💥 Why more is more: how to use words to shatter the short attention span myth: w/ Stevie Lutgen, Senior Manager UA at Radish Fiction 👓

“We discovered that the more words we put in front of people, the better return on ad spend we were getting.”

Our guest today is Stevie Lutgen, Sr. Manager UA at Radish Fiction. In our conversation today, we talk about the importance of copy(especially long copy) in reaching and retaining valuable audiences. In a world where emphasis is placed on creating short digestible snippets of content, we were surprised to hear that Radish uses paragraphs and paragraphs of text in their most successful ads.

In our conversation today, Stevie walks us through the process of selecting, testing, and refining ad copy to best represent their product. We also look at how framing and identifying ‘hooks’ can be so pivotal to performance.

This week’s episode is a masterclass in how words, narratives and framing can so dramatically impact performance in a world where they are almost taken for granted.

KEY HIGHLIGHTS

Why more copy has correlated with stronger ROAS for Radish.

What Radish’s ads look like – and how they ended up with text-heavy ads.

Framing matters more than the story itself.

Watching Netflix for research.

How stories become ads: who on the team owns what part of the process.

How stories and narratives work for UAC

Telling stories on Snap and TikTok

How Radish is dealing with the absence of deep linking post iOS 14.5

KEY QUOTES

Framing is everything

And so we’ve just been discovering over time that us telling people, “Oh, hey, download Radish, it’s great, here are the reasons why” hasn’t been enough. And in fact, we discovered that the more words we put in front of people, the better return on ad spend we were getting.

Data tells a story that you shouldn’t ignore

We thought, “Let’s put in the first few chapters of our stories and see if that would get people interested and invested enough to kind of hop into the app ultimately and continue along that same trajectory.” It varied by story, but that had pretty astronomical results for ROAS and that’s what kept leading us down that road, of course, the data was there.

Understanding human behavior to identify archetypes that resonate

We’re all drawn to the same kind of tropes, characters, themes, motifs, over and over again, whether we realize it or like it or not.

Research Netflix for work

It’s not unusual for somebody to say, “Oh hey, as part of my day today during stand up, I’m going to watch this documentary,” or “I’m going to watch this new romance that came out on Netflix.”

Find what works and translate it across multiple ad types

We had tried so many types of videos, where we certainly harped on the features of the app– interacts with authors, interacts with the community, choose from many romance stories– that sort of messaging. And it has really paled in comparison to putting the stories front and center, as usual.

Refine your offering

The offering is not, “Hey, this is a free reading app,” the offering is, “Hey, this is a story that’s going to ignite your imagination and give you all of the passion and intrigue that you’re looking for.”

Make a seamless transition from ads to the onboarding experience

What we can do is curate an onboarding experience that puts those main themes right in front of people’s faces in thumbnail format. You hit the app, and boom, boom, boom, boom, boom, boom, here are our most popular stories or genres. So while somebody might hit the app and not get right to that exact story they were looking for, they could navigate there.

How to use onboarding to your advantage

We’re trying to incorporate a feel out survey at the very beginning of the onboarding experience. What are you interested in? Here’s a list of themes that we have available that we know are pretty popular, so which ones are speaking to you the most? And that can help us get them into a certain kind of onboarding flow that might expose them to stories that they’re more interested in.

FULL TRANSCRIPT BELOW

Shamanth: I am very excited to welcome Stevie Lutgen to the Mobile User Acquisition Show. Stevie, welcome to the show.

Stevie: Thank you, Shamanth. It’s cool to be here.

Shamanth: Excited to have you! Certainly because, of course, you come highly recommended but also because the last time we spoke, you expressed a perspective on using words and copy that I have found to be very unique and very interesting. So for both those reasons, I’m excited to dive into this episode with you today. So welcome.

Stevie: Yes, thank you!

Shamanth: Speaking of our last call, one of the things you said was, and I quote perhaps imperfectly, but this was the essence of what you said: “words are important in ways that I think are counterintuitive in a space where we usually say, ‘Oh, people have very short attention spans, keep things as brief as possible.’ That’s really not true for our product.” So say more, and tell me more about why this is the case.

Stevie: I think that it’s normal for people and advertisers to try to cater to this idea of the shortening attention span of the market in general. And I think that that still holds true in a lot of ways, like sticking to what your core offering is and keeping your language kind of punchy and relatable might be really important. But it’s going to depend on your product and Radish, the app where I work and where I buy media, is a serial fiction app.

So for our product, in particular, our offering is spicy, interesting, cliffhanger based stories in serialized fashion. And we’re finding the more of that content, that actual writing and that narrative we can get in front of people– no matter how much reading it ostensibly is– our ad performance improves.

Shamanth: And presumably, that’s also because you guys are a reading heavy product itself. And so in many ways, your ads are providing a preview of what the product experience is like, by having words in them. Can you talk about what your ads typically look like? And what the structure of your typical ads are like?

Stevie: Yes. So that can, of course, depend and vary pretty widely by channel but our biggest channel right now is Facebook and feed based advertising, where people tend to be in a reading mood. So what our ads typically look like are an image that is attention grabbing, and really speaks to the central themes of the story. So you can imagine those classic romance novel covers that you might see at the grocery store, with Fabio and his hair waving in the wind. We kind of have a modern equivalent to that kind of imagery that tends to capture a lot of people’s imaginations and speaks to the story and maybe the central conflict of the story. And that’s certainly kind of number one.

But number two is, in these specific feed ad units you’ll see a bit of copy at the very top like any typical Facebook ad. I think it’s like a 60 characters or something at the very top there– above fold copy snippet. But if you hit that ‘see more’ button and you expand our ad, you get up to 6,000 or 10,000, whatever Facebook’s limit is– I actually don’t know– you get that many characters of content and we’re adding full chapters of the stories there. So that’s what they look like. It’s a lot of words that you do have to expand, but then otherwise kind of like a classic romance cover type imagery.

Shamanth: Yeah. So there’s literally hundreds, if not thousands of words in this ad copy that’s about the image. Obviously, the image draws them in and people presumably read the entirety of that story. Certainly, the interested readers do this, obviously, that percentage is out there. And I know you talked about “Look, there’s this preconception that short attention spans, yada, yada,” so, I’m curious: what the process was of getting to a point where you have hundreds or maybe a thousand plus words in there. Was there a point where you guys were like, “Hey, let’s start with two lines, or maybe let’s get into a paragraph, and then let’s just go really big.” Can you speak to what that process was like?

Stevie: Yes, certainly, yeah. Because if you saw one of our ads and you expanded it, it absolutely looks overwhelming. It looks like something that no modern consumer would ever stop to care about. So we did start with this notion, like, “maybe what people are interested in, is the narrative of the stories.” So we were experimenting with putting very short snippets of those stories in the ad unit. And I’m talking just a couple of sentences, maybe a piece of dialogue that we felt could really draw people in or a description of a scene, but either way, it was very short. And we saw okay results.

But then we started to think, “You know, maybe our offering does go beyond just this one snippet of narrative and maybe we need to inject more words here to kind of gauge interest.” So we started experimenting with just a couple of paragraphs up to a handful of paragraphs and I’ll say that performance did not necessarily change at that point. It wasn’t until during one of our competitors snooping where we saw a competitor had started to kind of experiment with putting tons of copy in the ad unit, that we thought, “Maybe that’s something that we should do.”

But whereas the competition was putting a lot of disjointed paragraphs in there, just a lot of content but it didn’t necessarily all fit together.

Shamanth: Yeah. So in some ways, the ad can be seen as a standalone story experience. So in some ways they get to preview the product itself before they even install.

Stevie: Yes, definitely. And for some stories that had a huge impact on ROAS right away, and for others, it ended up taking a little bit more tweaking, but yeah.

Shamanth: Yeah. And I know you described the structure of this ad. Can you think of an example of an ad that you thought followed all of the practices that you described, you’re like, “Look, this has a chapter, this is almost a self contained story, so to speak,” and it didn’t perform? And then you guys went back to the drawing board and you’re like, “Okay, let’s rework it in this particular manner,” and then it performed. So are there examples of that sort that come to mind?

Stevie: Oh, yes, definitely. So the stories that performed really well, like I said, when we started adding one, two, three whole chapters of content there, we noticed that they had beginnings, or the start of the story, that were very hooky. The stories that were not doing so well, we still felt like the content here is interesting to people, certainly, because in app these stories perform very well. So people get invested in them, read several chapters in, invest in the app, that sort of thing. However, when we looked at that self contained ad unit, we realized for the less successful stories, that they got off to a very slow start.

My boss made an observation about one in particular. The story starts with a woman just sitting on an airplane interacting with a flight attendant and we really don’t know what’s going on. This isn’t very interesting. So from there, we started to almost go back to square one, where we had taken little punchy snippets of copy and that was the only thing in the ad unit.

Now, we started experimenting with different versions of those copy snippets above Chapter One, within the ad unit, such that if a user starts on a scene or an exchange of dialogue, that hints at a trope that they’re very interested in, which is huge in the content space.

We’re all drawn to the same kind of tropes, characters, themes, motifs, over and over again, whether we realize it or like it or not.

So once we started kind of pointing to those things in those paragraph snippets and putting them above that Chapter One content copy, it’s like, “Ooh, what’s going on here in this story, there’s this conflict between these two characters. It’s a little vague, but now that I’m getting into Chapter One, and it’s just a woman on a plane, not a lot going on, now, I’d like to keep reading and see what’s going on.” Or at least that’s our hypothesis.

Shamanth: Yeah. What is the hook for the story with the woman on a plane?

Stevie: This is counterintuitive or it took a lot of testing to understand this but the main hook, specifically for this story, over and over again we find is that we hint that this is going to be a romance that revolves around a Marine Corps member, and somebody who is seven years younger than him, it’s his best friend’s little sister, and his best friend was in the military with him. So you pull at those heartstrings a little bit. And you realize, this isn’t a story about a woman on a plane. This is a story about a woman who’s got this family conflict with this man who’s a Marine and somebody she’s loved for a very long time. So that kind of sets that stage.

Shamanth: Right. And I can instantly see how there’s so much more drama and tension. And the dude sitting next to his best friend’s little sister who’s seven years younger, presumably attractive to him. I could instantly see how that’s so much more dramatic than a woman sitting on a plane.

Stevie: Precisely.

Shamanth: Yeah. And if I had to think back to what I learned in Journalism School, and I don’t know if it’s the same phrasing you guys use, but this is like the lede. So you guys plant the lede ahead of the copy. So you’re not starting with Chapter One but you’re taking the highlight, putting it atop Chapter One so that’s the first thing they see.

Stevie: Exactly. And I was on my high school newspaper, so that really spoke to me. But I want to say we took cues from other content platforms. I do have to mention Netflix, for example. If you’re on any content platform that has video previews, if you’re hovering over certain thumbnails, for example, they’re going to be rapidly testing which segments of video are going to draw you in and drive your behavior. So that’s very core to how we operate as well, when it comes to testing these things.

Shamanth: I see. So you’ve spent time watching Netflix for research?

Stevie: Oh, yes. Yes, we do that.

Shamanth: Again, not to digress too much, what did that process look like of looking at Netflix and saying “Here’s what they’re doing and here’s what’s working?”

Stevie: That’s interesting. We use Netflix as a pulse for what America in particular is interested in. So it’s kind of like a matter of being the content consumer, first and foremost, and starting up with a program that gets you interested enough to keep watching it. And then paying attention to not just the preview that was given to you in the first place but the thumbnails that start to pop up as you revisit the platform, like what characters or themes are they highlighting more and more over time as I continue to navigate back to that particular show? So it’s a lot of having those observations and then coming together as a team and just talking through what all of us are seeing. We touch base with each other a lot that way.

Shamanth: Yeah, I can see that. It’s interesting how sort of observing what’s out there, how not common it is. It’s funny, I was talking to somebody that runs a gaming studio and this person is like, “Look, there are gaming studios where they spend one day a week, just playing games. And you would think it’s ridiculous but it’s not because we’re making games so we should be spending at least one day a week playing games.” So if you’re coming up with stories, that’s where you should be spending so much time, studying what’s out there.

Stevie : Absolutely. We talk about central conflicts, we talk about moments of big confrontation, we talk about resolution, and we all read together and discuss these things.

Shamanth: Absolutely. Oh. And if somebody says that in a stand up, are they somewhat obligated to share their notes or takeaways? How does that look? Is there a structure around that?

Stevie: That’s interesting, our content consumption within the company is important but that’s one of the things that actually is not very regulated or micromanaged. I’ll be clear, we have points of brainstorming every week where that information is incredibly useful. And so if you have teammates coming, and they’re saying, “Hey, I’m learning so much from interacting with this platform, reading this thing,” it’s a motivating factor for other people to do the same, or to find other platforms that give them similar insights. But it doesn’t become an assignment with work and a book report.

Shamanth: Right, right. It’s unstructured in many ways, it sounds like.

Stevie: Absolutely, to keep that viewing experience more casual and keep us all consuming content without it being this, “Oh, pressure, it’s a part of my job, it’s a drag.”

Shamanth: Yeah, just to segue just a little bit from how you describe things work on your team around this, who on your team identifies what stories to test or how the hook is framed, or what the ad experience itself looks like?

Stevie: So that has a long answer, because those are all different people and all the same people, all at one time. The story selection process starts with our growth team and our live ops team having in-app benchmarks for understanding how in-app performance works. Those benchmarks are coming from our data science team and they’re benchmarking against other successful stories within the app. So if a certain set of stories is displaying similar behavior, if you will, to our benchmark stories, then they become a candidate, essentially, for UA testing. And from that point on, what gets actually selected for advertising testing– “Does this story have legs with advertising, given a number of creative directions, a number of messaging directions,” that sort of thing– that’s what the UA team kind of does together.

We actually review stories from a list of already successful good candidates and we actually sit there and we think about “Is this something that we haven’t put out there yet, that we’re not sure if people are interested in but we think that we could market this really well,” or, “Hey, this is really big on Netflix right now or in the cinematic space right now and now we have this story that speaks really well to these themes. Let’s try to put that out there.” We go from that kind of qualitative to quantitative process when it comes to story selection. And then of course, from there, we have our own set of benchmarks for determining success, also very helpful, from our data science team.

Now, who decides what those ads look like: we have a Creative Strategist and a Creative Director and they work very hard to come up with a number of visual directions for us that might feel out certain aspects of a story. And a story can be kind of spun a number of different ways. And we find that our readers have their own imaginations while reading. And you’ll even see in the comments, “Oh, I imagine this actor playing this role.” “Oh, really, I imagined this actor” and they’re really quite different experiences. So we have a team that tries to peer into the imaginations of our readers and put out different versions of what might attract people visually in the market.

Lastly, you talked about who decides on those openers, I’m lucky enough to be heading up that copy funnel in the team. So while we have a lot of collaborators, both on our content team, pulling content in marketing, writing brand new openers for these stories, depending, I’m sort of the person who has a gauge on what themes might be trending well, and then how to test them, check our assumptions against them and conceive of new tests that pit one theme versus another, for example. And then we run those on our scaling campaigns if they’re successful.

Shamanth: It sounds like a lot of this stems from the content that is performing in the app. But there’s also a fairly detailed curation process of that to make sure these are marketing ready, UA ready.

Stevie: Yes, absolutely. Endless cycle.

Shamanth: Yeah, and I know a lot of what we talked about has been in the context of Facebook ads thus far. But I also know and understand you guys run ads on UAC. Considering that there’s almost no control that you have around the combinations of assets that get packaged into an ad, or the sequencing of those assets, how do you think about everything that you described in terms of having hundreds of words? How does that change in a UAC context?

Stevie: So yeah, we’re definitely limited with UAC, the number of characters we can put up there. However, the inspiration from those Facebook ads still does carry over nicely to UAC. For example, we’ll just take the very shortest, the most scandalous, the most punchy copy snippet or dialogues snippet, and we’ll make sure to have a selection of those available in the UAC unit. But additionally, we try to vary up the other copy that could be served. So we tried a one-two or one-two-three punch as it were, and also have a couple of other core kind of offerings available.

Number one, this is a free app, you can read a lot of content in Radish for free, you can read several chapters for free, if you don’t want to pay for anything you can wait to unlock. So we feel very comfortable putting that free language out there and it converts well. But also it’s an app where you can interact with authors in the community. And while that copy hasn’t traditionally done the very, very best, we can see it rise to the top for example, like the last time I spoke with you, the narrative snippets were really killing it on UAC. But in the last week, what’s risen to the top is actually a lot more conversation or a lot more copy about how you could have a dialogue with an author in this app, this app is a community.

So I think understanding your core offering, what sets you apart, as usual, from the competition is important to have cycling in there. But at the same time, we keep those narrative snippets, and then we make sure to have wildcards present. And as long as we’re touching all of those bases, we really don’t mind how copy is getting scrambled up, we kind of let the algorithm do its thing.

Shamanth: The essential promise of leading with a hook that draws a user in remains the same, even though the tactical execution is just different because of the nature of the platform itself.

Stevie: Absolutely.

Shamanth: And out of curiosity, how does this approach change on the other platforms you’re on? Let’s just say Snap or TikTok or, for that matter, if you’re on programmatic networks. Can you speak to how the structure or the approach changes for any of the other platforms you guys might be on?

Stevie: Yes, so for channels like Snap or TikTok or very story based platforms, meaning like stories ad format, I would say that it is largely the same. A lot of the same theming kind of carries over. And something that I haven’t even touched on yet is our success with video.

In these video formats, we find that providing still some glimpse into the variety of the stories available, that’s still helpful. But largely in video, again, creating a cinematic trailer or a kind of cinematic emotional experience that hints on the themes and motifs and central conflicts of these stories, is still the most effective. So for example, a video that says, “Hey, read Radish, it’s free. We have coin sales, you can interact with authors, oh, and we have a bunch of shifter romance stories” that’s always absolutely paled in comparison to a very moody, cinematic, full of stock imagery that’s kind of vague, type video with character narrative overlay. So some dialogue between characters that’s very heated.

You start watching a video like that, and you’re kind of like, “I don’t know what’s going on here,” which is also really counterintuitive. For advertising, we’re kind of trained on this notion, you need to let people know what your offering is within the first five seconds, otherwise, they’re gonna look away. And that’s the thing with Radish,

So that’s kind of how the learnings from Facebook translate over to these more video heavy formats, for sure.

Shamanth: Certainly, right. And I think video allows you to tell a lot of these stories in a much more graphic, much more feature rich way. And also, just to go back to your example of Netflix, if I have to think of the last one or two Netflix series that I saw, they were because I saw ads for the series, not necessarily Netflix, the platform. And I can think of at least one that I’m watching right now, this is a Spanish series, which was on the side of a bus stop. And I was like, “Oh, this is cool. This is intriguing.” And they were not advertising Netflix the platform at all. So that absolutely makes sense.

And you talked about the different stories or different kinds of content that are present in ads that drive users to specific parts of the app or specific stories within the app. Do you find that some specific kinds of content that results in paying subscribers or high LTV users as compared to other kinds of content that eventually results in not so high value users or just non-subscribers? What have you found or seen?

Stevie: Yeah, that one’s pretty interesting. For us, sometimes it’s hard to tell if it’s the creative direction that’s creating that disparate behavior. And that’s why we do go through a few iterations of different creative messaging and testing before we decide “Okay, yeah, this story might be an interest driver,” for example, “with really high IPMs, but for some reason, it’s just not driving investment.” That particular example isn’t all that common, by the way, although I will say that I continue to see this theme with vampire stories in particular, they seem to drive a lot of interest, yeah. But then people get in the app and they don’t want to invest in the way that somebody who’s very much into boss romance, office romance, or even werewolf shifter romance, those types of people tend to invest more in Radish.

So yeah, we do see that behavior and it’s certainly interesting. But we try to consider not just, “Oh, you know, people must be interested in vampires but everybody who’s interested in vampires doesn’t ever want to open their wallet.” We know that that’s probably not a totally fair assumption, either. Though, we try to explore how the in app experience might be driving that behavior. For example, we might explore, “Okay, we need to acquire additional content that’s about vampires and vampire romance.” Because it’s not just a single story that drives interest in Radish, it’s understanding, “Hey, this platform can provide a number of stories and experiences that speak to me and that resonate with me.” I doubt you would keep going back to Netflix if they just had one show that you liked, other platforms would probably catch your attention. So we try to take that into consideration as well.

Shamanth: Yeah, so the depth of content is important for the longer term retention and LTV of a user. Even if the ad just elicits interest, the ad can result in higher IPM, but maybe the work needs to happen to ensure there’s enough content in the app itself.

Stevie: Precisely it’s like that optimization funnel, like with any product, never really ends. But you have to think about just the many, many sub niches out there and how that might be affecting your upper funnel metrics. For sure.

Shamanth: Certainly, and having that funnel in place, I imagine quite a bit of that has been a function of being able to deep link users to specific stories. Somebody who watches an ad about vampires goes right into that particular vampire story through deep linking. At least on iOS, that’s going to go away, actually, potentially this week. I think it started rolling out, it’s not yet on my phone, but it could be any moment at the time of this recording and certainly by the time we release this, it’s likely not going to be available at all. So in the absence of deep linking, how are you thinking about adapting to the potential break in the funnel, so to speak?

Stevie: So we’re doing it in a few different ways because, first of all, yes, you’re exactly right. We do run AAA ads where you can’t deep link, we do test running without deep link, but consistently, even if a non deep linked ad unit, for example, outperforms a deep link to add unit initially, the deep link to ad unit does have more longevity. So yeah, it’s a concern 100%. But the way that we’re thinking about that is, let’s do what we can control.

First of all, we’re lucky to have an app that has very strong audience demand around specific stories. So right now, at very least,

Almost like taking that a step back/a step further, we’ve been, probably like a lot of other companies, I know that this has been hitting me personally as a consumer–

And we can kind of capture that information and habit.

And then last but not least, we’re working really hard on deeper recommendation engines within the app, such that if you’re clicking through a story, and we see that a user is really invested, they get X amount of chapters in, we can say, “Oh, hey, you’ve gotten to the end of content for this particular story,” or “You’ve gotten to a certain goal post in this particular story, these other stories may be of interest to you, based on a number of things.” So we’re trying to rely less on the kind of invisible data that’s been helping marketers and developers make a lot of these decisions and instead we’re putting that power back in the users hands or otherwise trying to find intelligent ways to surface new options to them.

Shamanth: Sure. So you’re moving a lot of that decisioning within the app. Whatever was happening somewhat automatically via Facebook is now actually moving within the app. And like you said, a lot of the control will stay in your hands. It’s not as fickle as what things are going to be like with Facebook. Certainly, that lets you guys control your fate and destiny just going forward.

Stevie: Right.

Shamanth: Stevie, this has been very, very instructive and certainly very counterintuitive compared to the way in which the vast majority of marketers approach their marketing and their ads. This is perhaps a good place for us to start to wrap up. But before we do that, could you tell folks how they can find out more about you and everything you do?

Stevie: Yeah, if for some reason they want to learn anything about me, I’m certainly on LinkedIn. But a great place to start is by looking at Radish the app or our website– which we’ve just redone– and just having a good time, having fun. The content is there to have fun with.

But certainly, if anybody’s interested, we’re also hiring at Radish and we are looking for Senior UA Specialists. And we’re looking for people who are curious about this process, are curious about the way that we collaborate within the company, and use it as a jumping off point that supports and helps us explore new channels. Because we think that there’s a lot of opportunity out there to capture more readers. And especially versus the gaming background that I’ve had, finding users for Radish is kind of an adventure. Creating ads is definitely an adventure.

Shamanth: That’s certainly very creative in ways that you just described just now. We will certainly link to your LinkedIn, Radish, and of course, the position you guys are hiring for in the show notes so folks can check those out. But for now, this is perhaps a good place for us to wrap. Stevie, thank you so much for being a guest on the show today.

Stevie: Oh, yeah. Thank you so much, Shamanth. I’m flattered and I hope somebody gets something out of this.

Shamanth: Certainly a lot of people will get a lot out of it, myself included. Thank you again.

Stevie: Thank you. I hope that you have a great day.

A REQUEST BEFORE YOU GO

I have a very important favor to ask, which as those of you who know me know I don’t do often. If you get any pleasure or inspiration from this episode, could you PLEASE leave a review on your favorite podcasting platform – be it iTunes, Overcast, Spotify or wherever you get your podcast fix. This podcast is very much a labor of love – and each episode takes many many hours to put together. When you write a review, it will not only be a great deal of encouragement to us, but it will also support getting the word out about the Mobile User Acquisition Show.

Constructive criticism and suggestions for improvement are welcome, whether on podcasting platforms – or by email to shamanth at rocketshiphq.com. We read all reviews & I want to make this podcast better.

Thank you – and I look forward to seeing you with the next episode!