Google AdWords Insights and Strategies from Mike Rhodes

Display is certainly a lot more nuanced, you know? If Search is a game of checkers, then Display is a game of chess.

But each keyword has roughly the same number of people searching. Which is what makes the whole game work. And you’re not going to change that number dramatically from one day to the next. You’re not going to cause ten times more people to go search for beard oil tomorrow.

But on the display side, so 2 million plus websites out there, everything from Oprah, CNN, and ESPN all the way down to tiny little blogs and forums that you or I have never heard of but some of them are all about beard stuff or beard oil.

And you can basically go teach Google to go find those sites or even just pages about that stuff. It could be an article on The New York Times talking about how to care for your beard. You can tell I’m an expert on the subject. This is a year’s worth of growth for me.

And tell Google, “Go find this content. Go find these articles. Go find these pages. Even go find people who are interested in beards or interested in baseball.” Based on their behavior over time, you can tell Google to figure out who those people are and follow them around.

And then, you’ve got these other 2 million sites. Again, some of them, enormous sites. But you can just be running ads and all different types of ads. And then, you’ve got YouTube, and then you’ve got apps, which is an area that nobody goes and plays in. You can put your ads inside other people’s apps, whether you have an app or not.

So, from a branding point of view – not that either of us are brand marketers.

(RELATED: [Parts 1 & 2] The Display Grid: How to Scale Your AdWords Display Campaigns Profitably with Laser-Focused Targeting and the Right Choice of Ad Type)

(RELATED: The Remarketing Grid: The Science of Ad Retargeting Audience Segmentation)

Remarketing isn’t just about where the people came from. I know people think, “Oh, but I don’t do much Search, so there’s no point in doing remarketing because there wouldn’t be many people there.” But you can remarket to everybody that came to your website. If that’s what you want to do.

(RELATED: How To Write Blog Posts That Sell)

Yeah, I think if you’re going to lose a chunk of money month one, it’s much easier for a lot of people to go, “Well that doesn’t work. I’m going to go over here and master Facebook.” And if Facebook’s working for you, that’s awesome. Do it.

But I’m a big believer in not putting all of your eggs all in one basket. Maybe only two or three baskets. Don’t try and do 15 things and don’t run off and do Pinterest and this and LinkedIn and do everything, but at least diversify it a little bit.

Now, that’s somewhat safe. If you are doing this for clients, that can be an okay place to start because you can tell your client with some certainty, “This is where your ad is going to be seen.” Whereas with some of the other methods, you don’t actually know where your ads are going to show up, which can cause some issues sometimes.

But because you’ve never tried all of these other different ways of targeting, you’ll never know that maybe you can get them for five over here, and ten over here. It’s all about testing, and testing, and a bit more testing.

(RELATED: [Parts 1 & 2] The Display Grid: How to Scale Your AdWords Display Campaigns Profitably with Laser-Focused Targeting and the Right Choice of Ad Type)

So, you have to feed that into Google and it’s a little bit of a roundabout way of getting all of that data inside of Google, but once Google has that, then Google is going to decide when somebody comes searching for, I don’t know, “Buy beard oil,” Google’s going to say, “Oh. Hang on a minute. There’s this merchant over here that’s got all of these products that are about beard oil. I’m going to show that graphically, I’m going to show an image of the product, with the name of the product and the price.” And you typically will see five or six or eight of those across the top of the Google Search results.

The artificial intelligence that is being leveraged on the AdWords platform is changing. Is it not? Can you talk a little bit about that?

A.I. is going to change everything. Not just AdWords, it’s going to change how agencies work. It’s going to change everything.

From an AdWords point of view, Google bought this little company for a tidy sum of about 480 million, I think it was, called DeepMind in England, who had about 110 of the best A.I. machine learning dudes on the planet and they snapped that up for half a billion. It started off playing Atari video games and it would basically play the game and learn to play the game just by trying different things and seeing what happened to the score, but it would teach itself.

It’s this complicated game called Go. It’s an ancient Chinese game on a big grid. There are more possible moves in the game of Go than there are atoms in the universe. It’s a complicated game. They said, even 10 years ago, it would be impossible for a computer to beat a human at Go. It’d be impossible to teach a computer how to play this thing.

Well, Google DeepMind built a computer that taught itself how to play. They fed in every move of every game ever played and it went, “Yep, I’ve done that. Now what?” They basically had to build another computer so they could play each other 100 million times, so they could both get really, really, really, good. Then, they played the world champion and everybody predicted that the computer would get smashed and it won five to one.

(RELATED: 3 Advanced AdWords Tactics That Increase ROI)

There are even today some places where you can almost go in and hit the “easy button” inside of AdWords to allow Google to basically control your spend. What are those different opportunities?

Bidding is probably the easiest one. You need a little bit of data first. So, it’s really good to understand how the machine works and to do it manually first so that you’ve done grade one and grade two before you let the computer take over. It’s good to know the basics.

But then once you get a little bit of data, it’s like, “Hey Google, I need $7 leads, please. Off you go,” and it’ll take a week or two to learn and sometimes it gets it horribly wrong, but that’s less and less and less these days. It’ll start getting you leads. There’s a new model that’s coming out later this year, which is pay per conversion, so not bid per conversion, but literally only pay.

This new model will be: “Did you get the lead?” “Yep.” “Okay, 10 bucks please.” And only pay when you get the lead. So, they’re starting to automate more and more stuff. They will write part of your ad for you. So, that’s on the search side. This thing called dynamic search ads or DSA. So, you basically point Google in the right direction and say, “This chunk of my website, please. Keep an eye on it and if someone comes searching, but I don’t have the right keyword, then you figure it out and you show an ad and you decide the bid and let them see my stuff, please.”

(RELATED: The Remarketing Grid: The Science of Ad Retargeting Audience Segmentation)

I don’t tend to think in terms of percentages. I don’t draw the pie chart and say, “This much here,” because that frustrates me if I’ve got a marketing manager doing the same thing. It’s like, “This much to TV and this much to radio. And digital can have this little slice.” If I can prove that you’re getting a 9:1 return on this little slice, why wouldn’t we have a bigger slice? “Oh, well, because we’ve got this money allocated for TV and this money for radio.” “Yeah, but it’s 9:1. Give me more money and I’ll go get you…”

That requires a different mindset. And for someone my age, if we’re dealing with a marketer that’s a bit older and went through uni 20 years ago, then it might be a little bit different and there might be a bit of an education process there. If it’s a younger marketing manager, it’s probably not going to be the same hassle with that. But test a lot of stuff and see what’s working and double down on that. And then go ask for more money for them to invest more money. So, it’s all about investing. Investing in experiments is probably the best way I can put it.