Hello, and welcome to the Theory of Marketing Relativity Podcast. I'm Jess Burton, a marketing specialist here at Epic Marketing Consultants Corporation. Let's talk all things marketing and merchandising with our Epic team. Guest starring on this week's podcast is the co founder and chief data scientist at Trust Insights.ai. Chris Penn. Thank you so much for joining me today, Chris.
Chris:Thanks for having me.
Jess:I am very excited. I know I when we were talking about this topic of going viral. I was like, what makes a video viral? Like who can we talk to you about like the science, the the science of like marketing or whatever behind something going viral? Because I think especially with everyone being home, and everyone being on social media, and everyone being tuned into the internet, so many things have gone viral in the last year. Even what I like to refer to as mini viral, that it's such like a hot button topic sometimes.
Chris:Let's do this, let's agree to not use the word viral anymore. Especially in terms of a pandemic, which has killed over a million people. In terms of your exponential growth of content, I certainly think that, you know, there are definitely things we know about all the different algorithms that are out there. But yeah, let's I was specifically talking about SARS, co v2, or it's the disease that creates COVID-19. Let's stop using viral.
Jess:Agreed that was one of those words, like at the beginning, I'm like, I never want to hear this again, I was like, we need to come up with a better term for it.
Chris:Exactly. You know, when it comes to understanding how things get seen how things can follow this, this, a bunch of different angles to it, but it fundamentally comes down to two major things. It is the algorithms that are involved in accelerating, you know, something being seen, which are self fulfilling prophecies, you know, as something gets seen, it helps it get seen more, and so on and so forth. So a big question is, how do you get that started? And that comes through your distribution networks that comes from essentially getting it seen by people who can help distribute it much faster? Where do you want to start? Which one is more interesting to you just to start in?
Jess:I don't know. I feel like the big one for me, like so for me, I noticed that things go crazy with um, like video platforms, things, you know, anything video related, whether it's TikTok, or YouTube, or even. It's funny, a couple of my friends don't have TikTok, but they see TikTok videos on Instagram and Twitter. So like, Oh, yeah, I saw that TikTok, and I'm like, but you don't have a TikTok. Um, so I think like vi eo platforms are a big one.
Chris:So we don't know what's inside of TikTok's algorithm. They developers have not disclosed any information about that. What we do know, however, we do know very well, because Google has published academic papers about about what is inside of the YouTube algorithm for how it chooses things to recommend. So there's a bunch of things that go into it, there was a 2016 paper that you can actually find over on Google's website, that shows you the machine learning model that makes up the recommendations. So there's, there's I guess, four or five major chunks. The first is, there is an embedding, which is a fancy way of saying, looking at your browsing history, your viewing history. So for any user, when you sign into YouTube, a good chunk of the recommendations come from things like your your previous history. There's a second vector of what are called search tokens. These are the things you type into YouTube, whether they be you know, in the search bar, like you know, you're looking for past episodes of Supergirl, or how to make a French omelet. All these things those are those are categories and tags as well. There's a geographic component to the recommendation engine, obviously, it's not going to show you things that are not geography, your geography available, but also looking at what users around you are talking about. There are things like age, the inferred or actual age, if it's in your profile, your gender, things like that, the language that you're searching for. And then inside of the videos themselves, Google takes a look at the basics, title, the language, the videos in the description, the tags that you use, and then a critical one that almost everybody forgets about is that Google looks at closed captions. And it looks the language that you use inside so if you were to take this episode of this podcast, Load to YouTube and submit a closed captions file with it. The words and phrases that we use in this will help people find this video, even if you don't put it in, say the the title or description. So if you were to say, you know, something ridiculous like throat wobbler mangrove, which comes from a Monty Python sketch, and somebody was searching for that, if we provided that in the closed captions exactly, you know, as as said, Google would know, to help to serve as part of that as part of the search. So if we're talking about helping our content, get exponential growth, we have to work with what we know the algorithm is looking for, we know the algorithm is looking for all these different components. So how do we influence these? Well, obviously, with language embeddings, that's pretty straightforward, right? That is making sure that we're providing closed captions, we fill out all the fields that you could possibly fill out when you're when you're submitting it. But another big part is that, that browse history that embedded video watches you are for at least a period of time until you don't engage with it. If you follow a topic or a channel, or even just browse a channels, content, more content from that channel will be recommended to you. So one of the things you have to do is you have to seed The, the your audience, essentially with content, that they can establish that your channel is something they should be seeing and their recommendations, you will probably not see, you know, recommendations from channels you've never visited before, right? Because Google says I don't see any logical reason to show this unless it matches those search tokens. Now, if you've done a good job of seeding your audience, then when they search for something that you are searching about, your channel will show up higher in the results because you not only have the match on the search token, you have the match on the history as well. So how do you do that? The easiest way to do that, honestly, is to create content and distribute it and and share the heck out of it, maybe even put some ad dollars behind it to create that history. So I'll give you an example. In the last couple of newsletters I published, I've been featuring this video on how to understand UTM tracking codes part of Google Analytics, front and center in the newsletter front and center, you know, this content stuff. And what that does, is it for anybody who clicks through and watches at least you know, 30 seconds of it embeds my channel in my video in their YouTube history. And now, anytime they search for like, analytics, if they do that on YouTube, I have a slightly higher propensity to show up. So it helps me get discovered based on YouTube's algorithm. So that's the technical side. Does that make sense?
Jess:Absolutely, I, um, oh my gosh, I guess it was like back towards the beginning of the pandemic, I made a TikTok that basically was saying, like, ladies, if you want to hit to your boyfriend to propose to you, you need to start sending him videos, and images on his social media platforms have engaged engagement rings of rings stores of jewelry stores, because eventually the algorithm will take over if he's opening them, and he'll start getting served ads and recommendations for different things. And I think that is like a part a huge part of recognition that people don't take into account is that like user history, and then it kind of like, on each platform bookmarks you to like certain types ofcontent. So when YouTube sees that you just watched a hour long video of you know, hiking through, you know, the Appalachian Mountains, they know that they now know that okay, you're in the hiking. So our suggested videos are all going to be about either the Appalachian Mountains or hiking or hiking gear reviews, or all that kind of stuff. It's that, that and I watched your UTM tracking code video, because Nancy was explaining it to me. And then she was like, Chris does a great job of explaining it, you should watch it. And I was like, Oh my gosh, this makes so much sense now, um, and then I continued to get served articles for UTM tracking, tracking. Um, but yeah, I definitely think that it is super important to understand how all of that works. I think it ties in nicely with, you know, knowing a little bit about SEO, knowing a little bit about like cross channel consistency. So I always tell people, because I'm on like, the social media and the things. I always tell people, you know, know where your audiences know that they're on LinkedIn. But the chances are, they're probably also on Facebook. You know, your audience, it's on Instagram might also be on Twitter. So you should cross promote it and find them so that you're meeting them where they are, and they can get your content and they can see it and there's more of a chance that they're seeing it multiple times. And now all of a sudden you've got 10,000 video views in a week or whatever. I just made that number up.
Chris:Yep. So that's, that's one part. The second part is, you know, leveraging the distribution networks. Again, this is something where you can see pretty plainly how well it works on YouTube. The number of successful channels that do cross channel collaborations is was like 80 90%. Because people understand that their the who their crossover audiences are. So for example, if you like the cover band first 211 you're going to see them do collaborations with Violet Orlandi with Cole Rowland's, you're gonna Halocene, Lauren Babic. So you're gonna see all these people doing videos with each other, because they understand their audiences appears, and that the people who like first film probably will also like Violet Orlandi. This is just basic, just basics of distribution, figure out who the Who else has your audience. And especially if it's non competitive, we even if it is competitive, it's it, there's still no blended interest, maybe you'll work out a barter arrangement with channels of similar size to trade audiences to say like, you know, you appear on this show, you know, I'll appear on that show. And you cross promote. Podcasting has been doing the same thing for what, 15 years now, your podcasters doing spots on each other shows guests guest appearing on each other shows and things like that. And a lot of that works really well. If we're talking about taking things to exponential growth you want you can do it one of two ways you can either, you know, try and hope to get placed on channels much larger than yours, which will have some level of effect, but very difficult to do. Or you can go out and get 10, you know, do 10 collaborations, the channels of equal size where you can see like, Okay, if you got 1000 subscribers, and I got 1000 subscribers, let's let you know, let's cross promote. And then that way, you can get to larger numbers, but you will also again, because of the nature of watch histories and search tokens in the algorithm, you're you're probably going to be better off, you're chasing down 10 peer channels, rather than trying to punch up to a person that's got a million subscribers, because, A, they're getting 500 pitches a day, and they all suck. And B they may not have as refined an audience, you if you want people to like and share and and comment and essentially drive that engagement algorithm, you've got to get that engagement. So a channel that has 1000 subscribers, but 100 of those subscribers engaged with your stuff is way better than a channel has a million subscribers and the same 100 people engage because the percentage of engagements going to be much lower. So there's a big chunk of that, when it comes to that distribution site. That's, that's part two, I would say of how you get something to have exponential growth.
Jess:I am so glad you mentioned like percentage of engagement. Um, I noticed and I have conversations like this with our clients all the time about how it's not necessarily the size of your audience. It's about the engagement percentage of your audience. And that's I always like delve into an explanation about micro influencers, it's not so much that you want to have 10,000 followers. But if you have 1000 followers, who 50 or 60 of them, every time you post content are commenting, and liking and sharing and like building on that community, you are doing a phenomenal job versus if you had 10,000 people and like you said those same 60 people are commenting, well, that's great. The way, like it's not seen the same, it's not as valued. Um, I think that is a very big important part of getting your content out there.
Chris:Yep. And then the third component is temporal is time. Specifically, one of the things that you can do is look at, and this is this is, you know, for more things are education based, as opposed to entertainment based, is looking at search trends. And then using these statistical forecasting software of your choice to project forward, what is the likely trend going to be around that term? So for example, right now, we're in the midst of doing some ads for my data science for marketers course. And I have things like business analytics course data science course, all these different search terms, fed them into one of our SEO tools to get search volumes, and pull that match that with our trend data, and then did a forecast. When is each search term going to trend? And then knowing when it's going to peak in the next, say, 12 weeks? I can say, we did this last week I sent to my ad agency, okay, this week right now, the term business analytics course is going to be the most searched trend, I need you to increase my bid prices by 20%, just this week for that term. And the week after that, it's going to be this time of the week, I just made this term. So I changed my bid prices on this algorithm. And you see the results right inside of Google ads, you can see like yo, this term is outperforming Well yes, because I know based on statistical analysis, when people are going to be most interested in that term. So a big part of you know, exponential growth is not only having the right content, but also having it at the right time when the audience is most interested in it right now, for example, people are not terribly interested in a holiday gift guide, right? A come October 1 people be very interested in a holiday gift guide. That's a very simplified explanation. But we don't as marketers think, think enough about when the audience wants something, usually marketers are too busy worried about their own numbers, I gotta make my numbers for this quarter, I gotta get my you know, my lead numbers up this quarter. And you start trying to force things on the audience that they don't care about. It's, it's not the right time for it. So you have to take a big step back. And you know, do some analysis, do some qualitative research and stuff to figure out, oh, this is actually what my audience wants right now. And in the near future? What do I have to offer them, and if you don't have anything to offer them, you need to go and do your homework. Because if you don't, you're just going to be looking at, you know, continued poor performance, it astonishes me that, you know, marketing is the only profession where you can fail 99.2% of the time, you know, and getting a click through rate of 0.8%. Like, you failed 99.2% time, and that's a success in marketing. Like, how pathetic is that? Like, my video got 100 views? That's a success? Well, no. When, you know, when when any of these other entertainment channels publish a video, they get 10,000 100,000 views immediately. Why? Because they're providing something that the audience actually wants. And we're not as marketers, we're just not?
Jess:Absolutely, it's definitely about the a lot of it is value based, a lot of it is value to the consumers consuming the consumer. I know, we I talked about all the time, we, um, a couple of us here are photographers, and that's one of those kind of like, side hustles, whatever, that it's all about the value you provide, because how can you put a number on wedding photography? How can you put a number on graduation photos down? Well, it's hard, it's and you have to create a product that people are going to be willing to pay for. It's the same concept, like people will watch things that they find extreme value in. And it's also I know, at least with, you know, doing stuff on social media, you have to catch them in the first three seconds. Not even I think it's like shorter than that now. But you have to catch them. Right? As soon as the video starts, or else you've lost them. Because everybody's attention span is that of a goldfish now.
Chris:No, I disagree with that. Yeah, we no such thing as binge watching. If we had short attention spans,
Jess:that is true.
Chris:That is very true. Not sit down and watch an entire season of This Is Us, right? Or any of these things? If you had no attention span, we have no attention span for crap. Yes. Right. We have great attention spans and things that make us feel stuff. And ultimately, that's what it comes down to is does what we're doing have any emotional appeal, think about all the things that do have exponential growth, all those TikTok videos and stuff, what do they make you do most of the time, they make you laugh, right? I don't want to make you shake your head or make you do something that inspires that emotion. I saw a video earlier of this guy, you know, I tried to be to give people the benefit of the doubt. But I'm watching this guy's video from from last week from the the gasoline shortages, truck trying to pour gasoline from the pump into a long tube sock. Think about that for you. There's another one of a woman pouring gasoline into a one of those mesh laundry baskets, like what's wrong with you, even if you're like, this was normal hips, like that's full of holes, it's not gonna hold liquid. And so, but you even just talking about this, you can see that it inspires emotion. And as you know, one of the reasons why, for example, we have such a problem with disinformation and fake news and stuff like that is because it inspires emotions, right? We'd like to think we get really angry turns out making people angry and afraid, are like guaranteed keys to engagement, if you want great engagement, just make people angry and afraid all the time. It works great. It's not good for people, it's probably not good for your brand, it arguably makes the world a much worse place. But if you've just got to hit those numbers, I mean, one of the things that when you look at, for example, Facebook's mission statement, they, they're their core values, be bold, move fast break things, you know, etc. And no point does mission statement is ever say make the world a better place. And so their algorithms do exactly what their culture dictates. And so if you want to, if you want that outperforming material, they want the exponential growth, you go with what works and is what is allowable on that platform. And the only thing that restrains you is your own sense of ethics and morals. So give that some thought. I know there are certainly plenty of brands that that do very interesting things and funny things too. That create that engagement. If you want to see your great social media engagement things, go look at like Arby's, or Wendy's and what they're doing on social media, right? They just, it's hilarious. It's fun, sometimes a little mean spirited. But they do it in such a way that it creates that emotional reaction and then look at your own content. You know, what the things I always ask people is, as a simple test, if, if you did not work for the company you work for, would you follow your own content? Would you enjoy your own content? Most be like, No, I don't want to read my own newsletter. You know, if you were watched, you know, this podcast, like if you if you and I were not working for our companies, would we want to watch this content? Would we want to watch ourselves. Another good test is ask your significant other if you have one. Hey, you know, if I wasn't in this thing, would you watch this? And you know, the significance is I wouldn't watch I'm not gonna watch it with you into like, Okay, well, you know, it's not emotionally engaging enough content to compel the average person to to view it, as opposed to like, you look at Vic Krishnas TikTok on how vaccines work on that the mRNA vaccine do phenomenal. It's like 30 seconds long at most. But not only did it do a great job of communicating accurately the science behind an mRNA vaccine, it was funny.
Jess:His was the one with the forks, wasn't it? .
Chris:Yes, exactly. And that's, again, it's one of those things that you remember it because it evokes strong emotion. And there's no getting around that there's no shortcut to that there's no algorithm that's going to do that for you, it comes down to understanding your audience and knowing what it is it's going to create a small, strong emotional reaction with them. There's a reason why, you know, for example, other than YouTube, Pornhub is the pilot top one of the top video sites online, it's not because there's educational videos, you know, about, you know, mRNA vaccines on there. But it all is about evoking and inciting an emotion. And it works, right, if it didn't work, would be at the bottom of the pile with a bunch of other, you know, B to C sites that nobody pays attention to. We have to acknowledge the reality that if we want our content to outperform it has to be emotional.
Jess:So aside from video, because I know video is like king and Nancy preaches that does all the time that like if you can do video, you can do a lot of stuff. Aside from video, do you think there are any other kinds of content that garner attention? Maybe not as fast but faster than others?
Chris:There's tons of things that garner attention. And video is just a format, right? podcast episodes. You know, every time for example, like Ilan musk shows up on some random podcast and stuff like that, and just goes completely off the handle, right? It's engaging content, people want to hear what this person has to say, there are no shortage of papers that have done well, there. In this whole pandemic, there are tons of academic papers that have had exponential growth because they did something that was useful. In particular, they helped assuage some people's fears about how does the virus work? Or how does, how do these vaccines work or the test results from the vaccines, you know, I pay very close attention to the pediatric trials that are going on, because I want to know how this vaccine is going to react in my kids, right? Is that compelling information, though, it's a little on the dry side. But I'm not going to read the hell out of it and share it like crazy. Yes, because it's something that ultimately people are going to care about. It's going to has that emotional hook. If a human being can consume it, it can be engaging, right? It doesn't matter what it is, where it is, what format, if, if it provides that emotional impact and that value, people will consume it, and they there's no shortage of PDFs that have done well, of interactives that have done well. You know, the new york times they're they're scrolling interactives are brilliant, and they're not videos. You just scroll them the mouse or swipe your finger up and down the screen. You see these visualizations change. I mean, there's even just like short audio snippets and things. As much as I personally don't enjoy the platform. The social audio platforms or clubhouse have done well, depending on who the room host is and stuff. So that Yeah, there's, it doesn't matter what the format is, it depends on on, are you creating something that people would find emotionally engaging?
Jess:Absolutely, I think it's very important to put audience value first, and everything else can kind of come afterwards.
Chris:Something really simple. Like this is just an old Calvin and Hobbes book. Yeah, any clip out of here. Any cartoon out of here and put it on a simple format, it's gonna do really well. Why because it's entertaining, right? But it's entertaining in a way that is, it matches two or three different emotions that Same time because there's a lot of nostalgia along with humor in there. And again, is that a format that is super tactical? No, it's a it's a hand drawn cartoon. But does it work? You bet it does.
Jess:Absolutely. Chris, thank you so much for joining me. If people want to find you, I'm sure you are all over the internet, where can they find you sign up for your newsletter, where's the best place to look?
Chris:So there's two places first place, go to trustinsights.ai that's the company and our Wednesday newsletter is a lot of fun. There's usually there's fresh data in every single issue. This past week, we're looking at TSA checkpoints, data. So you know how many people went through a TSA checkpoint that day? It's published by the US government. And I think it's one of the best measures of the economic recovery. Because it's not speculative. You can't you know, you don't bet on how people are going to be going through the gates like no, this is the number of feet that walked through. And you can see pretty clearly we're getting close to, you know, back to business as usual. And the second place is my personal website, christopherspenn.com And I have a new book if anyone wants what's not new, it's it's now the third edition of AI for marketers, if you want to learn more about some of the technical stuff we're talking about today, but doesn't require a technical background. Go to AIformarketersbook.com.
Jess:Awesome, thank you so much, and hopefully we can have you back sometime soon.
Chris:Thanks for having me.
Jess:Thank you all for tuning in this week. We hope this theory is relative to your marketing needs. Make sure you subscribe to get notified of our latest episode.