Episode 283:

Artificial Intelligence (AI) is a hot topic but many agency owners wonder how, exactly, it can benefit their shop. They get how it works but they think that either a) they don’t understand it enough to deliver on it for clients or b) they can’t afford the AI tools out there for agencies. The truth is – now is the time to experiment. Most agencies are at the infancy stages of learning about AI and how to use it to serve clients better. But we have to get up to speed or we will find ourselves left behind. The good news? It’s within your grasp.

It’s with these challenges in mind that I was inspired to reach out to RJ Talyor and have him on the show. RJ’s SaaS company, Pattern89, works with agencies and brands and improves their ad effectiveness and efficiency by measuring over 49,000 characteristics of a digital ad. With that data, they can provide insights on what’s working and what isn’t, as well as the predicted shelf life of any ad. Through their analytics, they’re also able to predict with over 95% accuracy whether or not a piece of creative is going to work.

In this episode of Build a Better Agency, RJ and I discuss the many ways AI insights are quickly becoming a necessity for agencies of all sizes. We look at the specific data it can provide, as well as how an agency can begin to experiment while convincing even the most dubious of clients of its importance. We also talk about the way the creative life cycle is changing, how agencies approach AI wrong, and what is coming next in the world of artificial intelligence.

A big thank you to our podcast’s presenting sponsor, White Label IQ. They’re an amazing resource for agencies who want to outsource their design, dev, or PPC work at wholesale prices. Check out their special offer (10 free hours!) for podcast listeners here.

AI for marketing

What You Will Learn in This Episode:

  • The specific data AI for marketing can give to an agency
  • Why AI doesn’t undermine creative
  • How small to mid-size agencies are leveraging AI marketing insights
  • Why AI is quickly becoming a need-to-have for agencies of all sizes
  • How the creative life cycle is changing
  • How iOS 14 is changing the way marketing will work
  • Why the speed of change is disruption A/B testing
  • Ways agencies can begin to dip their toes in AI
  • Where agencies get AI wrong
  • What is coming next in the world of AI
“Beauty is in the eye of the algorithm.” @Pattern89 Click To Tweet “Data sits between creative and performance.” @Pattern89 Click To Tweet “In three to five years, AI is going to be everywhere.” @Pattern89 Click To Tweet “The idea of evergreen creative is going away.” @Pattern89 Click To Tweet “Data can help an agency guide a client over a subjective opinion.” @Pattern89 Click To Tweet “We always have to be at the point where we’re predicting what is coming next.” @Pattern89 Click To Tweet

Ways to contact RJ Talyor:

Additional Resources:

Speaker 1:

It doesn’t matter what kind of an agency you run, traditional, digital, media buying, web dev, PR. Whatever your focus, you still need to run a profitable business. The Build a Better Agency podcast, presented by White Label IQ, will show you how to make more money and keep more of what you make. Let us help you build an agency that is sustainable, scalable, and if you want down the road, sellable. Bringing his 25+ years of experience as both an agency owner and agency consultant, please welcome your host, Drew McClellan.

Drew McClellan:

Hey everybody, Drew McClellan here from Agency Management Institute. Welcome back to another episode of Build a Better Agency. I’m excited to share with you the insights that our guest is going to have for you. He’s got a really interesting lens and perspective into the business. Speaking of an interesting lens into the business or perspective, I want to remind you as we are rolling into 2021, I want to remind you that we have a tool on the website, absolutely free for you to use, that will allow you to do your own looking through a different lens at your agency. So basically it’s a health assessment of your agency, and it looks at five key areas. Your account executives or account managers, your agency’s finances, your agency’s biz dev efforts, your happiness as an agency owner, and how you’re managing the team. So it basically walks you through a series of questions and in five or six minutes you’re going to finish the questions. You will get instantaneously a report that talks about where maybe you need to spend some more time and where you’re super strong.

Drew McClellan:

You will also get that report by email so that you can keep it for your very own and you can set some goals. And then what happens is, based on your score, then you’re going to get some emails from us that are going to point you to other free resources for you to shore up wherever it is that you have the most room to grow. So to take that assessment and to get all of that support and all of that help, you just need to go over to agencymanagementinstitute.com/assessment, assessment is single, not plural. So head over there, take the assessment, and then take the data that we give you and begin to think about how you might tackle some of those issues as you march into 2021. This would be a great leadership team discussion, a great discussion for your next traction meeting if you have one, and we are happy to make that available to all of you. So head over the agencymanagementinstitute.com/assessment.

Drew McClellan:

All right, let me tell you a little bit about our guest. So RJ Taylor is the founder of a software as a service company called Pattern 89. And what they do is they, and I’ll let RJ explain it to you in a far more eloquent way than I will, but basically what they do is they partner with agencies and brands and they use machine learning and artificial intelligence to help you improve the ads that you’re putting out into the social channels. And they do that by measuring over 49,000 characteristics of a creative ad and being able to tell you what’s working and what’s not working, what is getting reaction from audiences and what isn’t. They’ll also be able to tell you, predict how long the ad will be effective, they will help you write the creative brief, they’ll help you iterate the ad. So it’s really interesting.

Drew McClellan:

But what I want to talk to RJ about is what he knows and learns based on and watching all of their clients’ ads and brands go through this process. So what is AI teaching RJ that would be valuable for us? How is all of the algorithms and all of the changes that are happening in the technology, how is that impacting the way we think about creating and testing ads? And some other questions like that. So I’m anxious to learn what he thinks we need to learn, because I think this is a place where many agencies are still in their infancy. And so my job, of course, is to pick his brain on your behalf, and I’m ready to do that. So let’s get to it. RJ, welcome to the podcast, thanks for joining us.

RJ Taylor:

Thanks for having me so much, appreciate it.

Drew McClellan:

So tell the listeners a little bit about your background, about Pattern 89, and how you came to have all of this insight into how agencies are using AI, things that we can be doing with it, and all the other things that we’re going to talk about today.

RJ Taylor:

Sure, sure thing. Well, I’ve worked in the digital marketing space for almost 20 years now, and started at Exact Target at a pretty robust partnership program with agencies, and then worked at Salesforce, and then I’ve continued to work in digital marketing. And across all that experience, what I’ve found is that marketers just need to know what’s going to work from a creative standpoint. And they’re doing a lot of A/B testing to figure out, “Does this one work, does that one work?” And it almost goes without fail that once you find the creative that works, it doesn’t work any more. Because trends changed, means changed, things changed, et cetera, and then we’re back to the drawing board.

RJ Taylor:

So it was with that idea in mind, or that problem, that we solved it with Pattern 89. And with the advent of AI, machine learning for both computer vision and natural language processing, we can understand what’s going to work. And so started playing around with some new algorithms about 4 1/2 years ago and built a business around it, and now we’re serving clients across the globe, and it’s been a lot of fun to blend my creative background, because I have an English degree, I’m a liberal arts guy, with the latest in technology, it’s a win-win on both sides of blending those two together.

Drew McClellan:

So give the listeners an idea of what, if they’re not familiar, what Pattern 89 is and what it does.

RJ Taylor:

Sure. What Pattern 89 is is a software platform that’s built on a data co-op. And what that means is that over 2,000 brands have connected their accounts, Facebook, Instagram, Google, other digital marketing platforms, to Pattern 89. We analyze it with 49,000 different creative dimensions to discover what’s trending up and what’s trending down. And we’re over 95% accurate at predicting whether a piece of creative is going to work. So a good example is, let’s say we’ve got an upcoming campaign for a fitness brand. Should I use a man or a woman in that fitness brand? Should I use a man with a blue shirt on or a red shirt? Should he have a beard? Should he have glasses? Should he be outside or working out inside? All these are creative questions that we might A/B test ourselves into, or you can use artificial intelligence to predict what’s going to work. And that’s ultimately what we do, is predict creative performance. We’re over 95% accurate at doing it and getting really granular as well as large, validating the creative ideas that are actually going to drive performance.

Drew McClellan:

So if I had an account, if I have an agency, I have an account, and I am going to be placing ads on behalf of a client, I can put creative up there and you’re going to say to me, “No, Drew, not the glasses,” right?

RJ Taylor:

Yeah, there’s two ways to think about it. One is, when you’re going to pitch an idea to a client … Actually, sorry, three different ways. One, when we connect an account up front you can say, “Hey, here’s what’s working, here’s what’s not working,” for your brand. And then use that to, number two, create a creative brief that is building on the things that are working for you. And so you can actually go into the creative session with a customer to propose, “Hey, you need more pictures of men with beards and glasses, plaid shirts, doing yoga outdoors in front of a mountain. And here’s the data underneath that.”

Drew McClellan:

Yeah, I don’t think that ad would work. I’m just saying, I’m just saying. But anyway, the plaid yoga thing, you lost me there. But okay.

RJ Taylor:

Yeah, I don’t know, it just …

Drew McClellan:

I’m with you. So when I plug my brand in, your software then goes and looks at what we’ve already done that’s been working and not working. So you’re basically going to say, “Hey, this ad underperformed this ad, here are the elements that seem to be getting you more attention, more clicks, more whatever,” right?

RJ Taylor:

Yes.

Drew McClellan:

Okay.

RJ Taylor:

Yeah, and the thing it’s going to do is benchmark you versus yourself, but also against the other 2,000 brands.

Drew McClellan:

Okay.

RJ Taylor:

So we’re going to identify patterns and outliers there to say, “Hey, here’s some ways to differentiate yourself,” as well as, “Here are the things that are in line with what’s working.”

Drew McClellan:

Okay, so I interrupted you with my plaid yoga guy. So number one is, you’re going to be able to look back and tell me what was working for us. Number two, you’re going to help us build a creative brief and suggest what we should do. And then I stopped you before you told us number three.

RJ Taylor:

Yeah, well, number three is, that creative brief will then generate creatives. And let’s say you come up with 20 candidate creatives or even 10, we can then tell you which creative to put into which format in front of which audience and identify what is it about the creative that’s going to drive performance. So rather than you just launching 10 or 20, or even if it’s two, you can launch the right one.

Drew McClellan:

So I can hear the creative directors who are listening to us …

RJ Taylor:

Oh, yeah.

Drew McClellan:

… screaming into their whiskey glass right now saying, “You’re destroying the creative process, the …” Et cetera, et cetera. I’m sure you’ve heard it before, so what is your response to that concern that using a tool like this or using AI at all, because this isn’t just specific to your tool, is watering down the creative process?

RJ Taylor:

Hm, yeah. Well, you know, it’s a fair concern, and I, just again for background, I have a masters in creative writing, so I really love the creative process and creativity. So I really come at this from the creative’s perspective. With that in mind, every creative really cares about ideas. And they want to figure out what idea is actually going to drive performance, and I describe Pattern 89 as the same as a recipe that you might give to three different Michelin starred chefs. And those Michelin starred chefs would create three different things from a basket of goods. I’m thinking of Chopped, that TV show, if you’ve ever watched it, but it’s true. Three different creative directors are going to come up with three different ideas. But what we’re going to do is use AI and data to tell you, “Here’s the recipes or the ingredients that you should work within.”

RJ Taylor:

So if I could tell you the color green is trending up, then you might want to use the color green versus blue. And humans, we’re just going to pick whatever we like, or maybe what the customer wants, instead of what’s actually driving performance. So if I can enhance those creative ideas, validate them with data, or a lot of times in the agency world we’re arguing with our customer about what to do. And it’s their opinion and the customer’s always right and they’re paying the bill. But what if you could say, “Hey, it’s not just RJ’s ideas as he’s sipping whiskey.” Instead it’s RJ’s ideas backed on predictive data that’s over 95% accurate. “Here’s why you should do it.” And we hear from our customers, “We want to differentiate from our competitors.” But then we come up with something different and then the customer’s like, “Wait, this doesn’t look like an ad for yoga pants.”

RJ Taylor:

“Well, did you want to differentiate or did you want to look like everybody else?” And you pay me to come with new ideas, now we can validate those ideas with data. So that’s how I get there. I’ll tell you, I have been, in the early days of Pattern 89 about 3 1/2 years ago, I made the mistake of walking into the creative directors’ world and trying to say that creative is going to be taken over by machines, and I learned that lesson early. And so we have to-

Drew McClellan:

Yeah, I bet that was painful.

RJ Taylor:

It was, it was.

Drew McClellan:

Yeah.

RJ Taylor:

But it’s good. So we just have to align ourselves with creative directors. Because there’s the creative team and the performance team, and both of them are pointing at each other saying, “The creative’s not good,” and the performance team’s saying, “We need better creative,” or whatever the case is. So we’re the data in the middle that can hopefully have the creative people do more creative ideas and express those ideas in a way that are actually performant.

Drew McClellan:

And I’m assuming that the AI, and again, your tool or another tool, the AI’s not just about visual things, but it would be about word choice and things like that, headline length and all kinds of stuff like that, right?

RJ Taylor:

Yeah, exactly. And it’s across images as well as video. So got a creative idea, how long should the video be? Should it be six seconds, 10 seconds, should it be long form? What copy should you pair with it? Should you use emojis, should you not use emojis? Special characters, tonality, those types of things. So we’re not going to tell you what to write or what to create, but the recipe around that is exactly what we can fulfill as creative people, and then performance of course matters. So it’s a little bit of both.

Drew McClellan:

Yeah. You know, I came up through the ranks as a copywriter too, but for me I never lost sight of the fact that, and I talk to agency owners about this all the time, it’s not our job in the creative department of an ad agency or a marketing agency, it’s not our job to create the Mona Lisa. It’s our job to create creative that drives whatever action we’re trying to drive. And so if I want to write a mystery novel, I can do that on the weekend. But my job Monday through Friday is to use words to create action and connection. And so if a tool could say to me, “Boy, if you use the word ‘pal’ versus the word ‘friend,’” that’s helping me do my job.

RJ Taylor:

Yeah. Yes, it’s true. It’s also scary and there’s this human thing that you’re describing there. And I really like this phrase that I’ve used a number of times, beauty is in the eye of the algorithm. You’re just saying, we want to create the Mona Lisa. It’s like, well, you don’t want to create the Mona Lisa like a Facebook ad. What would be the point?

Drew McClellan:

Right.

RJ Taylor:

So instead you want to create something that drives performance that also speaks to your brand. And so if you could do both, then that’s great. And it’s the Mona Lisa? Well then, you deserve a gold star, because that would be great. And the other component here is, a beautiful thing might resonate on a billboard or a beautiful thing might stand out maybe in a television commercial on the Super Bowl, and those are different mediums than an ad on Facebook or an ad on Google or a Pinterest ad or Snapchat or something. So you have to know the medium. And I don’t know, it’s tricky to balance those. But that’s a good point.

Drew McClellan:

Yeah. So agencies are challenged, small to mid-size agencies are challenged to figure out, they know AI is out there and there’s conferences about it, there’s things they could read about it, but they’re trying to figure out how to actually use it. And their clients aren’t Budweiser or Coca-Cola. Their clients are the local butcher, baker, and candlestick maker, or a regional bank. So how are you seeing, stepping away from your own tool for a second, how are you seeing small to mid-sized agencies leveraging AI? And is it your opinion that it’s a need or a want as we walk into 2021?

RJ Taylor:

Yeah, well, it’s a tough question to answer. I think it’s a need to have. And I’ll tell you what, it’s not a need to have right now, it’s a need to have for three years from now. And I think agencies of all sizes are going to be using artificial intelligence and it’s going to be something that everybody expects.

Drew McClellan:

Yeah.

RJ Taylor:

It’s kind of like a car without cruise control, or without driver assist, like the rear backup thing. You get into my 10-year-old car, and literally my wife has run into a few different things behind her because our other car, which is our family car, has the rear view thing. So it’s like we just expect it to be there. And so I think that a lot of especially small or mid-sized agencies as well as some smaller brands, they’re like, “Oh, we don’t need to have it right now,” because it feels a little bit luxurious or whatever. But if you’re not investing in AI, I don’t think that they’re going to be in business in three years. Or your competitors are all going to have it.

RJ Taylor:

And the other lesson I learned in my career, and I’m not that old but I’m 42 now, but in 2007 … So I ran mobile at Exact Target and Salesforce, all the mobile stuff. And in 2007 I was third in line to buy an iPhone when it first came out. Third in line in Indianapolis, where I live. And the news cameras showed me on the TV. I was pretty embarrassed about that. But in 2007 12% of Americans had a device with Internet, 12%.

Drew McClellan:

That’s hard to imagine.

RJ Taylor:

It is hard to imagine.

Drew McClellan:

Right.

RJ Taylor:

And back in 2007 we would say things like, I’m sure an agency would be like, “Well, what’s your mobile strategy? Tell me about your mobile website.” And we would talk about WAP pages and put a little “M” in front of everything. “Well, we have m-commerce.” And now if you ask a brand, you say, “Well, tell me, what’s your mobile strategy?” They’ll be like, “Do you mean my marketing strategy?”

Drew McClellan:

Yeah, you mean like my strategy? Yeah, yeah.

RJ Taylor:

Yeah, it’s the same deal. AI’s going to be everything. Right now we’re like, “Well, what’s your AI strategy?” There are AI conferences, and in three to five years there’s not going to be that. It’s just going to be everywhere. It’s going to be an enabling technology that is expected. And you can look at things like, Mobile Marketer just integrated their … They just got bought. There’s no more Mobile Marketer any more because there isn’t a mobile marketer that they’re speaking to. So it’s the same thing. So I don’t know that it’s need to have right now, because we’re still in early days. But it certainly is need to have in three years. If you’re not investing in it now, then I think you’re going to be left behind by your competitors who are investing today.

Drew McClellan:

Well, I think now is the time that it’s okay to experiment with it and to acknowledge to your clients that you’re experimenting with it. And so it’s time for you to play right now. But as you say, in three, four, five years everybody’s going to expect us to be proficient at it and have the tool set and all of that. So now is the time to start dabbling, right?

RJ Taylor:

Yeah, exactly right.

Drew McClellan:

So I’m guessing that you see a lot. When you’re looking at the data sets in, what did you say, 94,000 creative factors or something?

RJ Taylor:

Oh yeah, 49,000, yeah.

Drew McClellan:

Yeah, right, which I’m going to ask you to recite in a little bit. But anyway, so I’m sure that as you’re looking at the algorithms and the data, you’re seeing some really macro patterns about what’s working and what’s not working and all that. So I want to ask you about that, but first we’re going to take a quick break, and then we’ll come back and I want to hear about some of the things that you’re seeing, that if we had our brands plugged into your software we’d be seeing too. But because you’re on the podcast, you can just tell us. So we’re going to do that right after we come back from a quick break.

Drew McClellan:

When it comes to conducting a client satisfaction survey, your agency has three choices. The first one is, adopt a “don’t ask, don’t tell” policy and just roll the dice. Your second option is to do the study in-house, and the third option is to use a third party to conduct your client satisfaction survey. If you decide that you’re ready to invest in protecting your client relationships and improving your win and keep ratios, we believe there are some benefits of using AMI as your third party research partner. Number one, we know emphatically that your clients will tell us things that they just won’t tell you. The reality is they’re going to speak more freely if they’re not talking to you directly. They don’t want to hurt your feelings, and they don’t want to get into a big conversation about it. So a third party is a safe place for them to share their real feedback.

Drew McClellan:

The second is that at AMI we don’t have a bias about any particular clients. We don’t know if you like them, don’t like them, if they’re a pain, if they’re your favorite. And so because we understand the agency business but we don’t come into those conversations with any preconceived notions, we can absolutely give you unbiased and unfiltered information based on what your clients tell us. And you know what? We know agency clients. We can hear what they’re saying and we know which threads to pull on as we’re talking to them to get more information for you and more insight. Your clients will be comfortable talking to us because we speak their language. If you’re interested in having AMI do your customer satisfaction survey, head over to agencymanagementinstitute.com and look under the “how we help” section of the website to learn more. All right, let’s get back to the show.

Drew McClellan:

All right, we are back, and RJ and I have been talking about artificial intelligence and the importance of that for our agencies, and that now is the time to experiment. But right before the break it occurred to me that he’s seeing all of these data points. And so in terms of what’s working right now, granted that’s a window of time, so this is not what’s going to work for all of 2021, but I’m curious, RJ, what are you seeing that is maybe amongst the data sets or the 49,000 creative characteristics or features, what are you seeing that’s surprising you right now? And just FYI, we’re recording this at the end of January of 2021. So just if somebody’s listening to this in 2025.

RJ Taylor:

Yeah.

Drew McClellan:

Probably do not follow this advice.

RJ Taylor:

Yeah, well, because we have access to such a big amount of data, we research and then publish creative forecasts on a monthly basis. And this creative forecast report on this, this is all available through our site. But the first thing is, the creative life cycle is changing.

Drew McClellan:

So talk to us about that.

RJ Taylor:

So average creative life cycle of a piece of creative is around 36 days.

Drew McClellan:

So when you’re talking about a creative life cycle, what you’re talking about is the period of time that it takes the ad to saturate the market, right?

RJ Taylor:

Right, right, and then fall off in performance.

Drew McClellan:

Okay.

RJ Taylor:

And so it’s about 36 days.

Drew McClellan:

And so what is the pattern that takes in those 36 days?

RJ Taylor:

Well, launch and then you see a spike in performance, and then it typically levels off, and then once it has about four days of performance that are under your average, we consider it the performance ending, because it then just trails off. And then you’re below average, and then it’s ultimately taking your performance down. So 36 days from start to that point.

Drew McClellan:

And you said it’s changing. So is that longer or shorter than what it used to be?

RJ Taylor:

Well, the first thing is that 10.4 days is what top performers are seeing. So they’re actually actively cycling through creative, so it’s more of like a week and a half cycle for those top performers, versus the average is 36 days. So it varies by industry, and so we track 26 different industries. So industries like sports and entertainment, that’s a lot faster, versus automotive or education are in the two-month range. So that’s one. It’s getting longer for some industries, it’s getting shorter for others. In retail it’s really, really going quickly, whereas in automotive it’s actually getting a little bit longer. So understanding how frequently you need to refresh your creative is, I think it’s critical. Because as you’re proposing campaign schedules to a brand or trying to pitch them on different ideas, understanding how frequently your creative lives.

RJ Taylor:

And again, you have to think about algorithms. Algorithms are caching the content and de-promoting content that they see multiple times. So freshness of creative really matters, and it matters by industry.

Drew McClellan:

But 10.4 days, holy buckets.

RJ Taylor:

Yeah, isn’t that crazy?

Drew McClellan:

Yeah.

RJ Taylor:

Those teams are typically larger enterprises in the retail space, where they’re just popping creative out because they’re trying to stop the thumb scroll. They’re trying to get content up and in people’s faces with weekly deals or with promotions every weekend.

Drew McClellan:

But part of this data that is valuable for us is, if we are proposing a campaign or a year-long retainer or something with a client, we have to know or at least be able to talk with the client about how often we should be refreshing that creative. Because obviously that’s man-hours and that’s going to change the price.

RJ Taylor:

Yes. Yeah, well, and then there’s just iterations too. You don’t have to come up with a whole new look and feel for everything, but maybe you have the same models with 10 different treatments or 10 different poses or whatever the case is. So there’s iteration on theme, versus just having … I think the idea of evergreen creative is really going away. I wish evergreen creative worked, but with iOS 14 I think that some of the evergreen creative we’re using in retargeting and all the privacy initiatives that are coming out, I think that that is going to change in a big, big, big way.

Drew McClellan:

So talk just a little bit more about that. What do you think is coming down the pike in terms of what iOS 14 is giving us clues to, the privacy issues, the cookie thing, all of that?

RJ Taylor:

Well, I think it is about privacy. And consumers want privacy, but counterintuitively, consumers also want relevance.

Drew McClellan:

Right.

RJ Taylor:

So they want relevance but they also want privacy. And with iOS 14 and other privacy initiatives, that’s going to be a step harder for us to get to that relevancy. Because consumers are going to have to double-opt in or opt in many times by clicking and allowing permission for people like Facebook to track us. And in our heads we think, “Ooh, that doesn’t sound good,” except for when I stop getting relevant communications, you know? So as some of these new changes roll out, I anticipate that consumers will be confused by them and then opt out, or actually it’s going to require an opt-in, so they’re not going to opt in, they’re just going to dismiss notifications allowing for their data to be shared because they don’t understand it or they feel like … They’ve watched the Social Network movie or whatever.

Drew McClellan:

Right.

RJ Taylor:

And so what’ll happen is our creative is going to be even more important, because we’ve got to get relevant, engaging, interesting, thumb-stopping content that people want to click on without knowing everything about them. So that’s a little bit trickier, right? So how do I do that? It’s not through audience targeting any more, through some sort of automated feed that pulls in all the products that you used to look at.

Drew McClellan:

Right.

RJ Taylor:

“Are you still interested in this lamp, RJ?” “Well, yes, I am,” and so it follows me around the Internet until I have bought it. I haven’t bought it yet, but it’s going to follow me around. That’s not going to happen in the future. Instead, they’re going to have to get really creative in the way that they’re showing content to me so that I continue to want to engage with it.

Drew McClellan:

Well, it also speaks to, stepping away from paid ads for a minute, it also speaks to the importance of that brand connection that people have. I’m a Disney guy, so I want Disney to follow me everywhere. I want them to know what attraction I get on, I want them to know how long I’m willing to stand in line, I want them to know how many Mickey ice cream bars I consume in a week period of time. I want them to know all of it so they can just give me more of exactly what I want. But there are probably very few brands that most consumers feel that strongly about, right, that they would give that sort of untethered access?

RJ Taylor:

Yes. Yeah, you’re totally right. So we’re going to instead have to figure out what’s the latest trend in creative? So it’s January right now, and normally people are doing Super Bowl ads or Super Bowl promotions, or Valentine’s Day or President’s Day or whatever.

Drew McClellan:

Right.

RJ Taylor:

How do I make those creatives resonate with me in a COVID or post-COVID world? We track creative trends on a monthly … We track them on a daily basis, but we can’t really, either the system updates on a daily basis but our work’s done on a monthly basis. And last March the presence of people washing hands and washing objects, I don’t know if you remember that, back when COVID first came out everyone was like, “Don’t touch your face and wash your hands all the time.” So what happened is, on the Internet 600% increase in images of people washing things, washing hands, water falling, faucets, that kind of stuff. Also a giant decline in group shots, so where there was always a diverse group of people all hugging and happy at some sort of festival or whatever, nope. And now we’ve just gone to selfies and everyone’s alone.

Drew McClellan:

Right.

RJ Taylor:

And a 38% increase in electronic devices in … Excuse me, that’s not true, 30% of ads now have an electronic device in them. They have a phone, an iPad, a computer, a gaming set. I’m trying to come up with a clever way to describe it, but the whole idea of looking down is now potent, because all of us are looking down in these ads because 38% of them have some sort of device. There’s a loneliness and an isolation to it all. And so how do we tell this uplifting story or aspirational story but also speak to what everyone’s going through? From a creative standpoint that’s what we’re trying to understand. Because it has to do with cost.

Drew McClellan:

Right.

RJ Taylor:

So because we’ve seen increases in certain areas, supply and demand dictate that things become more or less expensive. So funny things, like right now cats are causing more … Cats in your ads, everybody talks about cats on the Internet, well, those are actually more expensive in your ads. So if you have a cat in your ad you’re causing the cost to go up. Same thing with [crosstalk 00:31:41].

Drew McClellan:

So go with a llama or a goat, yeah.

RJ Taylor:

Maybe stick with the dogs for now.

Drew McClellan:

Right.

RJ Taylor:

But exercise and fitness, those are also rising in cost.

Drew McClellan:

Sure, January, right?

RJ Taylor:

Yeah, yeah, yeah. Things like reading or a lot of nature imagery is doing really well. So oceans and mountains and big wide open spaces.

Drew McClellan:

Yeah, all the places people want to go.

RJ Taylor:

Yes, yeah, yeah.

Drew McClellan:

Yeah, makes perfect sense.

RJ Taylor:

It does. And so oftentimes we would look at last year, like, “What did we do last year? Let’s hit around that.” Everybody says don’t say “new normal” any more or something, so you can’t say new normal, but whatever we were doing last year or the years before, you can’t do it this year and expect the same results. Because beauty is in the eye of the algorithm, the algorithm’s going to dictate what content gets seen, and then also our consumers need to have things that they want to click on and engage on. So we track that type of stuff to understand what’s going to drive performance. And then even things like headline characters, five to 15 characters drive the best performance. Bodies, 40 to 60 characters and including a hashtag. The lightning bolt emoji is performing well. The alarm clock emoji is not performing well right now.

Drew McClellan:

Does it ever really? I mean, really have we ever been pro-alarm clock?

RJ Taylor:

I don’t know, maybe back to school, I don’t know.

Drew McClellan:

Maybe.

RJ Taylor:

I’ll have to look at the data.

Drew McClellan:

So with everything changing so fast, how does that impact the effectiveness of things like A/B testing? Is that still relevant, or are things happening so fast that you can’t even test before something’s changed?

RJ Taylor:

Well, yeah, I think A/B tests are dead, and it’s kind of sad to say that, but I do think A/B tests are a thing of the past. And multivariate tests, same deal. I used to run Taguchi tests, which is crazy amounts of variants, in my role on the strategy team with Exact Target. Instead, we really have to learn what’s working in real time, and that’s where AI can come in. And this will turn into a sales pitch real quick, so we can back off of that. But that’s ultimately what we do have to acknowledge from a creative standpoint and from the agency’s standpoint. Once you do an A/B test and you pull the data and present to a customer, we’re talking about, let’s say best case three weeks.

Drew McClellan:

Yeah.

RJ Taylor:

The Kardashians have done something else in that three weeks. Something else nutty in the world happened. And so whatever you learned and considered a best practice is no longer either of those things. Instead you have to think of this agile mindset, and iterative and predictive is where we need to be going from a creative standpoint. So I think that A/B testing still can be used if you’re testing between two potential winners. But oftentimes we A/B test to find the winner and the loser. Instead I’d find a winner and something that’s doing better than the other winner. So you can use your data to whittle it down to two options and then launch those two options. But currently A/B testing’s often used when people agree, your client, or disagrees.

Drew McClellan:

Yeah, it’s a proof point.

RJ Taylor:

“Let’s test it.” Well, great, you’re just going to waste money, time, effort, energy, resources, and you end up at the same spot. So instead you can use data, predictive stuff to supplant the A/B test or the need for it.

Drew McClellan:

So one of the things we’ve talked about a couple times in our conversation is this idea that data can help an agency help a client over a subjective opinion or thought. And honestly, I think that happens inside agencies too, I think. I worked with a guy once many, many years ago, and he was always like, “Well, I don’t listen to the radio so we shouldn’t suggest to the client radio.” And I was like, “Well, you’re not a teenage girl so it doesn’t really matter what you listen to.” But how do we work with clients? How in this new world with all of this data, how do agencies and clients work together to rise above the noise and to actually do good work in the digital space?

RJ Taylor:

Yeah, one of our core values at Pattern 89 is transparency. And of course it’s buzzy, or maybe even cliché at this point, but I think a lot of times when working with clients we try to tell a story. And of course we need to tell a story, but I think we just go in full transparency. “Here’s what we’ve learned. This one’s a winner by 1%, this one’s a loser by 1%. This one’s 80/20.” So we can understand, “This one won and lost. How much did it win by? Let’s actually show you the data.” Versus trying to craft a narrative around what we want them to believe or that makes us look the best. I think authenticity and transparency are just being demanded across everything that we’re doing, from culturally, from … We can go into all sorts of things, but just culturally we want transparency and authenticity, and I think that that extends into relationships we have with our clients. So I do think it comes down to transparency, and, “Let’s both look at this data together. Am I looking at this in the wrong way?” And use that to build trust and then ultimately joint success.

Drew McClellan:

So if an agency is just getting ready to dip their toe in the water of AI, and honestly everyone talks about data, blah, blah, blah, blah, blah. But as you know, a lot of clients can’t find their own data, don’t know how to interpret their own data, don’t know how to harness their own data, don’t know how to share their own data with your agency. So I think there’s a lot of talk around this, but there’s a lot of confusion about how to actually begin to experiment with it, how to get it done. So assuming that your product is off the table, not that, and you’re advising a small to mid-sized agency of where to step into AI, are there some places that you’ve seen, because you’re in that space every day, where agencies can at an affordable price, because that’s part of the other challenge, is that clients don’t want to pay for it, agencies know they need to learn it. So are there some places where agencies can go to begin to play in the sandbox and begin to educate clients about the value of that sandbox?

RJ Taylor:

Yeah, well, the best one that I know is the Marketing AI Institute.

Drew McClellan:

Oh, yeah, Paul’s.

RJ Taylor:

Yeah, yeah. Paul Roetzer, he’s got all sorts of webinars and educational pieces, that kind of stuff, that you could attend with your customer. The other thing that they’ve done is they’ve come up with a questionnaire, it’s like a quiz, one of those Buzzfeed-y quizzes. This is not [inaudible 00:39:04] to them, instead it’s just easy, is what I mean, where you can say, “Hey, here, I’m trying to explore,” this or that or the other. And it’ll actually generate, “Here are some things you can do.” So it’s a great way to get educational content but then also you can take their quiz and then get guided to resources or maybe a starting point. And they actually answer this question better than I’m going to do, which is, how do I get started? And they have some rubrics that you could just take and use alongside your customer and say, “Hey, here’s the AI Institute recommendation on where we start.” So I would go there.

Drew McClellan:

Yeah, it’s impressive. So listeners, if you have not gone there, we’ll put a link to it in the show notes. If you’re an AMI member, inside the marketplace there’s a significant discount for you. You can actually get certified as of different aspects of AI. It’s a very impressive thing, and they also have a conference where they talk about AI from a marketing perspective and an agency perspective. The first year it was in July, then in 2020 of course they couldn’t have it. And as far as I know, the plan is to have it this summer, but I don’t know the dates have been announced yet. So go to the show notes and check out that link. So I agree, that’s a great resource. Where do you think we get in our own way when it comes to AI?

RJ Taylor:

My experience is the “gotcha” agencies. So we have a lot of agency customers. We have worked, maybe this is not the right thing to say on a podcast like this, but it’s the “gotcha,” meaning it’s the agency people who are looking for the ways that AI doesn’t do things perfectly. Meaning, like our platform is over 95% accurate, which means it’s not right 5% of the time. And oftentimes we’ll find some clients as well as partners will look at that 5% and say, “See, this is why AI doesn’t work.” And I think we’ve got to look at the 95%. And of course I’m going to say that because I’m trying to sell something here, but we have to look for the opportunity.

RJ Taylor:

And I think that a lot of times, especially in agency land, we’re selling humans. We are selling time and human time, and so if there’s a threat that a solution might take away from that, then I think we’re looking for the negatives. And I think instead we need to be looking at, “All right, how do we still sell time,” and we want to sell the most important time, strategic time. I’m going to sell ideas and I’m going to sell brain power, and how can I use the tools that are out there and automation, et cetera, to augment that? So that’s where I’ve seen potential agency partners go sideways, because it’s like, “Oh, I found the one thing that was wrong, and therefore I can’t trust anything, and I’m going to wait for three to five years to try this again.” So that’s my frank answer. Maybe it’s not the most glorious.

Drew McClellan:

Well, the other side of the “gotcha,” first of all, agencies shouldn’t be selling time the way we did back in the ’50s. We should be selling value and solutions and results, because that’s the only thing that our clients can merchandise up the food chain to the CEO to make sure we get to keep our jobs. So putting that aside, I think the other “gotcha” is that if agencies are using AI, if they’re in a combative relationship with their client and they’re using the “gotcha” to show their client that they were wrong, that also to me doesn’t seem to serve the agency’s purpose long-term.

RJ Taylor:

Yep. Yeah, you’re totally right, yeah.

Drew McClellan:

So given that, what you said about that we always have to be at the point where we are predicting what’s coming next, what do you think is coming next?

RJ Taylor:

Just broadly?

Drew McClellan:

Well, in the shared AI/agency space.

RJ Taylor:

Yeah, well, what I do think is coming next is, and I don’t know if you’re familiar with GPT3 or deepfakes and natural language generation solutions. I think that where we’re headed, and I think it’s a little bit further off, is where a human effectively creates a creative brief but then a machine fulfills it by creating a video or an image or even writing the text that is aligned with the brief. And I’ll tell you, that scares me a little bit.

Drew McClellan:

Yeah, I’ve seen some videos of it and it is creepy.

RJ Taylor:

Yeah, it is. It’s creepy, there’s potential for that technology to be racist, to ignore empathetic or brand related things. You mentioned Disney earlier. There is one sideways, one wrong decision by that machine can create a disaster for a brand like Disney.

Drew McClellan:

Right.

RJ Taylor:

And so we as humans need to understand potential as well as understand brand, empathy, ethics, those types of things, and make sure that we layer them on top of any technology, but especially this creation, the content that it can create. So I think that’s coming, and so understanding how it will or won’t impact or change our businesses, I think, is really important. I’m keeping my eye on that pretty intently because I’m very interested, but also I think it’ll change marketing again.

Drew McClellan:

Yeah, I think so too. All right. So one of the things, as we’re wrapping up I’m thinking about the things we’ve already talked about. So one of the things you told us was that campaign effectiveness is getting shorter and shorter and shorter. So how do you believe, given there are certainly budget constraints, there’s client tolerance constraints, how does an agency balance creating new creative all the time and meeting the client’s needs, both in terms of results but also in terms of budget and all of that? What are you seeing agencies do to find that balance?

RJ Taylor:

Yeah, well, I think iteration on theme is the easiest way to achieve all of those. And instead of getting to a creative that’s approved or the one creative, I think we come up with 10 of just literally iterations on theme, and those iterations are different enough that they look different to the algorithms that are judging and scoring them, and different enough that they’ll feel a little bit different to the consumer. But they’re same enough that you can get them all through in a single approval, whether it’s single approval by client or by legal or whatever the case is. So I think literally just a multiplier, it could be a 5X or a 10X, of iterations on themes. Small changes to wording, changing out the emoji, literally it’s just a change-up.

Drew McClellan:

So how much change has to happen for the algorithm to go, “Oh, that’s new”? So is that like the man in the flannel shirt doing yoga, the lady in the flannel shirt doing yoga, the baby in the flannel shirt doing yoga? Is it that?

RJ Taylor:

Yes, yes, it is. To be frank with you, I don’t know because I don’t know what the algorithm’s doing. And I imagine that algorithmically they’re changing algorithm all the time. But the man doing yoga and maybe five different types of plat.

Drew McClellan:

Poses, oh, sure. Different plat, right.

RJ Taylor:

Yeah, yeah.

Drew McClellan:

So it’s both word changes though, like headline or body copy, or maybe a hashtag, or as you said, an emoji. So it’s character changes, but also visually making some changes.

RJ Taylor:

Correct. Yeah, or it could just be assembling your video in different iterations. So let’s say you’ve got 30 seconds of video, you could make five unique versions of six-second videos for 30 seconds by splicing different pieces together. So your creative budget doesn’t have to be exploded, but instead it’s an editing exercise.

Drew McClellan:

But you’ve got to know that up front, right, while you’re doing the photo shoot or while you’re taking the time to write the different headlines or you’re shooting the video, so you know that you’ve got enough footage to slice and dice into different things. So again, this is about us measuring twice to cut once and be thoughtful on the front end so that we can meet the client’s needs further down the campaign.

RJ Taylor:

You got it.

Drew McClellan:

Yeah. Okay, this is awesome. RJ, this has been fascinating, as I knew it would be. So if folks want to learn more about you, they want to reach out to you, they want to learn more about Pattern 89, what’s the best way for them to do that?

RJ Taylor:

Sure thing. Well, we’re just at pattern89.com, and we publish our creative forecast on a monthly basis, and those are free and available. So that’s an easy way to learn more about Pattern 89. If you want to reach me, I’m just [email protected]

Drew McClellan:

Okay, awesome. And listeners, again, if you’re an AMI member, you want to head over to the marketplace because we’ve got a special pricing deal for you inside the marketplace with Pattern 89. So scoot over that way. I will not put a link to that in the show notes, of course, because it’s behind the firewall, just remember. So RJ, this has been great. Thank you for coming on and helping us think about this future that is both in front of us but also kind of right here, right now. So I appreciate your time and you sharing your expertise.

RJ Taylor:

Thanks for the opportunity.

Drew McClellan:

You bet. All right, guys, this wraps up another episode of Build a Better Agency. I want to remind you that we are going to be gathering together, knock on wood, for our Build a Better Agency Summit in August. Paul Roetzer, who RJ and I were talking about with the Marketing AI Institute, is going to be there. So you can hear from him, right from the horse’s mouth, he’s going to be speaking. So I would love to see you there with us. Going to be about 250 agency owners and leaders, we’ve got plenty of space so if we are still at the point in our world where we’re social distancing, we’ve got you covered. And you can read more about it by heading over to the agencymanagementinstitute.com website and looking up Build a Better Agency Summit. So we would love to see you there.

Drew McClellan:

Huge shout-out and thanks to our friends at White Label IQ, they are the sponsor of the podcast. And as always, you know that if you’re looking for somebody to help you augment your team for PPC, design, or web dev, they’re your guys. So head over to whitelabeliq.com/ami to get the special pricing that they have just for you. All right, I’ll be back next week with another guest like RJ to get you thinking a little differently about your business. I highly recommend that you go check out their forecast reports, they’re awesome. And I will see you next week, all right? Thanks for listening.

Drew McClellan:

Thanks for spending some time with us. Visit our website to learn about our workshops, owner peer groups, and download our salary and benefits survey. Be sure you also sign up for our free podcast giveaways at agencymanagementinstitute.com/podcastgiveaway.