Bringing Data and Creativity Together to Create Better Content With Anastasia Leng From CreativeX [AMP 230]

As a marketer, do you need to create content that applies insights derived from data and research? If so, pay attention to the right data and apply it the right way to produce the most effective work possible. Today’s guest is Anastasia Leng from CreativeX. She talks about where marketers get misled with data and how to merge data and creativity to create content that connects with customers. Some of the highlights of the show include:

• Data-backed Content: Objective way to understand what’s in each content piece
• Performance: After putting piece of content out, what has happened as a result?
• Views/Variables: Marketers should move away from biases and assumptions
• Trust gut or data? Marketers want to be right; get comfortable with being wrong
• CreativeX’s Mission: Enhance and elevate creativity expression through data
• Consistent Content: Number of clicks vs. what reflects brand and audience
• Best Practices: Creative quality and distinctive brand assets to increase sales
• Cheat Sheet for Content:
  • What should your definition of creative quality incorporate?
  • Brand right away; marketers have 2-3 seconds to make impression
  • Don’t waste money by running same piece of content across channels
  • Get around brevities

If you liked today’s show, please subscribe on iTunes to The Actionable Content Marketing Podcast! The podcast is also available on SoundCloud, Stitcher, and Google Play.

Transcript:

Ben: Hey, how's it going this morning? Anastasia: Hey, I'm doing great. How are you? Ben: I'm doing fantastic. Do you mind if I ask, where are you based out of? Anastasia: Wow, it's surprisingly a charged question. Pre-COVID, I split my time between London and New York, spent an equal amount of time in both, and was truly bi-coastal or bi-country I guess. Then when COVID hit, I had to make a decision on where to plant myself. I chose London thinking, oh, it's a nice way to spend a couple of weeks in London, and have basically been stuck here ever since. Currently, I am based in London. Although when the world opens up, I'm hoping to get back to my transatlantic lifestyle. Ben: Sure. Very cool. I grew up in a military family. My mom's side of the family is from the UK. We moved back and forth a few times. We lived just outside of Oxford, and I graduated from high school just outside of London. Anastasia: Lovely. Ben: Yeah, I'm very jealous. Anastasia: Where else did you live as part of your military traveling the world? Ben: All over. I spent a lot of time in North Dakota, which is where I'm based now. This is where CoSchedule's two offices are here in Bismarck and Fargo. I also lived in Colorado, Utah, California, and a few different places in the UK. I was born in Norridge. Been all over, but I love London. Amazing city. Anastasia: It wasn't military, but my family traveled a lot. I was born in Russia, then moved to Vietnam, spent a little bit of time in Hungary, and then Bahrain. And then of course, the journey culminated ending up in New Jersey of all places. Very anti-climatic. Ben: All roads lead to New Jersey. Anastasia: All roads lead to New Jersey, of course. Since then, I have gotten the chance to travel for myself and now spending quite a bit of time in London and New York. Ben: Very cool. What we're going to talk about in this week's episode is creating data-backed content, which is a hot topic in content marketing circles. But it’s maybe something that people have a lot of questions around what that actually means, and what are we actually talking about when we say data-backed? Why should we all care about this? The question I have for you, when we talk about data-backed content, what are we actually saying or what does that actually mean? What does top-notch, data-backed content actually look like in the real world? What's an example of this type of content that a listener might recognize? Anastasia: There are a couple of questions in what you've said. When we look at what data-backed content means, at least what it means to us is two-fold. The first one is having a projective way of really understanding what is in each piece of content. If we look at the way that marketing has evolved, we have tools that help us analyze and really dissect every little thing that we do in structured ways. We know these are the different audience profiles we're reaching. These are the different sites we're targeting or different times of the day. We have ways of bucketing information. When we analyze the content or creative we put out, it's still very, very subjective. The first part of getting to data-backed content, our belief is how can we structure and understand the kinds of content we're producing and analyze it either thematically or by an object in a way that we're looking at things consistently from a piece of content to piece of content? That's the first part. Maybe the better way of saying it is technology-backed or whatever the case may be, but it's really about having a common language for valuing every piece of content you create. The second part of data-backed content of course ties down to performance. Performance will mean different things to different people. But it's really about saying, hey, you've put out this piece of content into the world, what has happened as a result of that? That result does not have to be short-term. In the marketing world, we tend to prioritize analyzing through the data that is immediately available even when that is not the right data to analyze for. We might optimize for clicks, conversions, views, or likes even if those metrics don't actually mean anything to us. The second part of data-backed content is understanding what is the long-term metric that really matters? How can you start to get creative, and whatever that KPI is, closer together so you can understand the relationship and the journey these two things take together? Ben: Sure. It sounds like there are two sides to it. There's data that supports the content and its execution. There's data that also guides performance and measurement. Anastasia: Yeah. When we talk about the first bit, what we're trying to move away from and marketers should be trying to move away from is this view of saying, hey, this piece of content did really well. It did better than this other piece of content, so we're growing to draw some inferences about what is about this individual piece of content that did well and use that to form a bunch of our decisions. The problem with that approach is that it naturally plays to some of our worst biases. I'll give you an example. When we first started the company, we were working with a major CPG brand that put out a lot of hair care products. One of the first pieces of analyses we did for them is we tried to evaluate all the content they were producing for different variables. We looked at the presence of people in the content down to did the model's hair color matter in terms of driving better performance. What we saw was when we got the data and the results—we were sat in the room with about 50 marketers who worked on this brand. We say, hey, do you think hair color makes a difference to marketing performance? They said, yes absolutely, it does. We said, okay, great, which hair color do you think drives impact for this brand of yours? We showed four options. We showed blonde, brown hair, black hair, and red hair. We went through the room and got people to vote. And 90% of the people in the room said that they believe models with red hair would perform best in their marketing campaign. When we asked them why, we heard lots of great answers. We heard that red hair was the most unique. It was the one that was most rare to see. It was also the one that offered the highest contrast. Lots of reasons that you could find yourself buying into. But ultimately, the data didn't back any of that up. My point here and what this is really trying to illustrate is we need to get to a place where we can really analyze content more objectively such as we're not letting our own biases drive our understanding of what is working and what is not working in our content strategy. The way that we think about it is that if we look at content and that first part I mentioned of breaking it down, it's very important to become objective and specific about saying, okay, this type of content, we're going to bucket this as content that really illustrates the human part of our brand versus this part of the content is content that we feel is very consistent with all of our brand values. Again, it's hard to come up with examples that are specific to everyone on the call. But hopefully, people understand that it's really about objectively understanding what is in your content either from an object point of view—does it have our product, does it have our brand, does that have a call to action—as well as more of a thematic point of view. Ben: I think that makes sense. What, in your opinion, do you feel leads marketers toward trusting their gut over trusting the data? As marketers, we like to think of ourselves both as creatives certainly, maybe not numbers people, but we like to say that we're data-driven. If that's the case, what do you feel steers people wrong? What kinds of tendencies are they giving into that are steering them toward acting on assumption versus acting on something more concrete? Anastasia: I think there's a trait or behavior that is on us as marketers. But the flip side of that is there's also something the industry is doing which is not helpful. I'm hoping to tackle both of these. The trait that impedes us sometimes from embracing data is a very fundamental human need to be right. I studied psychology at university. This is pop psychology for anyone listening, but I remember one of the things that always stayed with me is that the two most basic human drivers that we all collectively share as people are the need to be liked and the need to be right because they fundamentally provide some justification or belonging to us being here on this lovely planet. There is the trepidation of hesitation of embracing data to its full capacity is sometimes the fear that the data will prove us wrong. What does that say about us? That's especially challenging when you're in a marketing role because part of the beauty of being in marketing is your ability to understand, dissect, analyze, and ultimately try to predict and even create human behavior. When you know there's something that can prove you wrong, that's uncomfortable. What we've seen in the companies and the marketers who have really done a great job here is the idea of saying, I'm going to use data to be right more often or to be more certain about being right. But the initial hesitation is this will prove me wrong and that's scary. I do want to say that a lot of the way the industry has been talking about diffusion about data and creativity is categorically unhelpful. What I mean by that is when we look at the intersection of these two things—our company mission at CreativeX and the purpose of as to why we exist is we want to help enhance and elevate creative expression through data. That doesn't mean that creativity is now dictated or driven by data. What we've seen a lot of companies talk about—which to us fundamentally counters the way creativity works—is this idea of I'm going to tell you if there's a cat in your creative and that drives click-through, therefore, you should put a cat every ad. Fundamentally, we don't believe that that's how creativity works, should work, or will ever work. Especially if you think about the extent to which we have data and how much of it we have. We've talked about big data, the data, the haystack problem, and all of that. There is the view of every data point we get automatically needs to be applied. When it comes to creativity, that's just not how it should work. The way that we try to look at it is to say, rather than going to this object-based detection which is really about hey, we're going to tell you having cats, lemons, or the color red drives more clickthrough—which again we think is categorically impossible to apply in any meaningful way—let's think about how we use technology to measure larger concepts in a way that is customized to the kinds of questions and things that the marketing teams are asking. Let's use that to drive conversations about impact. Basically, what we're trying to think about is how do you paint a canvas and a structure that says, these parameters lead to the best type of creative for your brand, your industry, that specific channel, that specific campaign objective, and then have the people who are creative—put that inside on steroids. A good example and something we do with a lot of our brands is we try and think about what does creative quality mean to you? If all the marketing teams sat in a room and they said, hey, this is an example of what we consider a good creative. It's creative that has our brand visible. It's creative that is optimized for the environment and the audience, which is different when it's Snapchat versus TikTok versus YouTube. It's a creative that shows our product being used. It's creative that humanizes our brand. Again, I'm obviously making all this stuff up. Then, we start to think, okay, let's take all of this and let's create an analysis that says, hey, here are the creatives that meet your quality component and those that don't. Let's see if those are the right things that actually help move the dial. Ben: It's extremely easy to use data to confirm our own biases rather than to use it to objectively lead us toward the best course of action. I love the hair color example that Anastasia mentions because it shows how true that is and what the consequences can be if we don't catch that happening within ourselves. Everyone wants to be right, but as marketers, I just think that example is such a great reminder that sometimes we need to get comfortable with being wrong more often so that we can use data to make us right more often in the long run. Now, back to Anastasia. It sounds like what you want marketers to get away from—and this is something that marketers understandably do a lot. They will look at a piece of content that did well, identify one thing about it that was unique from whatever else they're doing, and then they will erroneously assume that that one thing was what drove the result. Then they will run that thing into the ground when it wasn't even the thing that drove the result in the first place. It was an assumption. They thought it was data-backed because they did a thing and they saw a result. It sounds like there are some fallacies between correlation and causation going on. Anastasia: Yes. Obviously, people can't see us right now, but I am vigorously shaking my head in agreement here. You're absolutely right. The problem with this is when we look at a piece of content, the way it happens is exactly as you described. We’ll identify something and we’ll say, hey, creative A is different from creative B because it has a person who has red hair in it. That's why we need to put redheads in all of the content that we produce. The danger here is that creative could've performed better because it was correctly sized for the platform we ran on or because it introduced the brand super early. Something that actually is not visible or really easy to perceive by the naked eye, which is where the use and the value of technology come in. The danger of this, other than from all the stuff we've talked about is the institutional danger here. Those fallacies are very, very difficult to undo. Going back to that example we started where we talked about that hair care company where 90% of the audience believed that casting people with red hair is the critical business decision they need to make for all their content. When we drilled deeper into that, what it turned out was that there was one person on that team who was very senior who advocated that belief. That perception basically filtered throughout the whole organization where people took it as gospel that we need to cast people with red hair in our content, otherwise, we're not doing a good job, which was someone's subjective assessment which proves to be not really the right one. Ben: That's really interesting. In that example too, it seems like everybody looks to the person who is the highest up on the org chart and follows what that person says because obviously, they got to where they were by being good at what they do and probably being correct most of the time, at least would be the perception. Which is probably not true because quality, creative work, and marketing, you get to a positive result by making lots of mistakes very quickly. If we can identify that as being a core driver, how these biases tend to perpetuate, these misperceptions get taken for gospel, on an individual level, how can a creative professional or a marketer not only get comfortable with being wrong frequently but to actually embrace that as a positive in itself? How do you fall in love with the process of being wrong in order to get the results that are going to be what you want? Anastasia: We're starting to stray from marketing and maybe talking about human behavior and human dynamics. I think it's very difficult. The way that I think about it—both from a personal level as well as the company level—is on a personal level, I think about it as you have to be wrong so you can be right. You can't get to the right answer if you don't know you're wrong. One of the things I remember my manager saying to me very, very early—I spent about five years working in Google and one of my managers said this to me. She said, your quirks in your 20s become your calling cards in your 30s. This was a conversation that was around getting honest feedback about your performance. None of us like feedback especially when it's critical. We all love good feedback, I hope. But the point that she was trying to make is it's the things you don't know that ultimately hurt and impact you sometimes in very pervasive and long-lasting ways versus the things you do know because at least, those things you can fix, address, or somehow incorporate. Taking this back to marketing. The issue is if you're not even aware of these things, then how can you truly be a good marketer? How can you truly put out great content if you're not actually able to really look deeply within it? Again, great content doesn't mean content that gets the most clicks. We actually think it doesn't, for the most part. It means content that is consistent with your brand vision. It means content that is reflective of your audience. It means content that's very specifically set up and optimized for the different audiences and channels that it's going on, not content that has a cat or a lemon in it. These are much broader questions that the industry likes to talk about because it's much sexier to say, lemons drive more clicks or whatever the case may be. It’s seeing this as a necessary step of the journey. Sometimes, when I've been uncomfortable with getting data, I try and take a step back and think about if a competitor had this data and they acted on it or incorporated that, would that enable them to do a better job? Is my inability or unwillingness to embrace it actually hinders my own ability to be successful? Ben: For sure. I think you're absolutely correct that we are straying from the execution and the practice of marketing, but what is marketing other than an understanding of human behavior? If these things sound they’re disconnected, maybe that is where the relevant connection gets made. It's super important though to be talking about these things because I agree, these are things I don't hear people in marketing circles typically talking about because it's way easier. Also, sometimes, it is probably easier to explain to a stakeholder or to a client, this thing did well because the model's hair was red or because it had a cat, it had a lemon in it. It had something simple that someone could point at and be like, yeah, you're right. That ad does have a cat in it. Then suddenly everything has cats. That's way, way easier than to look at something with a more critical eye and really drill into what was driving performance. Say a marketer finds himself in a situation where maybe it's with the hair care company. They're looking at surface-level things and assuming that the loudest or most respected person in the room, whatever they think is correct. How can you get away from that? In order to get away from that, if a marketer is in a position where they are assessing the performance as a piece of content, what are some types of questions that they could start asking about that content to get to a better answer other than just looking at it and drawing a conclusion that any normal person could have? Like looking at an ad and being like, oh, the model's hair was red. Clearly, that's why it succeeded. You could look at anything in that ad that you wanted to. If that's all you're going to do, you could say that literally, any part of it drove the result. Anastasia: Yeah, absolutely. The journey starts with actually not looking at a piece of content or an ad individually, but only looking at this data collectively. Because when we look at a piece of content—creative A versus creative B—what we're simply doing is plotting two points on a graph and drawing some line between them. There are infinite ways to get between two points. When we look at any chart or any trends, it has many, many points that ultimately tell us, this is the direction in which things are going. The only way to truly get this analysis is to move away from this creative by creative analysis, really a step back and say, let's look at every piece of content we produced over the last quarter. Let's measure it consistently based on the changes we're making, the high-level things we're thinking about, and see how that's moving and impacting our business dynamics. Some of the things where we see people do this is, one is around this concept of creative quality. Creative quality and something that we think of creative health is ultimately very platform- and channel-specific. If you're a marketer who has other people above you whose voices sometimes ring louder than yours, there's a lot of industry research that's produced by all the major platforms out there—Facebook, Instagram, YouTube, et cetera—that say here is data aggregated across thousands of campaigns, thousands of videos and images that have been running around our platform that are industry-specific that point to the creative best practices for our platform. Facebook calls this their brilliant basics. YouTube calls this their A, B, C, Ds. But one way to start, especially when you don't have all the data or when you don't have thousands of creatives that you can analyze is looking at those best practices and saying, hey, are we doing these things? If not, is that because that's not part of our definition of creative quality or is there some other reason? That's one thing. We can talk more about what that is so people can have a checklist they can think about their own content with. The second bucket of things that we look at is what we think of broadly as brand consistency. Brand consistency means trying to be very objective about what is it about your brand that is unique. Again, there's a tremendous amount of research about the impact of branding on sales. Even a percentage point improvement in brand recognition has an outlier effect on sales. What a lot of the big brands that we work with tend to do is they have this concept of distinctive brand assets. Fundamentally, these are cognitive shortcuts that signal to a consumer, oh, this is the brand. For Coca-Cola, it might be the color red or the shape of its bottle. For McDonald's, it might be the [humming]. I don't even have to tell you who the brand is. Either of those things seen in isolation, it's crazy that when the color red is seen in isolation without any brand, you'll be, yeah, that's Coca-Cola. The point I'm trying to make here is every brand has some elements of this. They might not be as rigidly enforced, but fundamentally, we think that's a problem because ultimately, it erodes those cognitive shortcuts that you're trying to create between your brand and your consumers. The second pillar we see a lot of brands focus on is this concept of creative and brand consistency. What are those things that you're trying to create these cognitive shortcuts with? How often are you utilizing them? Over time, do you see that consistency actually ties into sales lift, brand awareness, brand recognition? Which of these distinctive brand assets ultimately help you get there? Ben: That makes sense. If any of our listeners are curious to dive into this a bit more with bringing their creativity and branding in data together to make better decisions—this will be a two-part question—one, are there any other resources out there that you would recommend they check out? Whether those be books, courses, research reports, or anything like that. Two, if they were interested in going to CreativeX potentially to get help with this, what would be the best place for them to find you? Anastasia: The second question is easier to answer. You could always find us at creativex.com. If you have any questions, info at CreativeX is the best place to start. Even though we predominantly work with large Fortune 500 brands, we do our best to take any aggregated data and insights and publish them on our blog to help inspire teams that are smaller to take things away that they can apply to their marketing today. In terms of resources, some of the best practices on creativity, there's Think with Google which does publish some of its own entrepreneurial research occasionally. There are some bits about creativity that are very powerful. There is a book by Jenni Romaniuk called Building Distinctive Brand Assets, which talks about the importance of creating these creative cognitive shortcuts and the impact this can have on your brand and ultimately your sales. That's a book that we read and talk about here quite a bit. For marketers that are looking to dive in as well, if it will be helpful, I'm happy to share a list of 3–5 things that they should think about for their own content without having to read these things. Almost like a cheat sheet just to make sure that their content is at least meeting the minimum bar to be seen and heard in these environments. Ben: Yeah. That sounds great. Anastasia: Cool. This is based on lots of analysis around what should your definition of creative quality incorporate. Again, the context we’re all facing, the uphill we're all facing as marketers is there is so much content out there, so much advertising. Consumers are overloaded with messages. How do you make sure yours sticks out? The first thing that we know and the context that marketers think about this from is the time we have to capture consumer's attention has decreased. That's across B2B and B2C. That's across thought leadership, [...], as well as ad content, et cetera. On average, we find that 2–3 seconds is really all you've got to make an impression. That doesn't mean your content—if you're running video—needs to be two or three seconds, but it means some of your core messaging and branding needs to come across. With that in mind, one of the key creative quality elements that we see our brands track is this concept of, you got a brand upfront or brand right away. Let people know that this content belongs to you. That doesn't mean putting a logo on it. Again, it might mean what are the other unique elements of your brand that you can put in in the first 2–3 seconds. If that's all people see, they at least start to build some of that brand awareness and brand association. The second one comes around content consumption. What we see on environments like Facebook and Instagram, the vast majority of content is watched without sound. If you have a video that you're running that doesn't have subtitles or ideally isn't optimized for being watched and understood without sound, that's basically wasted media spend. YouTube, for example, is the opposite where designing for sound is very critical because it actually helps amplify the message. Another thing that we see, and this is an unsexy one but one that we found makes a very big impact, is there's still a tendency to take one piece of content and then deploy it across every platform. But actually, an environment like YouTube is optimal for a 16 by 9 ad slot versus Instagram as 9 by 16. If you run a YouTube piece of content inside an Instagram story, you literally got about 80% of your real estate that's blacked out and you utilize a tiny percentage. Again, wasted money, not even giving your brand the chance to be seen and heard. Another one within that is under the spirit of making a quick and lasting impression is whatever the product is, find a way to show it, mention it, et cetera. Again, this is where people can get creative. It doesn't mean slap the product on it but show it early, show that consumption moment happening early. Then the last one that I'll leave people with is really get around brevity. There’s a number of pieces of research that shows that when you think about creative as a combination of visual sound and copy, there's research that shows that when a copy is under 280 characters, it does, on average, tend to drive higher ROI because it helps get your message across in a much faster way.