Christopher S. Penn – Approaches To Creating And Using Media - 2020 12 16
8:17PM Dec 16, 2020
Speakers:
Keywords:
machine learning
tillage
attribution
channels
people
questions
data
podcast
seo
technology
company
analysis
marketer
spend
market
customers
content
blog
learning
marketing
Welcome to ag media conversations. I'm John Blue. And this is a chat in December 2020, with Christopher S Pen, co founder and chief data scientist with trust insights.
Welcome. Thank you very much. Thanks for having me. I love love chatting with you, you learn some really interesting things. A few years ago, we were in Nashville together for the ag chat conference and stuff. And
I still have
some of the cover crop seeds. We planted them, they actually turned out pretty well. The the tillage radishes did surprisingly well.
Awesome.
Great. I do have. So I'm gonna start off with one question. I try to ask this in all the interviews. And when relatives ask, Hey, Chris, for what do you really do? What's your response?
It depends on who's asking. Typically, like I tell my parents, I just work with marketing data, I help companies fix their marketing data with my, my brother, who's an engineer a bit more about, you know, using machine learning to help companies fix their marketing data. But that's that's really about it.
Nothing, nothing crazy is a really you make money at this.
You know, that has never actually come up. I think people are aware enough to know that there is a lot of data in marketing. And to no one's surprise, a fair number of people got into marketing to avoid math, and are now at a point where like, Oh, we need math, we especially went to schools to avoid this. So we need to hire some people to do this. So as a result, it's not a surprise that I think that my relatives No, you can make money at helping people fix their problems. Great.
One, this is something I remember you've you've talked about in the past, and one of your early 2000 issues experiences was working with the at the student loan network, as as as it was listed, you're their chief technology officer. And you also started at the same time, the financial aid podcast, as it related to the student loan network. And and when you listen to the financial aid podcast, it was all content focused on students. But I think you actually created it for a different target. Can you explain that?
We actually, well, so it's complicated. The way student loans work, and anyone who's a parent knows is intimately familiar with just how painful This is. It's, it's a third party payer system. So you have the student who takes out the loan, or, and or the parent in the family, you have the lender, you have the school, and then you have the government. And the way it works is that in order to market to people, you need to market to financial professionals at the school, to convince them to use your company, as a lender, you have to market to the family to say, Hey, take out the loan from us. And the company I was at at the time was one of the first direct to consumer, internet, student loan companies. And you have to market to the banks as well to say like, yeah, we wanted to be this we were a marketing company, for banks effectively. So it was trying to market three different audiences. The podcast, was a, it was an attempt to use a new medium at the time, because podcasting had just come out on 2004 2005. And we gave it a try to see if we could reach anybody, and doing so in a very low cost way. Because, you know, I used to joke, our startup, our annual revenue was like the cream cheese budget, Sallie Mae, no way could we compete with them, you know, head to head dollar for dollar in traditional form formats. But we saw an opportunity there for us to be the leader in a very small new market, it would be the equivalent of, you know, a company these days, starting up on like, clubhouse right, or Tick tock tick tock has sort of jumped the shark and is now a mainstream media property. But getting in early on a platform really made it easy to be noticed. And to gain some headway when good because a couple years later, that, you know, we got profiled in US News and World Report as a leading edge, you know, financial services company, because we were using new media and it wasn't because we didn't want to use old media, we just couldn't afford it.
Great, thank you. You know, technology goes up and down. And it comes and goes, and what's your approach to one staying on the road to help your customers and then to going off road to see what's new or what's happening?
You know,
it's funny, technology's one of three pieces you need, right? So if we go back to like 1964, IT consultant ha Leavitt created this diamond framework, which has since been simplified to people process and technology, right. And those are sort of the three ingredient You need to make something work. When we look at technology and onroad offered the why, and the what largely stay the same. They're the onroad pieces, right? Why are you doing thing? What are you doing?
broadly?
It's you go off road at the how. So? Why would you be doing? Why are you in the business you're in? Right? Is it to feed people? Is it to supply restaurants? Is it what is the thing you're doing? What is it? You do? And then how do you market it? You know, and, yeah, it depends, you know, agriculture is very similar and a lot of ways to finance and that you could have, you know, direct to consumer companies, and you have b2b companies, if you sell all of your, you know, your feed to Tyson Foods, you really don't interact with the consumer, right? You don't sell chicken feed directly to the consumer, other companies do, like I have chickens in my backyard. And we have to go by feed for them. And you know, dried mealworms, which is just freaky stuff. And, and those companies do market directly to the consumer. So when we think about why and what that stuff is pretty timeless, and you get that through talking to customers, when you talk to a customer say like, hey, why did you pick this company? or Why did you pick that company? Why did you pick this vendor? What goes into your vendor selection process, those are the things you need to understand. And then the How is where you can get creative, we can find new technologies, or even just evolutions of existing technologies. So for example, this week, we've been doing a ton of work in Google Ads with one of our clients. And there's so much it's so different now, even from just 18 months ago, that it's like a it's always a brand new place to market. And if you know what you're doing, and you have access to the financial resources to do it. It's a pretty incredible experience. You can do you know, machine learning models, trying to detect which customers are likely to convert, and do you know, 15, way AV testing of your ads and stuff and some crazy cool stuff. And I, I would encourage people when you're looking at these platforms, the platform comes after the why and the why if you don't have that down first, if you don't have on road salt, offer, it's just gonna be a disaster. But then yes, then you get an offer. Look at the different ways people interact. And this is the thing that I feel like, that people get wrong a lot is they look at the technology, and they don't look at the audience. Ask your audience, where are you spending your time? Right? Where do you spend your time online? And then go to where your audience is, especially if it's you have a subset of like, these are our 10 best customers, if you know where your 10 best customers who they are, pick up the phone or text them or something and ask them like, hey, do you spend time in slack? Do you spend time on discord? Are you on Tick Tock ion parlor? Where are you? And where do you go for information? And that will that should govern what you do with the technology much better than just trying stuff out for the sake of trying it out?
If you had 15 minutes to talk with anyone in history, who would it be? And why?
Ha.
That's interesting, because there's so many people you could talk to. It depends do I get the ability to document it and record it? Because I have a whole bunch of questions for some, you know, substantial historical figures. Like why did you do that? or What did you actually say? You know, it's that time of year? Yeah, I think it would be really interesting, assuming that I spoke Aramaic, which I want to go back to, you know, 30 ish ad and see who exactly was the historical figure of Jesus Christ? And, you know, did that person exist as an individual entity? And 15 minutes, okay. Okay. Supposedly, a bunch of people, including some who did so long after your death, said you said these things. Which of these things are actually true? And is this what you actually meant? Because a lot of the oral traditions passed on back then, you know, it's like playing telephone. And then you have you know, Saul and Paul and like having visions, you know, 70 years after, after Christ's death, like yeah, I don't know that I trust your your, your vision. I but you know, the ability to turn on your iPhone, you know, turn on the tape machine, bring your iPhone with you, and record say, okay, which of these things do you actually mean, and what's your take on X and assuming that you could then credibly bring that back to the present time?
We very interesting. You may have answered this already in the in this the first couple of questions, but I want to go ahead and ask it and see if there's anything you want to add to it. You know, there. There are a lot of many digital marketing tools, tips, tricks, conventions approaches, and this certainly creates a lot of noise. What's your approach to making choices and decisions in that noisy environment?
attribution analysis, attribution analysis, by far is the greatest single tool that you can use to assess your digital marketing efforts and figure out what's working for you, especially for channels where you don't know necessarily whether it's going to work out, but you you've run a test, you run a pilot, then you've got to looking at data, you run an attribution analysis and say, Did it work? Did anything? Did we get anything out of this? We do one for clarity.
Can you define what attribution analysis is for those who may not know it?
Sure. So attribution analysis is all about who gets credit for what? So if you were to look at a in your Google Analytics data, extract that data out and say, what are all the different channels that we're using? email, Facebook, Google search, paid ads, YouTube, all these different things? Or channels, or sources and mediums? How many of them drove conversions? Right, whatever the conversion is, for your company, whatever that looks like, how many of them drove conversions? And what how many did each channel generate? Because if we don't know, what's working, what's not, we can't calibrate our marketing efforts. So an attribution analysis looks at a lot of ways. You know, it looks deceptively simple. It's it's a bar chart that shows you know the channel and how many conversions either directly drove or helped drive. And you look at that and say, Okay, well say email was 51% of our conversions. Great. Did we invest more than 51% of our resources and email? Facebook ads? Were 10% of our conversions. Great. Did we spend 10% of our budget on Facebook ads? Now, if you spent 5% of your budget on Facebook ads, if you got to 10%, you know, lift that up, then you did? Well, on the other hand, if 40% of your budget was Facebook ads, you got 10% your conversions from it? You got a problem, right? And so when you have that data, you can then start making those decisions. Like, yeah, we put an awful lot of time into this podcast, you know, and, and no one's visiting those pages on our site is not converting. So if you are resource constrained, you have to ask those hard questions like, yeah, should we be doing this podcast? I don't know. And all of that is predicated, though, on having a good analytics infrastructure, having it set up well, so that you can track that data. Because if you don't have that infrastructure, then you're flying blind.
You You appear to have many projects and activities going on. You know, personally, I know you've it appears that you write a blog post every day, it may not actually be Daddy, but it sure seems like it and you've been doing it since I don't know the early 2000s whenever blog started to come on. Plus, I know you have other channels that you slice and dice for and customize for, like YouTube and Twitter and others. So what's your productivity tip to keep it all organized and going forward toward what you want to accomplish?
So we have a framework called the transmedia content framework, if you go to trust insights.ai, you can find on there somewhere. That's probably not as specific as I need to be. But it boils down to this video is the richest source of data you have, right, because within video, there's the picture that's constantly moving, there is the sound, which is the words, the audio, and then this derivatives you can make from that. So from this one video, that recording right now is the zoom call, you could split out the audio. Great with that audio, you now have a podcast, right you have an mp3 file, take that audio, put it in a tool like otter, if you go to trust insights.ai slash otter Ott, er, like the animal. Now you can sign up for a free account got it. It's got 1000 minutes for free. You put the mp3 in and it transcribes it using machine learning into into text. Great. So now, in one video, you've got a video for YouTube, you've got you can cut up the video for Instagram and other channels. You have the audio for a podcast, you have the text for a blog post an ad or a newsletter or whatever. And so you created four pieces of content out of one video relatively in a relatively straightforward manner. So like every day I publish a blog, a 1500 word blog post a podcast and the video people like how do you do that? And, you know, in 40 minutes, like I record the video, and then I and I follow this process to slice and dice so that I creating the content for the ways people want to consume it. Some people want to listen, some people want to watch some people want to read and this method helps you fulfill as many of those channels as possible.
Did you discover that first? Or did you evolve that over the last 15 years or whatever that is?
So this was proposed years ago 12 years ago by a friend and former CEO of mine Todd back in 2008. called content optimization. The idea can take take content just, you know, break it up into individual pieces. For us, though, I mean, it's something we've been advocating really for the last five years. Because it's, there's so many marketers out there who like, I don't know how to create more content, well, you've got a lot of content, you're just not using it. Well. One of the eye opening moments for me was, I just like 2015 2014 2015, when we were talking about the Marriott blog, and how the CEO of Marriott was one of the most prolific corporate bloggers. And then you go behind the scenes, and he doesn't blog at all, what he does, is he leaves a 10 minute voicemail message for his team, while he's traveling, you know, just his thoughts on that the industry, they transcribe these and turn them into blog posts. And so you know, suddenly build, marry out, yes, blogging all the time, you know, build marriages, leaving a message in the middle of the night, as he's going from, from place to place. But it's one of those things we think about, okay, how could you get more out of your subject matter experts, without making them work too hard, you fire up a zoom or phone call, or whatever, say, hey, just tell me all about this. Tell me what you know about this. Take customer questions and say like, for example, how would I get how to get greater yield out of red wheat? Right? How do I get a greater yield out of that? That's a good customer question. If you've got somebody on staff who knows red, hard, red, wheat, hard red wheat really well. They can say, Oh, yeah, you're an iTunes probably got a balance, right? Or your soils not draining well enough, or any of these things. You've got subject matter experts with knowledge locked in their heads, but they don't have the time to sit down and write this about, so just call them press releases.
What do you think the most misunderstood thing about press releases today?
That they work?
So which leads to my next question, then is the press release dead?
Um, press releases are useful if they meet a regulatory need, they are still the gold standard for the SEC regulation fd for financial disclosure. That said, beyond that, they really have no impact. Just a yesterday over on the blog, we or was yesterday. Yes, yesterday.
It was two days ago,
we talked about the effectiveness of press releases, and the you know, the the short answers out of 198,000 press releases, the average number of clicks that they got was zero. Right? You know, it's about traffic, you know, estimated is zero. And so I think that they are there, there's a time and place for them. Again, regulation certainly is one of those places where they're required. But beyond
that,
it doesn't move the needle anymore. It has no SEO benefit at all zero. And if you've got to take the you know, two to 12 $100 to go spend at least I would rather see you put that to Google ads, or I would rather see put that to YouTube ads or to an email campaign something where you know, you're going to get some return?
Well, we've put that to rest. You've done a lot of experimenting with, with machine learning. just for clarity, you do a lot of experimenting a lot of things in machine learning and has certainly come up in the last several years and and it shows on your blog posts and in your conversations on marketing over coffee with john wall. And you talk about the potentials of machine learning. And you give case study overviews. is machine learning today, approachable by smaller organizations, that would be like a two to 50 size person agency.
Chances are you're already using it. You just don't know it, right? I mean, honestly, in your personal life, if you use Google Maps, you're using machine learning, right? If you use Google, you're using machine learning. If you have one of these smart assistants on your desk, you're using machine learning in the marketing space. Google Analytics is powered on the back end by machine learning, especially the new version version four is a machine learning engine. Behind the scenes, you don't directly interact with the models or the the algorithms but it's working for you in terms of the ability for a marketer to sit down and custom build their own machine learning model. No, that's not really accessible to the non technical marketer. But also it may or may not be necessary. There are so many things that you know some of the basics you need to get down first in marketing before you worry about the advanced stuff. Can you get multiples of benefit from the advanced stuff? Yes. 100%. You know, the best attribution models are machine learning based models hands down, they delivered better results and better guidance. But the something I always say to clients all the time. All that data at analysis is is amazing and wonderful, but if you don't do anything different, it's a waste. Write Seth Godin has a really great expression if you don't change what you eat, or how often you exercise don't get on the scale, right? And it's the same thing is true with machine learning. If you're not going to change what you're going to do, how are you going to do it as a marketer? You don't need the technology because it's not going to benefit you, you're going to get an answer from it that you'll then ignore and just do what you're going to do anyway. And you'll get the same results you've gotten. In terms of practical things. attribution analysis, I think, is one of the most important applications of machine learning. I think for SEO, there's a tremendous amount of opportunity there now and the vendors that are in the space, for the most part, are not doing a great job. By the only vendor, I would recommend, be marketmuse. They are reassuringly expensive. But their optimization technologies are some of the best I've seen. Because one of the things that happens with a lot of you know, more traditional SEO tools, they are very keyword focused. And that's not the way search engines work anymore. Search engines look at topics and ideas and concepts. So again, going back to you know, if you are talking about growing crops, right, you're searching for growing crops. And you're giving guidance about tillage. You may not think to write about, you know, rainwater patterns. But rainwater is an essential part of tillage, right, depending on how much rain you get determines how much tillage you need, because it's how tightly compacted the soil is. And so Google would understand a search about someone looking for dealing with you know, rainwater runoff problems, and might bring up a post about tillage, even though you don't worry, you actually forgot to mention rainwater in your post, because it knows conceptually, those two concepts are interlinked. And so a post about, you know, tillage radishes would bring up come up in a rain water search. So there are strong applications for machine learning in SEO. But vendors are taking a long time to catch up. Okay.
I wanted to touch on a phrase used a couple times I've heard it. And you said, are we putting resources to work commensurate to what's working ie the results? Can you explain what that means?
That's exactly what we're talking about with attribution analysis. If you break down your channels, and look at the percentages, are you over investing time in a channel commensurate to the results you get from it? And one of the things that marketers go really wrong with here is on soft dollars, so hard dollars, easy to explain hard dollars that when you open up your wallet, yeah, your hand Mark Zuckerberg your credit card. That's hard dollar expenditure, soft dollars, or the time you spend building Facebook ads are creative, or you know, posting on Facebook and stuff. Every minute of the day you spend doing that as a minute, you're not doing something else. So there's an opportunity cost there. And so one things you have to assess is how much how many soft dollars are you investing in any given channel and add that to the cost? When you're doing your ROI computations, you may have a hard dollar positive ROI for blogging. But when you put in the hours that you spend on it, suddenly and you count those hours as money that's like oh actually has a negative ROI. Because you know, we spent five hours a week and you make, you know, 30 bucks an hour as a salary. And suddenly you're spending $150 a week in time on blogging. And if you earn $50 that week, from your site from from SEO from content. We're not we're not breaking even here.
Two more questions. What's the one purchase you've made that you can't live without
a house.
It'd be really kind of hard to get by and do a lot of stuff about that.
I think that's the that's the biggest item I've never heard anybody actually say for this question. But you're right. It's kind
of the practical, it's not the fun answer, but it is the practical answer.
And then where can people learn more about you?
You can find me at trust insights.ai for the company, and Christopher s pen.com for my personal blog.
