181: How To Use AI In Advertising To Accelerate Growth

Yeah, you didn't even know about some of them things.

frankly, health robots. But over the last 15 years, I focused a little bit more on building and releasing lots of different products and platforms, whether it's applied AI systems, ecommerce platforms, big data systems, consumer electronic products are consumer goods. And throughout that journey, you know, I'm an innovator like to build and design and actually release products. But the most important step, after you've designed the product is actually getting it into your customers

hands. And that part of it is really complicated. It's more complex than most people think. I mean, with your background, Darrell, you know, that you got to figure out the playbooks, you got to figure out who the audience is you have to have immediate strategy. And that's before all the legwork that has to come right after. So going through that so many times releasing about 11,000 different

product SKUs. I came to the realization that this is a huge problem that gets in the way of midsize businesses, small businesses, big businesses have a ton of money, they can hire the best talent, but not everybody's got that budget. Not everybody's got that time. Not everybody knows what to do. So we've build a tool or platform, an AI layer that sits on top of your marketing stack that helps you to take a look at what's going on, make the right decisions and optimize your own

results. So that's basically a through line between where I was and where I am. And frankly, where we're going.

handle 120 tabs. But I guarantee you, you probably have a method to your madness.

You know, we're talking about project management, Slack and all those tools, and then have one for me time. I've got my personal email on there. Anything My wife wants me to do open too many Home Depot tabs. That's what that one is for.

other areas of their life. But talk to us a bit about that journey 15 years ago into now.

intelligence in them. Now take it as you go forward, I got into my you know, doing my undergrad I went through robotics courses, talking about control systems, and feedback systems and the neural networks. Before it was called machine learning, it was just called back propagated neural networks where the depth of the system was determinative

of the outcomes you had. So I was building those kinds of systems, you really just for robotics purposes, to make sure that by robots could identify patterns and follow through paths. But as time has gone on, as compute power has gone up, as the GPU is a good lesson, like kind of overpowered, you're able to use it now to identify detailed enough patterns to know what objects are on the screen to be able to do speech to text transformation. I mean, that's,

that's not too complicated. But it's complicated enough, that if you're an Indian accent, versus an African accent was an American accent, you know, the systems have to be smart enough to know that you're all saying Netflix, right? You know, and I can tell that at home because when I say Netflix, and my wife says it, the remote still doesn't catch when my wife speaks. But the whole point is you've gotten AI systems like that, but we've evolved even

further, right? We're talking about generative additive

intelligence networks. We're talking about adversarial networks where you've got two AI's that battle each out each other out to figure out who's got the less error and usually that's what is used for things like Ali or even GPT-3, all of these more advanced systems to give closer to human level outputs in terms of creatives, text, decision making, chatbots chatbots actually fairly easy in the sense that most chatbots end up being heuristicly driven, right, there's some sort of

input. And there's a specific output set that they're, that they use. But compared to more complicated AI systems, whether it's navigation systems pattern, match recognition, object identification, GPT-3 type creative output systems, those are quite a bit more chunky, quite a bit more delicious. But for me, the long game is based on this realization, and frankly, it's the same thing with every machine. And AI is just another kind of machine and everybody knows what machines

are. They're things that humans make to make our lives easier. Even Microsoft Word can be called a machine, if you think about it in a certain way. But if we thought talk about AI specifically, we're using AI's now to make sure that we're eating properly, tracking

health. AIs are helping us make sure that anything that left for example, Amazon, they've built intelligent systems, so they know exactly when to order, and move products around, so that they've got the product in the warehouse at the right time before you go ahead and buy it.

But going into the future in the space that we're in right now, in this marketing and sales and growth oriented space of businesses, there's a lot of opportunity for using AI to make life easier for people whose responsibility is to help companies make revenue. So whether it is automatically generating textual content, or blogs, or whatever it is, if you want to do the SEO, long game, or even some startups whose jobs is to combine data sources, and spit out content using good

intelligence systems. But ultimately, for me, the goal is what I call the growth coordinator, the growth conductor. And this would be a super AI. Well, not super, let's not use that term, because the AI sticklers will chew me up if I say super AI, but this will be the intelligent system whose job it is to augment the growth manager, the CMO, the CSO of any company, so that they can ask simple questions like, I've got a million dollars, how should I

spend it? And the test system will tell them, where to spend it, how to spend it, and even kind of be like a coach right over their shoulder so that most people, agencies, users, companies can have a highest chance of success, regardless of what they're trying to do, regardless of the product you're trying to sell.

Because certain kinds of you know, if you look at everybody's day, there's a bunch of busy work that we always end up doing, that turned out are easy to automate. Now, there's a difference between an AI doing the automation and then, a robo Robo task worker like UiPath doing the work. But ultimately, you know, RPA, robotic process automation is doing that work. But it ultimately boils back to the same thing, as far as a lay user is concerned, computer

getting my work done. The truth is, in the long term, I think it's not going to be an issue, as long as we have a specific ethical framework in how these systems are built and released. And I'll be clear on that. If we make it so that just a few rich people have all the intelligence systems, the whole world is going to be poor, and there'll be like 1000, rich people. That's literally the world at its limit, call it 100 years

from now. But instead if we make it democratize, democratize, or we build for the mid market and below, what ends up happening is more people can use this AI of virtue of AI of this power of AI to meet their dreams without having to go through years building the experience that they would otherwise need. They would need to build smaller, highly focused, highly motivated teams, instead of spending 10s of millions of dollars hiring 20-30 people because they're just not sure which way to go.

It will really help eventually make individuals hyper powerful, but that's only if the way you release the AI's is so that these people the midmarket and and below have access to it. If it's only enterprise, if it's only the richest people, then you'll end up where there's a rich 1000 people and everybody else is a serf, or worse.

mover advantage. It was, you know, he didn't get to until about a year after I don't know if that was due to restriction or?

played with it. But we didn't want to roll fit into our platform, because we're doing something that we think is eventually going to be even more valuable to businesses. But absolutely there lots of companies now using GPT-3 as the basis of their business where they'd like Jasper or copied on AI, or you know, like four or five others who are in the general copy generation space in the content, civilization space.

And it's a big deal, right? You know, at that point, the person who has retrained in GPT-3 best gets the best outcomes. But like I said, not every AI is good, you know, frequently there's a human that needs to do a review step because frankly, humans are still the smartest creatures on this planet. Don't let the computers fool you.

And all three, this recent round that they built, that they got isn't the only round, you've got a few other rounds in the past. But surely that's sort of a team building a great product, pretty excited about where that's

you talk on Reddit. Right? You know, it's actually all different formats. And so they've got a niche to chasing it. As long as other people in the space think about some niche, they can go crazy, or maybe even law. They can, they can definitely win and win very well.

identify behaviors. With those attributes and behaviors, you can literally make configurations. So, for example, our intelligence system allows customers to have their own AI brain per customer. But that's because we've been able to identify, like I said, the attributes and the behaviors, they give the outcomes that the customers need. And all they need as a big input at that point is their own specific

business parameters. And this is a really key thing, when it almost all of aerospace, I know, for a while, you know, we would make single use rockets and throw them in the ocean. But a lot of aerospace engineers have always looked towards the holy grail of reuse, because we've

seen it everywhere. You know, Aerospace is not very different from almost any other kind of industrial strength engineering work, whether it's automotive or others in the sense that you've got to deal with a lot of complicated systems, simplify them, put them together at high quality, high reliability. And it's not different in aerospace. And it's definitely not different in the growth automation space.

usually 10-12-15 passes. But what that does is it gives you this muscle this memory that says, Okay, I'm iterating towards the goal. Goal doesn't have to be absolute perfection, you have to meet standards, you have to meet deadlines, you have to be on budget, so you can go for perfect. But you got to learn how to iterate quickly. So almost every kind of failure is another opportunity to learn and

iterate. And as long as you keep that top of mind, it will help you as an engineer, but also help you as an entrepreneur. Because if you plan your business according to iterations, whether it's sprints, or some other mechanism, then you can use a budget efficiently, you can hire the right time. You know, you don't want to wait until you're already selling too much before you hire a support person on your delivery team or an extra

developer, for example. But you want to be able to iterate, learn something key and use that key thing as leverage. And the best way to do it is learn the hardest things first, because the things that you already know how to do you need fewer iterations to get those right. However, on a personal side, when I'm when the stress gets really, really high when you're staring into the abyss and chewing glass. I take a nap on my tummy.

your DNA. Interesting to use that word in your company, literally, that and DNA in my world means definite natural ability. That's how I approach DNA. In other words, when I look at company culture, and I look at the entrepreneurial spirit, and I coach and lead growth for a lot of companies, or help lead growth for a lot of companies, I'm looking for the definite natural ability, because it helps us with our marketing, it helps us with our angle, it helps us attract that buyer

persona. But I'm asking you specifically how do you relax and get away from all of this?

lot of time. So in the mornings every day, I've got an hour and a half that it's just for getting the kids ready for school, but also we can talk we can do homework, we can kiss we can play and then the evenings. I just loved that whole timeout so it can pick them up. Take them to anything they want to do. We can watch TV together. We have a garden in my backyard. I planted 24 different crops this summer and me and my kids we planned We tend to we harvest

together, right? So right now the only thing left is our carrots. But we literally have dozens and dozens every day, we just walk outside, but I'm like five carrots come back in, and we're good to go. But these are the kinds of things right like, I love using my hands, but I love using it with my family. Whether it's on my, on my farm in the back of my house, doing renovations at home, or just playing, that's how I relax.

advertising quickly. And we started that off, we got integrations with Google and Facebook. But then you bring it out to customers. And you start to see that each business has a different kind of playbook. That is more likely to give him the best outcomes, whether it's a combination of advertising and social or SEO and ads or whatever combination of things, conferences, webinars,

partnerships, and so on. And we didn't want to go to up on all of that stuff right away, we figured that there was something more to do. So we built the first version of our

intelligence system. It's our Objective Rank AI engine, it's a time oriented regression model back propagating regression model, they uses about 11 different dimensions that are visible or useful in the advertising space, we're talking about timing, location, targeting details, keywords, behaviors, on site behavior, how many times they visited things on your site, and so on. And we built a ranking engine that basically says, Find me the needle in my haystack of data that I should be spending my

time and my money on. We built that system. And we're able to really quickly testing and our customers give people 2-3-6 times better results per dollar, in a handful of weeks. Now, that was good. But that wasn't the

end of that journey. Turns out that if you're a lead gen company, versus an E commerce company, that a few digital steps that exists for E commerce that don't exist for Legion, right, so if you're helping, like a health provider or something, you get them leads, but then what right, you know, if they don't have a really disciplined sales team, you get blamed for failure. And that kind of stuff, we had to go

through some of those pains. And so what we did instead was we saw that as an opportunity, right? Well, like I said, we're iterating iterating iterating. It's not all Eureka once it's Eureka one step at a time. And we found that for lead gen type companies, turns out what they actually need is some help on that part. Because what is it that businesses need revenue growth, that they're not doing it for, for sale, sales boils down to money in the bank,

right? And so we decided that we will help businesses get all the way to money in the bank. On ecommerce that's easy enough to do you integrate with the 26 different places that were integrated in, whether it's Google, Facebook, Snapchat, Instagram, Twitter, LinkedIn, Amazon, and so on, plus your on site analytics, we can use all of that data to help you to find the right audience, the right marketing mix, pick the right budgets get to the right

answers. But on Legion Post lead capture there was a hole and so we built an automation system that basically says you can lead, let the salesperson know right away so they can close the deal now, and let the customer know. So that relationship can be had. And that allows customers to go from 90% lead loss if they wait one or more days to be more like 40% lead last because at least now you can have that conversation. It's

a huge difference. We're talking about, you know, two to five times better lead qualifications so to say and that allows businesses to take their existing processes and just get better outcomes from it. Well, once we did that, too, and I'm

just keep going. I took those in those insights and we decided to expand because whether you're an E commerce business or lead gen business or brick and more business and enterprise business, there are certain few points that we've noticed some attributes and behaviors that everybody does. And that is there. We've identified 17 workflows, strategies, one, research, publishing, monitoring, optimization, attribution, follow through click fraud, and so on, we've

actually automated 13. But the remaining four other big matters of the space, the remaining four are content automation, which we don't really want to do, one, because there are other startups that might be doing it, so rather partner with them. But too maybe it's too much of a cake to eat. But then number two, the one that we are really excited about a strategy

automation. And this is the idea that for most businesses, they can tell the system Hey, I've got to do an awareness goal, a sales goal, conversions goal, loyalty goal, retention goal, what should I be doing to make this work, and the system, we're finishing up right now will actually be able to tell the customer, this is where you should run ads, this is how much you should spend, this is how many campaigns you should run through the AB testing to make sure your content is perfect. On

site and off site. And these are the things you need to do to scale to get you there. Oh, and by the way, I'm holding your hand all the way through. That's where we're taking our system all the way to. So that goes from being a growth, advertising growth platform, to really being the growth coordination platform I talked about earlier, where you come in, you answer a few questions, the platform tells you how to hit you goal. And a hits it for you or with you, as you go all the way through.

And then it goes. So whether it is in the agency world, where you've got either your team is doing a crap ton of work, but it's not so easy to scale without adding a lot more people. Now you can kind of have almost infinite scale, you know, in your company. If you're a small business, and you don't want to get started, I mean, we'll make your pay. But at the very least you can set the right expectations. There's nothing worse than a small business

coming into the space. And thinking that, hey, I've got I'm gonna make a million dollars and I want to spend five grand in ads. That that's not a thing. So all these expectations, setting issues that are hard to report, hard to speak about. You can have a system at least bubble some of it out. And then you as a person will take the guidance, but really steer the ship all the way to your goal.

process once again. It's the way it works. And it doesn't matter what platform you're on. We just didn't have to go get an engineering degree to get the know,

And he literally said you have to hire it was his word and his framework, but he says you have to hire Strategy three level thinkers. What he meant was you have to hire people that would have ordinarily been lawyers, doctors, CPAs and engineers is in his classification. Yet everyone shows up in the digital marketing industry with no degree you don't have to have one. And but what's interesting is you can learn the vernacular and you can take all the certification programs you want

to. But until you've been in the seat to do all the things we talked about strategy, plan sizes, contents, moments and process, execute, experiment, take the losses, the learnings come back and prove it. And the client wants one thing, like they give you a check, or they give you a credit card and they want leads where they want sales, e commerce or lead. Yep, that's exactly. They don't care about anything in the middle.

And to do digital well, is what I hear you saying is you're you're helping those of us who are committed to the game, do it better and easier. And then the piece that I want to go back to that you said that I want to make sure it's highlighted. And that it isn't to replace a human it is though, to add exponential capabilities to the person who already has a great strategy. Absolutely. And then assisting with the coordination. I love that.

in the right places. There's a lot of amazing stuff going on. And I really invite everybody to come take a look. But frankly, we you know, we're really appreciative for the program we've done with Google. They've got, you know, it's Google right. They're they're one of the only, or alphabet the parent company is one of the only trillion dollar club companies left. And we're getting some of that trillion dollar knowledge given to us as part of the programs. It's fantastic.

marketer? Or could this tool be used for someone just entering the industry? Tell us a bit about who your ideal user?

more advanced people. But actually, right now, you know, we are working on ways to make it a lot more self service, a lot more intuitive, a lot easier for a layperson to do the connections and start getting magic upfront without anybody on our team helping you along the way. So right now, it's more for intermediate and senior strategists. But over time, we do expect that the more SMB oriented people will use it, just not right away.

do as much of that work. But then going into the other part of the question, which was, you know, what is the ideal breakdown for goal hitting? It really depends on the business, right? Like I mentioned, what we've seen is you know, if you're if you're in, like a real estate industry versus manufacturing, the kinds of customers that are available to

you are different. And so the part of the strategy work actually includes finding the right customers that are hyper relevant to you, you know, like Account Based Marketing type of preamble work might be necessary, but if it's more ecommerce, you can usually get a bit more general. And that is just on the ad side. There's also the part of, hey, you know what, I've got my email list, who should I be sending anything to right now got a lot of tools

helping with that. But by tying into our platform and your Analytics, you can tie it all together and literally see ad to purchase through email attribution, that you wouldn't be able to see right now. And so give you some of that, you know, some of that magic asap from a reporting standpoint, but then when you're not looking, it's working 24/7, 100 times per day, looking at your data, shutting down the stuff that doesn't work, boosting the signal where

it's good. So you can give higher and higher ROI ROAS as you go. That's really our goal to make sure you can keep making dollars while spending spending as much as you need to and up

much more. Now, you know, to be very clear about the first piece, which was one of the average budgets that we usually see, we usually prefer and I have to use that phrase in customers, we're spending 30k or more a month on ads, but we have helped people as low as $500 a month on the SMB side is just there's some challenges in the SMB side, but $500 with a good strategy, you can still have some wins, maybe just not make a million dollars yet.

grow and do better. It's I've been an entrepreneur and small business owners, you know, 32 years now. So last 12 years blessed to help hundreds and hundreds on the agency side and easily north of 1000 to 2000 on the coaching and consulting side whether group or individual so amazing really appreciate the work you're doing the backstory the background

to talk to him. And I know what it takes behind the scenes, even though I don't know what you do and how you do what you do as an AI and engineer. I respect the journey to do 11,000 products. I respect what happened before the 11,000 got to the show. I res- I respect the iteration, ability and capacity to say, Oh, it didn't work. Let's do this. Oh, here's what we learned. Here's a drop this add this. Let's go try

again. And I respect that with any entrepreneur, software or otherwise, brick and mortar or retail, e comm or medicine. There's just something about the entrepreneurial DNA that matters to me, and I'm super passionate to host these conversations. And I'm again, excited that you took some time out to join me today and share with our audience. I gotta ask this last question

before we run got to last us. I know you probably got 8 million things you want to accomplish before you finish but if for whatever reason, if for whatever reason today it was to come to an end on this planet. What would you want everybody to remember you for?

advertising. So if we can have a platform that takes you all the way to your dreams faster, that's really what we're building. Because with that, you can have a better world. Take that and maybe use that as inspiration to go out and build something that people want to make the world better.