Listen on Apple Podcasts

AO: What was one unconventional growth tactic that you ran that did support surprisingly well?

Larry Kim: Hey Adam, thanks for having me today. Today I want to talk a little bit about Facebook ad costs and how to drive them down, using some weird tricks. Let’s face it, Facebook isn’t cheap, audiences used to cost a dollar CPMs like five years ago or closer to a hundred, $150 these days. So if you’re doing Facebook ads kind of the normal way, I think that it’s going to be very expensive and probably you’re losing a lot of money on that. So I’d love to share with you some weird growth hacks, ideas to kind of get more out of your ads by just employing clever ideas.

AO: We get the problem, we hear it all the time. Could you tell us some of the impacts? How would you quickly summarize a strategy that you have, that’s going to help us get around this?  

Larry Kim: Let’s see here, in Facebook ads, I think there’s sort of this playbook where you sort of target the people who you think the ad is relevant to. But I wanted to explore a different angle, which is targeting people that are who have no relationship, where you’re targeting interests that don’t necessarily correlate to your target buyer or it doesn’t really, really make sense, but let me explain you a little bit. So you’re going to have to use your imagination a little bit. A friend of mine showed me this ad and the ad was to buy a t-shirt. Okay. And the t-shirt said, “Life took me to England, but I’ll always be an Arizona girl at heart”.

So this girl, friend of mine, she’s living in London and she was born and raised in Arizona, but that was like 20 years ago. OK. And somehow, some clever advertiser is targeting her in London with this ad that says “life took me to England, but I’ll always be an Arizona girl at heart”. And she’s like, WTF is my phone listening to me? How do they know this? And so that’s an example of what I call inverted unicorn ad targeting. Don’t worry if you’ve never heard of this, it’s just a made-up word that I’ve termed.  

But in a nutshell, it’s a recipe that you can copy to create really, really crazy ads that you know, where people are like do a double-take and they’re like, how do they know about that? And it’s actually really simple, Adam. If you break it down. The first thing that ad is doing is it’s targeting people who live in the UK, who are women because it has to have like the image, it had a woman’s shirt on. There’s a demographic that you can overlay. It’s called recently moved. So women who live in the UK who’ve recently moved, but the question is, how do they know that my friend was originally from Arizona? Cause that’s kind of the punchline there, like “life took me to England, but I’ll be an Arizona girl at heart”. How do they know that? Basically, it’s very clever, they’re targeting unrelated interests.

So what I mean by this is they’re targeting people who live in England, who recently moved and who like a local business or club or organization that’s located in Arizona. So it could be a high school, it could be a sporting team. It could be a bar but like the combination of those three kinds. If your audience meets those three criteria, it’s not a bad bet to guess that that person originated from Arizona. Why? Because why else would they like some random school or bar or sporting team, like way out, so far away from the UK. Do you see what I’m saying, Adam?  

So this is kind of goes against conventional advertising theory. So conventional advertising theory says, well, you’re supposed to go after related audiences. There’s a suggestion bar in Facebook ads where it says “type your interest”. And you could type in people who like marketing. And then it’ll say suggest more interests. Like people who do email marketing or people who like HubSpot or people who like SEO, you see what I’m saying? But these are all related interests. Like they’re kind of a subset of each other. Or, at least they’re sort of concentric circles. In terms of the Venn diagram, what I’m suggesting is that there’s a new way, there’s a possibility of coming up with ad targeting where you target unrelated interests. There is no correlation between people who recently moved and live in London, but who originated from Arizona.

They like a certain high school in Arizona, there’s no higher propensity for that person to be from Arizona than it is from Massachusetts than it is for Canada. Do you follow me here?  

So what this targeting method does it does two things. The first thing it’s doing is making an educated guess by thinking if these two things, these three things apply I think I can spin a narrative of your life. The second thing that it’s doing is it’s making something, it’s creating goosebumps. Like people are commenting on this ad. They’re saying, “WTF,  is my phone listening to me?” How many Arizonians are actually live in the UK? Why am I seeing this? It’s just, it’s just crazy.

So when you make these guesses about someone’s demographics or interests or behaviors, based on what a plausible scenario targeting certain areas and kind of extrapolating some kind of a hypothesis about their identity and then that you’re usually right. Those tend to get very, very high engagement rates, like people commenting, liking, sharing, clicking, and then, of course, the way the Facebook algorithm works is higher engagement rates that means more impression share. That means more the ad server will be more likely to serve it at a lower, effective cost per engagement because you’re generating so much engagement from the impressions that you’re you’re accruing.

AO: One quick question that I have is just the size. It sounds like in this example, the limit, it’s like a super limited size. Is that been a problem for you now?

Larry Kim: So what you would do is you would roll this out, in a thousand campaigns, as “Life took me to Canada, but I’ll always be an Oregon girl at heart” or “an Oregon guy at heart”. It’s just an exercise in cloning, a gazillion campaigns with ultra-high engagement rates. Do you see what I’m saying here? 

AO: I did it, I mean, this is it’s brilliant and it makes so much sense. Could you help us with even more examples potentially around B2B buyers? Because I know I can think of examples in my head, but I’m curious if you can help us, like sure, sure, sure, sure, sure.

Larry Kim: Look, I use this in my marketing all the time. My company MobileMonkey, we’re a Facebook, we’re the world’s leading Facebook messenger marketing platform as well as SMS and SMS and webchat. But like a lot of B2B companies, we’re trying to get our name out there. And so we do a lot of content marketing like just blogging. And one of the things I wrote like a year and a half, two years ago. This was during the election, just after the 2016 presidential election in the US. It was just an article on fake news and how easy it is to spread fake news on Facebook. I thought that was a relevant topic because we do Facebook, our core audience is Facebook marketers. I just thought that was an interesting topic to write about.

And so we just wrote about it a story, how I just create a fake website called, “Citizens news” or something like this. And then I created a fake news article on that fake website. And then I created a fake Facebook page and posted the fake articles to my Facebook page. And then I just ran an experiment and showed how Facebook makes it very simple to like, boost posts, and then generate an insane amount of clicks and views and comments and likes and all this stuff. So that was sort of research. And then I wrote it up as a case study. So the blog post was written, “Here’s how you create fake news on Facebook and in four easy steps in 10 minutes or less”.

So that was the article that I wanted to create as kind of a PR stunt for link building and trying to get attention to your business. You see everything. So this is kind of a B2B brand building kind of thing. And so now I have this really great article that kind of goes through the step by step approach to spreading fake news. And I think that it’s going to be interesting to a wider audience of people. Now, I just need to promote the article. I need people to read the article so that they start sharing it and stuff like this. And the problem is I don’t want to spend a lot of money on this stuff, I’m cheap. So I limited the promotion budget to 50 bucks.

But we generate tens of hundreds of thousands of views for the article. So the question is, how did I do it? And the answer is I did it with Facebook ads, and I did it using the same inverted unicorn ad targeting method that I just described to you a minute ago.  

Adam, I had a hunch that this article would be more interesting to liberals. Because they were upset about the election, rightly so. And like, they thought they were going to win it and this unexpected thing happened. And naturally, people want to point the finger at somebody “Oh, see, it’s Facebook’s fault”. You know, like this is all because of Facebook’s and the Russians, and they’re doing all this nasty stuff on Facebook, and that’s why this happened, you know?

Like they want to sort of explain things. Does that make sense? 

All right, I’m going to target this ad for my blog story on my blog. I’m going to target it to liberals. All right. Like people who donate liberal causes, people who vote democratic, these are all targeting options in Facebook ads. The problem with that approach was that there were 55 million of these liberals. It tells you in the audience estimator how many people you’re targeting here. All right. And I was like, well, wait a minute, like 55 million people in this audience. And I only want to spend 50 bucks on this, on this PR stunt.

So that’s kind of what I describe as a sort of like your audience size is not commensurate with the amount of money you have to spend. If you’re even close, if you’re only spending 50 bucks, then you have no business targeting such an enormous audience. You’d be better off whittling it down from 50 million to 50,000 or 5,000, even. So that’s where the inverted unicorn comes into play. So I need some way to slice and dice these 50 million people. And conventional advertising says, well, why don’t you get the really, really, really like liberal people, like people who, what’s a liberal publication? Like people who also like MSNBC or something. 

You can find even more liberal people who are more likely to eat this stuff up. And I was like, no, let’s try a different strategy. Let’s try an inverted unicorn strategy where we target something that’s completely unrelated. So in this case, and this is kind of ridiculous. I targeted people who like a television series called “Star Trek. Deep Space Nine”. It’s like a very modestly popular show, like 20 years ago in the nineties. The Star Trek genre.  And there was this one episode where there was some kind of a Romulan Senator who, his name was Senator Vreenak.

And the whole point of that art or that episode was basically like fake news was, about fake news kind of using, how fake news was used in the 24th century to get these, the species into a war. So this is like a real niche, who the heck is going to know? Only a very, very small number of people. This is so niched. So I went after liberals, who also liked the steepest deep space, nine fan page, and it cuts it down to the audience to like 100,000 people. And basically in the ad copy, in the image, I used an image of that Romulans Senator Vreenak where he’s saying “this is fake”.  

It’s like a meme. But it’s a very, very obscure, it’s an obscure meme. I don’t expect anyone here listening today to understand that need, because it’s just such an obscure show, but that’s okay. Because I only have $50 to spend. And I don’t need all 50 million of my audience to recognize this. If it’s even just five or 10,000 or a hundred thousand, I’m fine with that. You see what I’m saying? Cause I’m only spent 50 bucks. So I combined this kind of imagery of an obscure, Romulan Senator saying “this is fake”. And then using that as the hero image for this article on fake news in, on Facebook for the 2016 election.

And that was my experiment, going after unrelated interests rather than correlated interests. That’s the inverter unicorn ad targeting method. And basically the thing went bananas. So within minutes of posting this thing, it generated, hundreds of thousands of impressions and tens of thousands of clicks to the website. And you can even see like the Facebook posts, it’s, there’s thousands of comments and likes and shares and I know that it worked because people usually don’t drop comments on ads. But there was hundreds of comments in the ad saying like, “ha ha ha. I love that episode”. Someone even incredulously, wrote a comment to the ad saying, “I don’t get it, how so many people in the world even know what this obscure reference means?”.  

But yeah, he’s opining in the comments saying ”how is this possible? Like, the person is so happy to be on the inside of that inside joke. But what he didn’t realize was the only reason why he’s seeing this ad is because I have had a hunch that based on the fact that you liked such an obscure fan page, that you, you would be in on this joke. 

AO: So good. And the fact that we know that he’s potentially liberal in that, in the way that Facebook categorizes it. So you put those together and it’s just like an explosion of, like, I can’t believe that someone gets it.  

Larry Kim: This is an article that I’m interested. And it’s a cultural reference that I actually understand. And look, you just have to think a little outside of the box here. You asked for a B2B example and every B2B company is doing content marketing. And every B2B company has a small budget for content promotion. this is an example where a little bit of creativity, targeting unrelated interests and then reflecting that back in the ad copy. So that it’s not just one interest that is drawing the viewer of your ad to engage with your content, but multiple interests simultaneously. So love of “Star Trek. Deep Space Nine Plus” a love of like political content. Those are unrelated. That it just opens up kind of new opportunities to engage with users in a much more effective way, than what every other freaking advertiser in the world is doing, which is targeting you based on your more obvious interests you’re following.

AO: I, man, this is so good. And I, like, I know that we can directly apply this just as growth marketers here immediately. So I thank you so much, Larry, for going through this example. And I love the humor in it. I think that’s a piece that we can’t miss and just the creativity and like, you just, you did a great job connecting something that, that you knew that that audience would find funny. And like, it sounds like you nailed that. And that’s something we can’t overlook that I know that’s probably the hardest part here. You didn’t pick, you pick the right thing to show them, but you did, but yeah, you nailed it. And I can’t believe that you spent $50 on that and got that kind of impression.

Larry Kim: So there was one other crazy thing that I did with Facebook that had a really amazing impact. It was kind of inspired by the Russians. The Russians weren’t just trying to promote content that was very one-sided to people who have held those biases. Like they were trying to create a fight. They were trying to create a food fight. The meaning is they would create content and then they would promote that content to the people like the vis of content, on the vis of social issues, they would then promote that content to people who held very strong views that were both for, but the genius part was both for and against that point of view. So what they, they kind of correctly inferred was that these Facebook algorithms, they really, really weight comments and comment replies, in terms of the weighting of their, ad and newsfeed algorithms.

So if you’re, this is going back to this point, if you’re just targeting the people who are interested in something. Then the comments which are so valuable, they have nowhere to go. They’ll just be like, I agree or great posts, but the stuff that the newsfeed algorithm likes to show to the surface is where there’s like lengthy, comments and comment replies. And comments, uploading, or downloading, like thumbs up or thumbs down of those comments.

Another experiment that we ran was just trying to promote content to both people who would eat this stuff up, but also people who would have an allergic reaction to that content. I did this with an article called, “Harvard study shows that working mothers, generate, oh sorry, raise more successful daughters”.

So “Working moms raise more successful daughters“. So that’s an incredibly divisive study because, first of all, stay at home moms hate that. Because they’re working moms too. They just happen to not be working in a conventional office setting. They might have it harder, and of course, you have the working mothers too, who will have mommy guilt. So they feel bad for having a nanny raise their children. And so they want to clean, they want to validate their life choices. So they see this article saying like, Oh, Harvard study shows that working moms, raise more successful children. So they’re going to engage with this. They’re going to write long comments.

Like, “this is why I work so hard, it’s to provide a great example to my daughter”, we targeted this, this piece of content to people there’s people who are female, but have CEO titles or managing director titles, really high powered women, parents. Everyone has children and home. So that’s kind of the people who are gonna eat this up and want to support and share this thing, like crazy. Right? Cause they want to validate their worldview. And then simultaneously we were spending 50 bucks to promote the content to the opposite. There are all sorts of targeting things like a stay at home mommy. There’s one Mommy CEO group worked for, for stay at home moms, just targeting the opposite, people who are not employed in a workplace, but have children and are raising their young children. And we just put 50 bucks to each group, then the sparks just flew, there were 64,000, reactions on that one post. Well, it’s because the magic is the comment replies and then the common engagement. So I just wanted to share that.

AO: It’s amazing. I mean, it just makes me think, man, it’s just, so it’s so creative, like you’re in touch with, with your audience.

This is what the Russians used. Like literally, they just wanted to create a food fight. Now, in terms of your own marketing, I’m not suggesting you start a food fight, but, there are definitely ways to spur legitimate discussion. Like if you’re promoting the keto diet, there’s going to be people who want, I don’t know, the paleo diet or something, so you can target others, people who have different strong views on something else. And a little bit of civil discourse, I think, goes a long way for these algorithms. So you just need to sort of think beyond this idea of targeting kind of the people who are likely to be interested in the ad. In the first example with the inverted unicorn, I was suggesting targeting unrelated uncorrelated interests in order to create a more, kind of captivating ad experience. And in the second example, I was suggesting targeting people with the opposing view in order to generate comments. And if you combine those two together, I think that’s a very effective way of using Facebook ads that still works in 2020.

AO: Larry, I’m excited. We got to try stuff like this in terms of just the two principals you said, and I cannot believe the low amount of money you spent. I mean, I just like it when I’m here 50 bucks, I’m like, Oh man, that’s not going to go far in, like, this is incredible. I wonder the equivalent, to get the equivalent traffic that you generated, how much money you had, would had to have spent in a traditional way of viewing ads. Yeah. Yeah. That’s like, you can’t, you can’t do it that way anymore because the costs are just too high. So you gotta get creative it, man, this is incredible. Our audience is gonna love this. Thank you so much, Larry. that’s all I got, man. We’d get, we got to have you on the show again and maybe a year or so when, as you continue with these really cool experiments,

That’s it. Another great episode The One Growth Show the official podcast of Growth Marketing Conference to learn more about upcoming events visit and subscribe to the newsletter. If you enjoy this episode, let us know. We’d really appreciate if you give us a 5-star rating, it’s super easy, just click the last star on iTunes and also share this episode on social media after all you want your network to know you’re the person they can always turn to for the best growth and marketing content, don’t you?