HOW TO GROW YOUR AUDIENCE WITH MACHINE LEARNING

Tips on how to grow your audience with machine learning with Patrick McKenna

  • We help people expand their audiences so that they don't see a diminishing return. And it's not just how to grow your audience, it's also finding those interests that actually work for your audience and your brand.
  • The good news is that if your business is growing, and hopefully it is, the artificial intelligence that you're building isn't going to eliminate your employees. In fact, you may need more employees. They may be just shifting in direction, but there's still plenty to do.
  • The main products for us are analyzing buying metrics and audience metrics.
  • What you get out of that data if you haven't broken those things out in the platform is very little. You're not really sure what specifically worked when you don't know specifically what worked on the keyword side or the interest side per age group and per device.
  • One of the tricks you need to watch out for on Facebook is if you put up a couple pieces of creative in there, Facebook will make a decision really quickly as to which one is performing better. And you need to kind of wait a little bit and give that second piece of creative a little bit more statistical significance because Facebook will pick a winner really fast. And you may throw a really nice piece of creative out the door too quickly.
  • And so you need to be testing, you always need to be testing.
  • The platform guys want to make this sound really easy. Google wants to make it sound really simple to go and measure and use their ad manager.
  • I get really excited about this direct to consumer situation. We call it the DTC, the dollar shave. Gwyneth Paltrow's brand, Jessica Alba.

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

HOW TO GROW YOUR AUDIENCE WITH MACHINE LEARNING

[just click to tweet]

HOW TO GROW YOUR AUDIENCE WITH MACHINE LEARNING

And it's not just how to grow your audience, it's also finding those interests that actually work for your audience and your brand.

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

Doug: Well, welcome back listeners to another episode of Real Marketing Real Fast. Today, we're going to be talking about testing, testing. No, the microphone is working fine, testing your advertising. So how do you know what's working, what's performing, what's moving the sales dial, what audiences are working on Facebook? Are they the same audience as working in LinkedIn, are the same audiences that are working on Instagram or YouTube or Google ads. So we've got a guest who is a specialist in this space, and they use machine learning and artificial intelligence to guide your path and guide their clients' path to making the right decisions. And significantly in many cases, it's going to be lowering the cost of advertising and acquiring a customer. So I'd like to welcome Patrick McKenna to the studio today. He is the founder and CEO of a company called Strike Social. It's a global technology enabled digital advertising agency based in Chicago.

Doug: Patrick has more than 30 years of experience in business development and technology consulting for major corporations. I'm sure you'll gather that when you listen to the episode, he co-founded and sold a telecommunications company, WCI in 1996. And then he was recruited to Microsoft as part of a 50 person team to develop the internet, audio and video industry. So by the end of Patrick's 12-year career at Microsoft, he had developed Windows entertainment division that now employs 40,000 people. So Patrick has spent his career mentoring and investing in startup companies. And in 2013, he founded the Strike Social, which was named the number 17 fastest growing private company in the US in 2017 by Ink Magazine. So Strike Social, develops artificial intelligence or AI-powered software and services for digital advertisers across industry verticals with over half the Fortune 500 companies benefiting from his solutions.

Doug: And some of the clients and the brands that leveraged the tools that Patrick's company Strike Social offers are companies like Beat, Xbox, Honda, Mattel, Lion's Gate, and other large financial institutions. So I'm sure you'll enjoy the discussion. We're going to get a little bit technical, but we'll come back and we'll bring it back to reality. So how you can take Patrick's tools that he used with his company and look to see if there's a fit for you to grow your brand, reduce your advertising cost, and move the sales dial. With that mouthful of information introducing Patrick, I'd like to welcome the Real Marketing Real Fast podcast today.

Doug: Well, welcome to the podcast, Patrick. I'm super excited to talk to somebody who loves sales and marketing, and someone who's a digital expert in AI. So welcome to the show.

Patrick McKenna: Oh, thanks, Doug. Thanks for having me, it's great to be here.

Doug: So when we say artificial intelligence, I mean that raises a lot of connotations to people. So do you want to explain it in your words what it is that you guys do?

Patrick McKenna: Yeah. I mean, I think when you hear AI, it's how do I replace human beings? And I think that if you're not working on that solution, then you're going to be challenged. I mean, the good news is that if your business is growing, and hopefully it is, the artificial intelligence that you're building isn't going to eliminate your employees. In fact, you may need more employees. They may be just shifting in direction, but there's still plenty to do. We take a look at, when we started building our artificial intelligence, it was really to help by using a machine, help to get rid of repetitive tasks. That was the first piece of it, I hate repetitive tasks. It's one of those things you wake up and you're like, now I got these things I got to do and here comes a campaign and I've got to click this, this, this, this and this. And you're doing it over and over, and over again every single day.

Patrick McKenna: And I think that's how people get really bored doing their job. And so we started there, and then as we started moving down that AI path, we started using it for things like mathematics and optimization and re-budgeting and forecasting and planning. Once you get in there and you start building this system, you start to realize that you can automate a lot of things. But our business, it's not like we lost headcount. Because we grew, the headcount continued to increase.

Doug: Well, I mean in looking kind of at the services or at least direction you guys are going, I don't know if this is too broad of a question, feel free to push back because we talked about that you can get nerdy and deep into the technology side. But you're saying you're using AI for gathering intelligence on social platforms like Facebook, Twitter, YouTube, LinkedIn, Pinterest. So how does that work, and what does that mean to our listeners?

Patrick McKenna: Yeah. The main products for us are analyzing buying metrics and audience metrics. You're running campaigns, you're multi-variant testing those campaigns in, let's say Facebook ad manager, AdWords or Google's DV360. And you're bringing that in with the clean data and naming convention so that it can be analyzed. And then what we did is we took about five years of data that we had and we were running globally. And we brought that into a group of data scientists and we built that team from one to three really quickly. And what they do is they program a box, and they are able to look at things that, the normal AdWords guru or Facebook guru buyer just can't, right?

Patrick McKenna: You're taking a look at audiences globally across campaigns and you're looking for similarities between audience groups, say 18 to 24, 44 to 55. And you're starting to look for if I'm trying to target a group of people. I'll give you an example, let's say that 18 to 54-year-olds are interested in UFC and the 40-year old plus crowd is still interested in the NFL. You want to target the 18 to 24-year-olds. If you go throw NFL in there, you're not going to perform very well. They're not looking at it. You've got to find that similar audience that the 18 to, and it could be 18 to 30, 34 that's going to be interested in that type of sport or that category that fits your brand. And you've got to be able to make those correlations.

Patrick McKenna: And so we do that across all the social platforms, and we help people expand their audiences so that they don't see a diminishing return. And it's not just expanding the audience, it's also finding those interests that actually work for your audience and your brand. And it's very, very challenging to do. On the front end, there's a lot of psychographic planning that goes on most of the time. And we can take care of all that stuff because we've analyzed what's going on in the buy side. Does that make sense?

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

HOW TO GROW YOUR AUDIENCE WITH MACHINE LEARNING

[just click to tweet]

HOW TO GROW YOUR AUDIENCE WITH MACHINE LEARNING

And it's not just how to grow your audience, it's also finding those interests that actually work for your audience and your brand.

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

Doug: Kind of makes sense. So just so we make clear for our audience, I'm currently running Facebook and Google ads as an example and we're running them with an agency. And I'm getting some analytics and Google analytics, and I'm getting some reporting. And I'm obviously looking at the back end of our shopping cart, and we can see the sales stuff. But that sounds like a pretty high level of view, but we don't really have the in-depth look to say, hey, maybe you should move over here. Is that what you're saying is that you've got enough outside data and intelligence to make recommendations that our ad guys aren't going to see?

Patrick McKenna: Yeah. And I think part of that is in the setup, in the way the platforms are set up. They're all a little bit different. So Facebook wants you in a bucket, Facebook says, put all the stuff in here that you think you know about your customer in terms of audience lookalikes and everything else. And ask our computer or one or two questions. And once we make that sale and our algorithm learns, we're going to continue to go find in our system more customers like that. And so what you get is a score. You get a relevant score. Let's say you get a three or a five or a seven. The better the score, the more on target you are with the audience on Facebook. But what you don't know is what's behind that score.

Patrick McKenna: If I'm getting a five, how do I get a seven, and what's behind that seven in the audiences? It's better if you're breaking them up a little bit more on Facebook. You don't have to break them up nearly as much in Facebook as you do in Google AdWords or DV. You need to do really extensive multi-variant testing within those two platforms on the Google side. What I mean by that is you need to take every single age group and the demo, male or female. And you need to break it out by interest, topic, keyword, affinity. Every single one of those needs to be a separate campaign. And then you need to go to device, mobile, desktop, tablet. And that's the only way that you're going to get the data back that says, hey, I'm on target with this younger crowd and middle age crowd and older crowd. That's the only way to do it on Google.

Patrick McKenna: There are two slightly different things. And then what we do to sort of add to that sort of, let's call it global social planning piece is they use different words within their targeting platforms. So let's say music is an interest in Google, and it may be called something different on Facebook. You need to be able to map that. If I'm doing well in music, what does that mean in Facebook or Snapchat or Pinterest or one of these other platforms? That's in the platform too. But to your point, downloading CSV and saying like, oh, hey, okay, CSV file is your, starting your analytics, right?

Patrick McKenna: You download this thing that's basically an Excel spreadsheet and you start doing pivot tables. What you get out of that data if you haven't broken those things out in the platform is very little. You're not really sure what specifically worked when you don't know specifically what worked on the keyword side or the interest side per age group and per device. You don't know how to keep expanding into that. Okay, they're interested in music. Well, what types of music are they interested in? And you can go down that rabbit hole really, really far, and especially when you start talking about age groups. You've got your classic rock, your classical music. Then you've got all the EDM stuff and everything else, so you've got to know because that's going to help you across all the other social platforms too. Does that help?

Doug: It does. Now, do you find that the different platforms may respond differently? If you're looking at comparing a Facebook audience to a Google audience or to a Pinterest audience that you can't take the same approach across all platforms even though you're targeting the same customer avatar.

Patrick McKenna: Yeah. Honestly, let's say you're a Search guru and you try to take that buying information that you know on Search over to Facebook, you're not going to be successful. You have to bucket over there. One of the things that we talk a lot about in what we see when we analyze someone's Facebook account, we'll see them spend a lot of time in their lookalike audiences. You upload an email list and you put it into Facebook, your first party data. And then you say, “Facebook, find me people like this.” What you'll see is people sit there, and what's gonna happen is at a certain point in time, you've run out of what we call net new.

Patrick McKenna: You need to be feeding that audience pool with new people or the algorithm is going to spin into that audience and your costs are going to go up. They're just very different platforms, and they do need to be bought completely differently and planned differently was to my point was that if you don't have a digital tool to do that, the likelihood is that your creative teams are going to do one creative and say, “Okay, here are our images and our copy for ads across all platforms, execute,” and then wonder why platforms are underproducing or your cost is going up.

Patrick McKenna: Yeah. We were just talking about the AdWords and the Facebook side. When I think of AdWords, I think of Search, and Facebook is very visual. You hit the nail on the head on the Facebook stuff. One of the best ways to perform better on Facebook is to make sure your creative as an exhausted at the audience. We have tools over here that we look at creative exhaustion, and one of the best ways to perform better in that platform is to do a lot of trade of swapping. One of the tricks you need to watch out for on Facebook is if you put up a couple pieces of creative in there, Facebook will make a decision really quickly as to which one is performing better. And you need to kind of wait a little bit and give that second piece of creative a little bit more statistical significance because Facebook will pick a winner really fast. And you may throw a really nice piece of creative out the door too quickly.

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

HOW TO GROW YOUR AUDIENCE WITH MACHINE LEARNING

[just click to tweet]

HOW TO GROW YOUR AUDIENCE WITH MACHINE LEARNING

And it's not just how to grow your audience, it's also finding those interests that actually work for your audience and your brand.

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

Doug: Oh that's interesting. That's brand new information to me.

Patrick McKenna: Oh, good. Well, that's why I'm on the show, I hope.

Doug: I'm surprised to see that you guys have LinkedIn there and Pinterest as well. That's not what I would have expected when I first dove into your site. I thought you would've hit kind of the main social platform is because the content required for each of those is different. When you start posting Facebook like content on Instagram, you normally get some pretty quick negative feedback from your audience.

Patrick McKenna: Yeah. So we don't do creative, right?

Doug: No, I understand that.

Patrick McKenna: We're an optimization company and we go to work with agencies. When we get a plan from an agency, it'll have a budget spread across most of social. LinkedIn is typically fairly small. I mean, the big ones are Facebook, Instagram, and Google. The rest of the guys are kind of battling it out a little bit. Pinterest is actually, their ads manager has gotten a lot better. And that's kind of a life cycle of these platforms. At first, Facebook wasn't very good, and now they've become really good. In fact, I've seen more budget going to Facebook these days than going to Google Search on YouTube. So it's interesting to see how these platforms evolve over time. Snapchat has got some work to do, I think.

Doug: Well, it's funny because I still hear people say, “Oh, Facebook advertising doesn't work.”

Patrick McKenna: That's not true.

Doug: Thank you. And you hear guys like Gary V. going, “No, it didn't work for you,” which is a totally different conversation. So how did you guys scale and build your company so quickly? I mean, you're obviously in the data business, so it's fair to assume that you took your own advice to grow your business. Do you want to share how you leveraged, how you grew your own brand?

Patrick McKenna: Yeah, sure. We looked at the market and we looked at what was going. YouTube was sort of the only place to do video on the Internet at the time. It was early in there. And we got started in March of 2013. What we noticed is that we can perform really well doing that multi-variant testing strategy in the platform. And one of the things that I think really propelled us to grow faster than the others is we modeled our business off of a flat margin, and we sold it transparently. And my model and motto was if I can't allow my customer to see what I'm doing, then I don't have a business. I should have enough secret sauce on the backend that I'm confident that my customer couldn't just go unplug us and perform the way we perform.

Patrick McKenna: And that's how we went to market. The way that other people were doing it was we're going to run the ads in our account, and you don't get to see anything but a final report. And what we said was, “You can sit in our account, you can watch exactly what we're doing. We don't care, and we're going to deliver to you in a completely transparent way. Here's your report, here's how much we charge” And Doug, as crazy as that sounds, maybe now, that was not happening at all in the early days. We were on a trend about transparency very early on before all the transparency ANA stuff really hit the market. And so we were just in front of it, and that's what really, really helped us grow. Do you understand that?

Doug: Yeah, I do remember that. I remember watching these platforms come out saying, hey, if you want to run Google ads and Bing and Yahoo and whatever the other flavors of the day were, you need to sign up. And it all ran through that common platform. And my concern always was if that platform fails, closes or underperforms, I don't have control of the individual accounts because it's all gone through their tool. So I don't have access to the data in the future.

Patrick McKenna: Well, yeah, and that's part of it too. For Facebook, essentially, you have to run in the brand's account because that brand account is learning with Facebook's algorithms in the background. You don't want a third party running your Facebook for you in their Facebook account, you want it on your account. But the other side of that is when someone's running in your account and they're just flicking you a bill, you don't have any idea what they paid for it. You can assume there's a mark up there because they're not working for free. But what is that mark up? That was the thing we wanted to eliminate. In the early days, we would hear one of our competitors come in and say, “Hey, we'll get you a 10 cent CPB.

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

HOW TO GROW YOUR AUDIENCE WITH MACHINE LEARNING

[just click to tweet]

HOW TO GROW YOUR AUDIENCE WITH MACHINE LEARNING

And it's not just how to grow your audience, it's also finding those interests that actually work for your audience and your brand.

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

Patrick McKenna: And maybe the agency would even dig in and say, “Well, what are you charging?” They'd say 15%. And we'd come in and say, “Hey, we're looking at your ads manager account here, and we think we can bring it in for 4 with our 15% margin.” And they would just be confused by that. The bottom line is the other company wasn't charging 15%, they were charging 80%. That's what started really making brands upset, and we just didn't want to be in that business. We wanted to just have a flat fare margin that people could live with and the performance needed to be there.

Doug: That totally, totally makes sense. Where do you think the low hanging fruit is for our listeners that are listening in and going, okay, I'm spending all this money advertising, I'm either doing it myself or I've got an agency doing it. Where's the biggest opportunity that they're missing?

Patrick McKenna: There are so many platforms and technologies and tools out there. I would say that the one thing that I think people are missing the most is, let's call it a level of happiness. You think you're getting a good CAC, you're getting a good ROI from your ad spend. And maybe even you're happy with the provider that you're using. And so you stop testing. And you're like, “Well, we're good, we've got a good partner here. I'm looking at your tool, it sounds good. It's going to be too hard for us to test.” It's not too hard to test, and you're in advertising. And so you need to be testing, you always need to be testing. I think that's the biggest opportunity. You need to get over the hurdle of we're good, it's all good, everything's running good because we see stuff like that and we'll come in and drop their CACs by 50%.

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

HOW TO GROW YOUR AUDIENCE WITH MACHINE LEARNING

[just click to tweet]

HOW TO GROW YOUR AUDIENCE WITH MACHINE LEARNING

And it's not just how to grow your audience, it's also finding those interests that actually work for your audience and your brand.

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

Patrick McKenna: And it's like you took us six months to get here and you lost that six months of opportunity. I'm not saying we're the best one, maybe I'm a little bit biased. And I actually think that. But it's frustrating to watch a brand or even an agency sit in a toolset and think we got this right.

Doug: Yeah. We're always testing. I carve off a small piece of every budget to test against the baseline, and so I get the testing side. What came to mind when you said, oh, we're good, is you're good until you're not good. You're sitting in your comfort spot and then you get kicked out of the chair, and then you have a problem because then you're scrambling trying to figure out what did I miss, what went wrong? You're playing catch up.

Patrick McKenna: I mean, and think about that. There are opportunities out there where you can literally lower your CACs by 15 to 50% and still pay a third party to do it. And I think what's happening on the brand side is they're just like, oh, I tried this agency or I tried this tool and it didn't work. And they get frustrated and they stop. And I just think that's a big mistake. Think about that from a bottom line perspective at the CEO, CFO level. And that's massive. That's the part I think that is the biggest opportunity in the space. If you're not really testing where your happiness level is, you're not going to know how happy you could be or how happy you are.

Doug: So what does that conversation sound like? I'm VP of marketing, and we're working with an agency and we've been working with the agency so that relationship's not likely to change. So what does that conversation look like between me and the agency and me and my CFO?

Patrick McKenna: Yeah. Well, I think on the CFO side, it's easy. It's, “Hey, look, I want to test something. I want to have a test budget,” like you were saying, I break off a little bit and test. I think on the agency side, the brands in a lot of cases need to be telling their agencies that they want them to test. And what it looks like is I want you to bring me new ideas, and it's hard to bring a new idea if Google is running your YouTube ad campaign. It's hard to be like, well, we're willing to try a third party, we're already working with the source. We're working with Google, how would a third party do? And then the conversation can end there.

Patrick McKenna: You've got to be able and to be willing to go to battle to get a test sometimes. And I don't think a lot of that is happening. It takes months and months to put a test up. You could literally take YouTube and split test in the same platform and watch it. You can't do that on Facebook. You actually have to make a little bit of a commitment. But on YouTube, it's just, okay, you run $10,000 budget, you run 5 with them and 5 with them. And it's that simple. And it's really weird, Doug, it needs to be happening more. And I think that people are so busy they don't want to go to the battle.

Doug: Yeah. I see so much in comments in terms of masterminds and various groups of people just trying to figure out how to get their ads to perform. And like you said, they don't have that transparency. They don't really know what they're paying for clicks so they can make an educated decision. And they may get the surprise when they change agencies to find out, hey, those guys were taking more margin than they said they were, but that's past history.

Patrick McKenna: Yeah. And then you hear the horror stories, well, we switched to another agency and they screwed everything up. We found one, we're happy enough. We're not going to-

Doug: Rock the boat.

Patrick McKenna: Yeah, stop testing.

Doug: But I think the great news is, is because you're a third party, so you don't have a vested interest because you said your model is based on what they're spending. But you don't have a vested interest in what platform they're spending. You're really the third party, like the police that actually provides the data. I've done that with email campaigns where I've got clients that are spending maybe 50 grand at a pop to do an email marketing campaign. And I'll go in and do the audit as the outside third party because I have no vested interest in what media they're buying. I'm just providing them the reporting and the analytics and the data to make sure that the agency they hired did the work they said and the publisher did what they promised to do. And that's it, I don't care what their numbers are, I just need to provide the right numbers.

Patrick McKenna: Right. And I think you're bringing up a good point too. That measurement side, it's really lacking. What we're seeing Facebook will go take a lot of credit for a sale. And then you'll go back into your CRM and you'll be like, “That wasn't you.” In the ads manager, it says, hey, I made a bunch of these sales, and then when you settle up. People need to make sure that they're settling up.

One of the things that are taking a lot of budget out of YouTube and going into Facebook is Facebook looks like I'm getting this massive ROI. And YouTube is more of either awareness or it's used for the mid-funnel to drive search results, branded keyword search results. And those are really inexpensive if you're doing it right. But if you haven't figured out how your YouTube campaign is driving up search queries and looking at those results, you may look at YouTube and just being like, this is just a waste of money. And I think that you got to really spend a lot of time on that big picture.

Doug: Well, that's interesting because I think of the sales cycle as people are buying and the way that the world has kind of changed for where I'm at in my buying cycle. Am I doing research? What platforms am I looking at to do research? And then, where do I ultimately end up making that final click that takes credit for the sale? And you miss all the data in between. So all you see is, hey, they clicked through my Facebook ad, so as you said, Facebook is to credit. But what I missed was they were in consumer reports and they did a Google search and they watched some YouTube videos. And they had all these other things that made the decision. It was finally the tipping point was I've made a decision, I like this brand. There was the ad, boom, I collected and bought.

Patrick McKenna: Yeah. And I think that's why Google has changed DBM to DV360. I mean, they're really trying to show that multi-touch model and how that awareness and then mid-funnel and then search, the whole picture. Facebook is kind of like crack. It's just, oh, got another sale up, oh, got another, oh, got another sale. You just start heavy and up in there until you start seeing this massive diminishing return there.

Doug: What some of the bad advice you hear in the industry around what you do?

Patrick McKenna: Around what we do or just like what we hear from the platform providers and the education that is given to the client either?

Doug: Either or both.

Patrick McKenna: The platform guys want to make this sound really easy. Google wants to make it sound really simple to go and measure and use their ad manager. And it's kind of like just put in what you want and we'll go make it happen. And Facebook is the same way. And one of the things that Facebook is saying just leave your budget the same every single day and our algorithm will adjust and make sure that we're spending it the most efficient. What they're not telling you is … Doug, this is happening at scale, massive scale.

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

HOW TO GROW YOUR AUDIENCE WITH MACHINE LEARNING

[just click to tweet]

HOW TO GROW YOUR AUDIENCE WITH MACHINE LEARNING

And it's not just how to grow your audience, it's also finding those interests that actually work for your audience and your brand.

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

Doug: Yeah, trust me. Just let me use your credit card and I'll spend your money the best way that it will benefit you, trust me.

Patrick McKenna: And I'm talking massive daily budgets. We'll go in and we'll analyze it, and we'll be like, you're selling like five times more on Friday than you are on Monday. But why do you have the same budget on Monday? You need to be heaving up big time on Friday, and you need to lower significantly this budget on Monday. And then it's like, well, that's not what Facebook told us. And it's like, well, I don't care. It's what the data is telling you to do, and you should go do that.

Doug: Well, and that's the good news is the data is speaking. That's the challenge with creatives. We sit in a room of creatives and we come up with an idea, this is what the campaign should look like. And then like you said to your point, we run multi-variant testing and we find out that this headline with this image and this call to action, this design outperforms what we thought would work. But who cares what we thought? What matters is what's moving the sales dial.

Patrick McKenna: And it's so hard to take yourself out of that, right? To use the word I when you're in that situation. Well, here's what I think. You can't use that word, you got to let the data tell you where to go.

Doug: Well, we've had that conversation where we've sat in a room and their VP of marketing says, “Well, my clients aren't that sophisticated, they won't find me that way in search.” We're saying, we're not speculating, we're just sharing with you the results.

Patrick McKenna: Yeah, we're not speculating.

Doug: He's arguing, “No, no, my clients aren't that smart.” I'm going like, “This isn't a debate, this is how they're finding you.”

Patrick McKenna: It's in the data.

Doug: Yeah, it's in the data, and we should do more here. And he's going, “I totally disagree. It's like, okay.

Patrick McKenna: Yeah. It's really hard to win those battles too when you've got, what it sounds like, someone who's got a really strong opinion about their customer. And to get over that, you're just going to have to keep presenting them with the data until he comes over the line or maybe switches jobs.

Doug: Or we go back to like you said, are we having fun? And if we're not having fun, then we move on.

Patrick McKenna: Yeah. There's that. Yeah, right. There's always those guys too.

Doug: We're not having fun. What are you most excited about in this space that you're in? I mean, AI and this whole world is changing and evolving so quickly, so what keeps you up at night that you're so excited about?

Patrick McKenna: I get really excited about this direct to consumer situation. We call it the DTC, the dollar shave. Gwyneth Paltrow's brand, Jessica Alba. I think you're going to start seeing a lot more of that where you're just direct to consumer. And I think that's really good for the consumer, and I think it's really good for entrepreneurs who want to start a business, I mean, try and imagine being having an idea for an energy drink or an energy bar or a makeup line or whatever 15 years ago. I mean, it's impossible because the CPG companies that were already there had the entire supply chain and distribution chain locked up. And now with marketing and with influence and with scientists, you can actually get a product that works well in the market to your audience. And I think that's really exciting.

Patrick McKenna: I'm sitting here staring at a bottle of Bulletproof on my desk here. This is just a guy that came up with a better mousetrap and sold the crap out of it. I got all the props in the world for that, that's the American dream, and that's better. It's better for the economy and for the consumer to be able to be competitive with someone who's got an ad budget that's $2 billion and a supply chain that they've had for 100 years. I love this, I'm super excited about that.

Doug: Yeah, I agree. I think that there's probably never been a better time in the history of the world to start a company. I think I would be more afraid if I was a major brand than being an entrepreneur launching a new product. The major brand is set in these ways, they've got the legacy systems, they've got all the hierarchy. They're slow to move, they're slow to innovate. Somebody new can come up and just eat your lunch, and you won't even know what's going on.

Patrick McKenna: And they're buying in. Those brands are buying into these marketing gurus, entrepreneur, DTC guys. Unilever bought Dollar Shave. Look what Gillette is doing. Let's be honest, buying a razor, they lock to the store. We've got a razor that's behind a door that someone needs to unlock. That thing, you can make in China for next to nothing and it's locked behind a thing. And now you've got Harry's and dollar shave, and there'll be more. And honestly, I think that's what's keeping inflation under control on the product side. I agree with you Doug, this is a great opportunity. It's a great opportunity.

Doug: So moving forward, is there anything that you can share kind of what your vision is to next steps in your company? Obviously, we're not looking for you to give away your trade secrets, but you guys have been growing at a crazy fast rate, getting all sorts of recognition for that. So where are you going next?

Patrick McKenna: Yeah. I have a group of entrepreneurs within the company. And one of the things that we do is we talk to other entrepreneurs every day. And we do that to sort of get out. What ends up happening is you get so into your thing that it's really hard to keep your head up. And so I like to have what I call a little team of people that has their head up. We look at a lot of things. I can't share specifically where we're going, but we are very interested in this DTC side of things. And we already know that we perform really well on the advertising side. I wouldn't be shocked to see us explore that more from a partnership level instead of just, hey, maybe you should choose us to do this for you, run the ads.

Doug: Yeah. That totally makes sense. I just talked to somebody who's that's their whole business model. They've given up on marking up the ads, and they're doing just strictly a percentage of sales. They're going, we're so good, we know we're good. We're happy to work with you, we're not going to mark up the ads, but we want X percentage of sales. And he said their business income is gone up by tenfold.

Patrick McKenna: Yeah. And that's been around for a while. That's kind of about affiliate type marketing. I think I'm more on the, hey, what can we own together? Kind of thing.

Doug: That's really cool.

Patrick McKenna: You know what I mean?

Doug: Yup. I offer I'm going to introduce you to somebody I think that you'd really enjoy because he's in a similar space, but just on the social side. I just talked to a guest recently in my podcast that's doing AI for CRM.

Patrick McKenna: Oh, wow. That's really interesting.

Doug: It just blew my mind. It's like, what do you mean I don't have to log into the CRM system? Well, he says, it listens to your phone conversations, your email and all your interactions. And based on the conversation, it starts to build intelligence on your sales funnel, your sales cycle. And then it sends you reminders like, hey Doug, you should talk to Patrick today.

Patrick McKenna: Have you been watching what's going on with … Are you a Gmail user?

Doug: Yup, I am.

Patrick McKenna: They're writing the complete sentence.

Doug: I know, I know.

Patrick McKenna: I love it. The world has turned into you got to do more for less. And that's full circle on this Ai stuff. We have to do that. I have to have a machine that outperforms the people that were doing it when we started in 2013. I have to have that. Otherwise, my business is not going to be future proof. You look at Google and this guy with the AI CRM, I need to be more efficient. I have to be personally more efficient with my time and my day because I got a lot to accomplish. So I love that stuff, I live in that stuff. I think it's awesome.

Doug: Yeah, so do I. I'm just amazed every day I log into Google and I see new stuff coming up so I've got the three choices at the bottom based on my conversation that I can auto-reply to, you know, sure, I'd be happy to or we'll do things or I haven't heard anything yet. You're right, it's using some intelligence. But to your point, it's not reducing my team. I'm not laying people off because AI is helping to fill in the blanks. It's giving me more time to spend having a conversation like you and I are having, which is a lot higher value than me typing out a whole email.

Patrick McKenna: Right, you hit it right on the head. I've got to get more, I've got to get more done. And I think that was the whole intention of the invention of the computer. I'm older, so back in the day, you'd get a computer and you're like, “Am I really accomplishing things faster? This thing is crashing, and I got to reboot it and turn it on.” Now, with the cloud and everything else, I'm super excited about what's going to happen in the future of just my own personal productivity.

Doug: Yeah, I'm a big Tim Ferriss fan and the 4-Hour Workweek book kind of wrecked my mind when it first came out. And so I'm thinking, wow, how cool is that now that I can operate my business almost 100% on my mobile phone, which means I can be anywhere in the world, and I don't need to be in an office that I can do everything with a handheld device?

Patrick McKenna: Are you on a four-hour work week?

Doug: No.

Patrick McKenna: Okay. I mean, I just can't even get past the title. I've seen Ferris on YouTube and doing body hacking stuff. I'm on like an 18-hour day.

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

HOW TO GROW YOUR AUDIENCE WITH MACHINE LEARNING

[just click to tweet]

HOW TO GROW YOUR AUDIENCE WITH MACHINE LEARNING

And it's not just how to grow your audience, it's also finding those interests that actually work for your audience and your brand.

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

Doug: I think that the title is definitely clickbait, but I think what it did is it kind of leveraged the idea of the old school. If we compare where your business is today with AI and you look at the way the world thinks, hey, get a good education. I've got three kids and I've got grandkids now. So get a good education, go to school, get a job for 40 years and you'll have a nice pension. Well, that ship sailed a long time ago. His book really just pointed out, I think that you don't need to be chained to your desk in a physical office at your employer's beckoning for 40 years. If you can get your work done in four hours a day or five hours a day. I don't work as many hours as I used to do because I fit the gym into my day. I don't fit it, it's in my schedule.

Doug: I'm two hours a day. By the time I commute, get to the gym, get home and get back to work. So big deal working in a traditional sense didn't work. So you're right, I don't think there's anyone including him working four hours.

Patrick McKenna: If he's working four hours a day, that guy is accomplishing a lot because people talk about it from all over the place. But yeah, that makes perfect sense. I totally get that. And I did a lot of that when I was at Microsoft with a smartphone. It's really nice to be able to have the kind of power in your hand that you have access to. And you can get a lot done doing that. I'm more excited about auto filling my emails right now.

Doug: Just make sure you send Google a note and say thank you. Let's drill down specifically onto your platform and your solution. I just want to make sure that I understand our listeners get what your offering is. So tell me exactly before I screw it up, what it is that we're buying into.

Patrick McKenna: Yeah. For YouTube, what we've done is we've taken that over to the DV360 part of YouTube. We've taken it mostly out of AdWords, we have some clients that are still legacy that use AdWords. And we built an essentially a self-serve platform that will automatically plan with just a few inputs, your media. And then it will build that really extensive multi-variant testing strategy. So to give you an idea, if we are doing it by hand, maybe we set up 40 to 100 strategies. When using the machine, I've seen it spit out as much as 6,000 strategies. When you're talking about a strategy, each strategy has an age group, a demo, a keyword, a target, an interest, whatever, each one separately of that. And then it's got a budget. When you start talking about managing several hundred to thousands of strategies, each strategy has a budget.

Patrick McKenna: We make a call into DV360. We build the plan after you give us a few inputs, we build a very extensive plan. We put that plan into DV, and then we call DV360 two times a day. And it's going to come back and say, hey, this is what's working, this is what isn't working. And it's going to readjust that plan, turn those things on and off, refactor all the budgets and offer some new strategies. Maybe that 6,000 goes down to 3,000, and it just keeps going and going, and going. One campaign that I saw last week, we did 6,600 optimizations on a two-week campaign that was $50,000 on YouTube. It's just humanly unattainable, right? You just couldn't.

Doug: It's impossible.

Patrick McKenna: You could not possibly do that.

Doug: I don't think even Elon Musk could do that.

Patrick McKenna: You could not do it. And you're talking about moving a few dollars into one and out of another, you would screw it up so bad. That's what we did, and that was to sort of move that bar again too, we're seeing in some cases 70% to 45% savings with our fees. We did it again. We came in the market, we're about that. And then people started to kind of like catch on a little bit, and then we did it again. That's where we're at on YouTube, completely self-serve. But you got to have a DV360 seat. On Facebook and the other social platforms, it's more like audience mapping. Let's say you're in a … We would go analyze your account, we'd look at what audiences you're using. And we would suggest similar audiences, and we would add those strategies into your Facebook account. But that's a more manual process, we haven't completely automated the Facebook feature of the platform. And that works across all sources, so we can do that on Snap and Pinterest and LinkedIn and Twitter and Facebook, all the major platforms for audience mapping.

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

HOW TO GROW YOUR AUDIENCE WITH MACHINE LEARNING

[just click to tweet]

HOW TO GROW YOUR AUDIENCE WITH MACHINE LEARNING

And it's not just how to grow your audience, it's also finding those interests that actually work for your audience and your brand.

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

Patrick McKenna: And then on the last thing, I think we talked about this earlier. There's a lot of creative swapping on Facebook, and so we have some trade of exhaustion tools that will say, hey, this audience is not responding to this ad because they're exhausted from seeing it. So you got to swap it.

Doug: Let's move on to the most important things, where can people track you down and find you?

Patrick McKenna: strikesocial.com, and then Patrick McKenna on LinkedIn. And I got a Twitter account and all those places. But we've got a really great Drift bot, I like those guys a lot. If you're not using that, that's a good one. And you can reach me right there on the website.

Doug: Excellent. Who's one guest that I've got to have on my podcast?

Patrick McKenna: You know what, I've had a lot of experience working with this guy. His name's Dylan Conroy. He's at a company called Social Standard, and they do influencer marketing. And the guy is just the consummate professional, he's got his own podcast. I think you two would have a good time kind of interviewing each other.

Doug: Okay. Well, Excellent. Really appreciate your time. This stuff, I'm so excited about it. I live in it, I know our listeners. I know you guys don't all live in this space, but I'm always looking for what's new and how can I move the sales dial. And this is where we're moving with our clients because this is what's moving the sales dial and working, like you said, optimizing the budget. I mean, people have less, in a lot of cases, less budget to spend. And they want to get the best results, so I would suggest head over to Patrick's site. Go take a look at their website, sign up, make sure you sign up for their newsletter on Strike Social. So thanks Patrick for taking the time.

Patrick McKenna: Doug, thanks for having me, it's been great.

Doug: There you go listeners, this is another episode. Hopefully, we stretched your mind and we've given you some new ideas you can take back to your boss or to your marketing team and have a look. And I look forward to serving you on our next episode.

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

HOW TO GROW YOUR AUDIENCE WITH MACHINE LEARNING

[just click to tweet]

HOW TO GROW YOUR AUDIENCE WITH MACHINE LEARNING

And it's not just how to grow your audience, it's also finding those interests that actually work for your audience and your brand.

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

Get in touch with Patrick:

Find out more about Patrick:

Links to other related podcasts and or blog posts:

IMPROVE SALES WITH AI POWERED CRM

HOW TO IMPROVE YOUR SALES STRATEGY