonna turn that channel off. It's still gonna be probably 80% of your budget.
But if you're not like, at least experimenting with other channels with 10, 15% 20% of your budget then you might run the risk of hitting that ceiling and like then starting to miss your targets.
[00:27:39] Savannah Sanchez: I agree with everything there. I would say, I don't know a brand that's not using TikTok right now, least on a small scale. Everyone's curious can we get additional efficiencies here? Have we maxed out on who we can run on Facebook and we need something new? So everyone's sent it to start. And like I said before, like if you have a product under a hundred dollars, I would say you should only be testing it, especially along with Facebook.
Facebook's gonna be a lot more expensive. Sure. More competition. So if you can figure out TikTok while it's still really cheap and your audience is there, I think that's the strategy. I was at least spending 50 50 on each channel if I was just getting started.
[00:28:20] Michael Taylor: Usually though, what you find is, if you I've seen this across almost every single brand is like the 80-20 rule, like the Pareto rule.
Like almost always you have one channel that dominates all of your, it drives almost all of your growth. And then you have one other channel that drives, 15%. And then the rest of the channels make up the other five. So it's very hard to escape that rule.
It's possible, right? Like I've seen people do it. But usually despite your best efforts most brands are reliant on just one channel.
[00:28:58] Ian Martins: How do you think about, and maybe I'll direct this one for Michael at first, How do you think about attribution right now? And I guess to be more specific, cause obviously it's a big topic, how do you think about like awareness,
right top of funnel investments versus kind of demand capture, more bottom funnel. And how do you think about attribution between the two and should I invest in awareness? A lot of our audiences earlier stage startups, obviously they want to make sales. They only want to spend money where it's gonna have an ROI positive return for them.
But, if nobody's heard of you, that might be difficult. So how do you think about navigating that and thinking through attribution?
[00:29:37] Michael Taylor: Yeah, so this is like my favorite topic, because since I was 14 everyone it's wide open, right? Nobody knows what they're doing.
We're all trying to figure it out, and it's, there's a suddenly a ton of room for creativity in how you do this. And there's a lot of advantage to be had, I think if you do figure it out. And nobody has, right? Nobody's figured out attribution yet. And if they tell you they have, then they're probably lying.
So I think roughly, the way you should be thinking about it. One of the questions you asked was like, brand, awareness versus performance. And I think that's, if you talk to a traditional marketer, like a traditionally educated, like qualified marketer, which I am not they would tell you read the long and the short of it like Les Binet and Peter Field it's this seminal paper that everyone quotes that nobody's read and and that tells you like you should spend
roughly I think it's 60% on on branding and then 40% on performance. But in reality, I like the bottoms up approach. I think go for the low hanging fruit first. If someone is searching for your product, You probably want to do Google ads first cause that's the very lowest, bottom of the funnel thing.
Those people definitely want the type of thing that you sell. So you know, if you can at least get that running, like you maybe usually can't beat the rest of the competition there. But like those are gonna be the most qualified customers and they're gonna be also ready to buy, like right now.
So that's gonna give you a lot of immediate feedback, which is useful. So that's usually the first thing. If nobody's searching for your product cause you sell like I don't know, like hips and neck ties or whatever, and nobody searches for that. Then you do need to go a little bit more to your audience and interrupt them essentially.
You can't wait for them to search for you. So that's when you look at the midterm kind of social networks that's when you're looking at, Meta ads and and you're looking at TikTok and stuff, like that's the type of product where you, if you see it then you do wanna buy it or if you see it a few times, eventually you'll buy it, right?
And that's midterm. And then I think that is interesting cause there is like a branding component to it. Not all of the impact is gonna be felt like the day that you run the ad. And if you show the ads enough, like to the same audience, eventually you'll start to get more and more direct traffic is how you'll see it come through.
And that is attributable back in some form to those ads. And that's where you get into things like marketing mix modeling which is, the new craze, the savior of iOS 14, but also incredibly difficult to do. And yeah. And then you're looking at also like long term brands.
Performance channels tend not to, they do have some brand effect, but not a huge amount. Most of their impact is felt in the short or medium term. But then if you are growing, you're becoming a bigger brand, and then you do start to do more brand type things, right? You might do a PRS done or you might do a TV ad or you might sponsor billboards or like for me, like branding, like for my consultancy, like branding is like writing blog posts because they very rarely lead directly to leads.
But every now and again, I get someone who's Oh, I've been reading your book for four months. And now I finally have the budget to reach out and, so I think you can invest in brand awareness early on. It's very difficult to attribute the performance of brand though.
So I would say like exhaust, the low hanging fruit first and then the medium hanging fruit , and then the stuff at the top of the tree.
[00:33:30] Savannah Sanchez: Love that.
[00:33:30] Ian Martins: Anything to add there. Savannah. Yeah.
[00:33:33] Michael Taylor: Don't if medium hanging fruits a real you get there.
[00:33:36] Ian Martins: It is now you can coin it. There you go. Cool.
Do you think, just one last one on the topic of attribution, cause it's a topic that at least it came up all the time with me, most particularly when I was focusing on page.
When do you think it really matters? and like how much attention and energy to spend on it versus it's starting to actually get in the way of you doing things. Because I think that there was was particularly just working with startups. There's this idea that you can actually get a true picture of everything and they start off with that starting point and.
and you can't and you might not do things you should be doing because you can't measure it, and you might be getting in your own way because of the topic of attribution. So how do you think about that? How do you have that conversation with clients and I imagine you, you've both experienced the same thing, so we'd love to get your take on like how much attribution is enough versus too much and spending too much time and energy on it.
[00:34:36] Michael Taylor: I was gonna, Yeah, that was interesting seeing how you do it.
[00:34:41] Savannah Sanchez: Oh, sure. I like to take a very basic approach to attribution. One thing that I like to look at is the market efficiency ratio, which is essentially your entire sitewide revenue. Regardless of what channel from divided by your total ad spend across channels.
Typically brands will say Okay, I wanna make sure that my revenue is at least four times my ad spend on any given month. And in that ratio we are profitable. Not necessarily they're trying to pinpoint exactly what channel it came from, because like we said, there's so many marketing activity you do that you can't attribute like to like this Instagram story.
You did this post, you did this blog, you wrote this email. But ultimately all those efforts combined, you wanna make sure in any given month that my total ad spend was only a fourth of my online revenue so that you're profitable. So that's really the North bar for most brands is not necessarily caring
exactly where I came from, one thing that's helped Claire, that is post purchase service. That's something I'm a huge believer in. There's a great shop app called Fairing, formally inquire. I'm not concerned. I just love them. And it's literally like something I look at for every single client. It's just end of when someone purchases something, they're asked, How did you hear about us?
And they can type or they collect Facebook, Instagram, TikTok, email, influencer, and we find that 50% of people will fill out the survey, like of all customers. So you get a really large sample size and then you can see, okay, 20% of my customers this month that they found me through TikTok. Cool. And what I've found is, especially for TikTok, where the attribution and pixel tracking is not exact, to say the least, I would say it captures maybe half of all purchases that actually come through.
We can spend on TikTok because we're validated from these post purchase per survey results that actually we have this much revenue and this percentage of people saying that they found us through TikTok. Whereas on TikTok Ads Manager, our CPC is way higher and we have way less customers than we thought we did.
So the post purchase survey often validates that this channel better than we thought. And if we did that, then we'd probably spend a lot less on TikTok because if we're just relying off what the platforms say probably just shut it off cause it's gonna look terrible that's how I like to look at it.
[00:37:07] Michael Taylor: Yeah, I think that's smart. Yeah I would say I'm like a attribution junkie. Like I've done like really complex stuff and learned how to build like Market Mix models and other stuff. But actually, when I do attribution, I keep it very simple , I think some clients obviously need really complicated stuff, but by and large, I think, if you've
some kind of directional steering from your tracking pixel. You understand in real time like this is getting more traction than that, right? I put this ad live and it's really bombing. And you don't need to like, wait for the marketing mix model or even look at the survey results half the time, right?
You can see it pretty quickly. But then I think surveys are definitely underrated. Actually. I know Fairing as well. I did a guest post on the blog recently. And and it's just a great tool. Like it's, there are obviously other survey providers, right?
I'm not sponsored either, but I think in general, if you don't have a survey that shows up after someone buys something or after someone signs up on your website then you're really missing a trick because something like most of the websites I look at like 60%, of the traffic is direct.
Especially if you're spending a lot on video ads. And those video ads are gonna, people not necessarily gonna click on them, right? Like they're gonna watch them and see them like a few times and then, What was that brand again? Okay. Let me just search the name. And they're gonna come through either through branded SEO, yeah
or they're gonna come through direct. Or maybe, friends will share links in WhatsApp, right? And that comes through direct as well. So I think there's this whole bucket of traffic and it's your biggest source of traffic that you don't know anything about unless you're doing surveys.
One thing I would say is that tends to bias towards more memorable channels. So if you are doing a lot of video, it will give credit to video because that's what people remember. And sometimes it will like bias against things like Google Ads, which are a lot less memorable and people like forget that they clicked on the Google ad.
But you, that's fine because all the other attribution methods are usually biased against video ads, right? So like Google gets too much credit when you're just looking at UTMs. So yeah, I think the key after iOS 14 is really to do multiple methods and you don't have to make it very complicated.
It's just a case of looking at the tracking data for kind of a directional steer to optimize, and then looking at survey data maybe doing some linear aggression, like marketing mix modeling stuff once you're spending more and maybe doing an experiment like a, switch it off for a few weeks, switch back on again, see what the differences and that's probably like good enough for 90% of brands.
[00:40:09] Ian Martins: Okay. We're coming up to time, so I got a couple more questions for you all. And then we'll start to wrap it up, But we're in Q4 now. If you buy media, this is a particularly expensive time to be buying media. What tips do you have for Q4 2022 for our audience?
[00:40:30] Savannah Sanchez: TikTok.
No, I'm just kidding. But , No I would say of course leaning into sale messaging, gifting messaging, we're start with that now. Like already making ads saying This is the perfect gift for your mom, for your best friend. Here's gift ideas under $30. I would advertise now because right now is cheaper.
Worse than two weeks leading up to Black Friday. Every week that goes by, it's gonna get more and more expensive. So start getting that brand awareness now and get people thinking about your product as a gift. Don't wait until a week before Black Friday to let people know that you're having a promotion or sale or that this, like start planting that today.
[00:41:11] Michael Taylor: Yeah. Do you know what like this is the advice? It sounds obvious, right? But so many clients, I saw this if you're not involved in Black Friday, like if you aren't in a DTC space, turn your ads off. You just don't wanna be in the market right now, or at least dial them down.
Because I see so many brands and they're like No, we budgeted like a million, $1.2 million for the year. So we're gonna spend a hundred grand a month, regardless of whether it's like Black Friday or like February when nobody is advertising, right? I would say if it's be aware of what else is going on in the market and know if you're a travel company like actually it's January and February the big times.
And if you're not a travel company, then maybe January would be expensive to advertise as well. So yeah, I would say think about it and usually there's like diminishing returns to this, right? Yes, it's more expensive during Black Friday, but that depends on your volume, right?
So if you need to deploy like $12 million this month. Then you're going to have to win every auction, right? You're gonna have to try and win as much inventory as possible, and therefore your costs are gonna be sky high. But if you've got realistic targets and you don't need to like, you don't need to deploy the spend and you've just got a sensible kind of a profit margin that to deal with, you can just dial down the volume, right?
So yes, you won't sell as much during Black Friday, but if that's profitable sales, that's better for the brand most of the time than just emptying your warehouse of inventory and selling all that stuff at a discount as well. So you're probably losing a ton of money.
So it depends on the brand and not every brand can be that flexible, but if it's like your business. Then be wiser about how you do it or, if you're cozy with the client, you can explain this to them and help them navigate this period rather than just like default thinking, Hey, we've got to hit this target.
We have to basically ignore all of these other things that are going on.
[00:43:22] Ian Martins: Perfect. Thank you. Everybody loves tools. What are your essential paid growth tools that you use on a regular basis or you would recommend for people to check out? I know we had one in the Shopify ecosystem that we touched on that we can't recall the name of that which Savannah?
[00:43:41] Savannah Sanchez: Fairing.
[00:43:41] Ian Martins: That's it. Yeah. What other tools do are essential to you and your work and paid growth?
[00:43:48] Savannah Sanchez: Should have got some sponsors for this. I know. Just kidding.
[00:43:51] Michael Taylor: I'm already sponsor . I will just wear any flag that people send me, say, just send me stuff .
[00:43:58] Savannah Sanchez: I keep it pretty simple.
I, like I said, I like the post purchase survey for Fairing. Most my clients will use some sort of third party attribution tracking platform like Triple Whale, which is great for reporting. Same with Northbeam, that's another popular one. But honestly what I do is most of video ads side and creating the UGC.
So most of my tool like CapCut and like doctor effects and Dropbox like my, and then Trello of a project management. So cause it's more on the video side, it's not necessarily the tactical Shopify apps, but I'm sure mine has some good ones to recommend.
[00:44:38] Michael Taylor: Yeah. Do you know what like I'm on the other end of the spectrum in that half the stuff that I wanna mess around with they don't exist as tools, so we have to hack them together ourselves, with like spreadsheets and API calls and stuff.
So yeah, like usually, we mix that stuff together ourselves. Like we actually had a team of four developers at LA that and a designer that, that basically just built stuff for us based on what we needed. But I, again, I think that's completely not needed unless you're doing a lot of stuff at scale.
And it is actually, basically for simpler, you keep it the less you have to worry about tools. Right? And if you are doing really complex campaigns, then you do need to some automation. Savanna was saying, like if you are testing thousands of ad creatives, nobody wants to do that work manually, right?
So that's when you have to start getting into I don't know, machine learning to identify what's in the ads. So you can tell. I don't know, these types of ads working or those types of ads that's the sort of stuff that you can do, which is becoming a lot more accessible now. Like we one of the things we did actually we moved away.
One thing I would say is don't build anything for yourself that already exists because it's gonna be like a nightmare to maintain, very expensive and that's not the business you're in, right? Unless this is you really cool to your business, you shouldn't build it in house.
So like ,even though we built a lot of stuff like we move, so these guys are the kitchen. I, again, not affiliated, but what they did was. They're like Zapier, but for ad management and stuff. So you can say like, when this, when you know when this ad is not performing, then turn it off, or send me a Slack alert if we're not hitting our target, that sort of stuff. So we actually scaled back a lot of our our tech and then just put it into that platform because then it's then they have to worry about maintaining it. So yeah, we actually, we went through this phase of building loads of stuff and then actually studying, scaling it back and just using tools that are out there instead.
[00:46:59] Ian Martins: Cool. And then as a final question for you both a lot of the folks in our audience, again, they are at startups and earlier stage startups, when's the right time to think about paid ads? Like at what point in your life cycle as a business should you be thinking about going into paid as a growth channel for you?
Is there anyone that maybe shouldn't be looking at paid growth at all and should be doing other things in your opinion, I'd love to get your POVs on that.
[00:47:33] Savannah Sanchez: Sure. I would say as one, as someone in the chat pointed out like product market fit is of course the most important thing and utilizing paid to prove that or disprove that is probably worthwhile at the start.
Sometimes brands will hit me up for ads and I'll go to the website and the website is just terrible. Just bad fonts. It's slow. So poorly, it looks terrible. I can't, and I'll be on the website for a few minutes. I'm like, I don't even know what they're selling. And they're asking me like, help me with TikTok ads.
I'm like, No, I need to go fix this, and this before you even start. So I would say being honest yourself and thinking am I ready for this? Does my website and my offer, my product show to the world? I think that's number one. And that's what I look for when I'm taking on clients that are brand new.
I'm like, Okay, have you figured out the basics? Do you look legit? Do I know what you're selling? Does your website function? Cause I'm surprised that the amount of people who come to me wanting ads and I can't even get their website to load or figure what they're selling.
[00:48:41] Michael Taylor: Yeah, sure. Like I'm a little bit more rogue with that stuff.
I actually, we did a bunch of stuff where we ran ads for products that didn't exist just to see if they should exist. So I think that is you the good thing about ads is like, it's not like buying tv, right? Like you don't have to spend $5 million for a Super Bowl placement on day one.
You can put $10 a day on your credit card for a few weeks and to just see generally are people clicking on this? Which one of the two variations I was thinking about are working? So I think it's valid as like a way to quickly learn as an actual acquisition channel.
I would say usually it's just one of the four major things. And there's a lot of businesses that, never, ever get like paid advertising working at scale. And like actually in every category there's usually like one client that only one business that like makes it in each channel, right?
So travel, you have booking.com where like world class at paid ads and I think Google's biggest customer right? And then everything you have Airbnb who turned off for their paid ads, because they said that they didn't see any drop in revenue, right? and Is it because booking.com, a bad at attribution?
No. Is it because Airbnb, like it's not because paid ads doesn't work, it's just because Airbnb found like their business is much better at virality and affiliates and referrals and like brand PR type stuff. And then you have TripAdvisor, right?
Which or Expedia who dominate the the organic search side of things, right? So I think usually, like in every category, there's one brand that does really well on SEO. One brand that does really well on paid ads. One brand that does really well on virality like kind of social stuff organic social and then one that does like really well on sales or like direct
partnerships and sponsorships. And so if like in your category, if there isn't someone in the paid ads bucket then you maybe there's an opportunity, right? But then you have to think about like, how does my, how can I change my business model to make paid ads work for me? Like booking.com have
thousands of people who they like hire like rocket scientists, right? To do all these AB tests and and like change, like everything about the website as well to make it convert on a short term basis,and they've made very different design choices based on that than say Airbnb, where it's more about unique properties and beautiful photos and things like that.
So I would say yeah usually your business model decides whether paid ads will work for you.
[00:51:36] Ian Martins: Great. Savannah, where can folks find you online, get at you if they need some help on TikTok or the socials?
[00:51:44] Savannah Sanchez: Of course, you can find me on my website, which is the socialsavannah.com.
My website also has my TikTok ads course. If you wanna learn more about media buying on TikTok and creating great ads for TikTok, my course is the ultimate resource, so that's on thesocialsavannah.com. I'm most active on twitter which is social_savannah. And then you could also find me on Instagram at thesocialsavannah.
[00:52:10] Ian Martins: And Michael, where can folks find out more about you and your book Marketing Memetics and, nerd out on attribution?
[00:52:18] Michael Taylor: Yeah. The book is Marketing Memetics. Yeah, it's marketingmemetics.com. Vexpower is what I spend most of my time on vexpower.com.
That's a training platform where you can actually learn a lot of this attribution stuff. And I make up fake companies for you to practice on and give you the data. So like that's definitely worth checking out. And then if you just wanna see me post random crap all day then go on Twitter, I'm a hammer_mt
[00:52:54] Ian Martins: Great. Thank you everyone for your time today. I hope you found this to be valuable.