Duane Forrester is one of those SEOs who every other SEO knows. He is a wealth of information on the subject, so much so that this interview runs a little on the longer side. Trust me, it’s worth it!
You simply cannot teach what Duane knows. Officially, Duane is the VP of Industry Insights at Yext and the former Sr. Product Manager responsible for Bing’s Webmaster Program.
Unofficially, he is one of the smartest SEOs around.
In this we discussed…
- Gen Z and their impact on search
- Understanding the knowledge graph
- Why the messy middle is important in SEO
- The value of vertical search engines
- Bing, does it matter?
I hope you enjoy this as much as I did!
Digital Transcript (edited for readability)
Introductions
Hi, I’m Jon-Mikel Bailey and welcome to Wellspring Digital Chats, where we interview marketing experts, we take their brain and we scientifically extract all of the knowledge and share that knowledge directly with you. It’s not as violent a process as it sounds. But anyway, today, we have a good friend of Wellspring Digital and I can safely say one of the only SEOs who has custom made a guitar that I have actually seen in real life. So that is a new one for me. Yes. So, uh, Duane, if you could, please take a moment and introduce yourself to these fine people.
Duane Forrester: Ah, thank you. First off, Jon, thank you very much for having me on the show I appreciate your time here. So for those of you who don’t know me, my name is Duane Forrester. I’m currently Vice President of Industry Insights for a company called Yext based out of New York. I personally am on the West Coast in California.
I have a pretty long career in the industry, about almost 25 years now. I ran operations for Bruce Clay Incorporated for a little while. Prior to that I ran Bing Webmaster Tools and was, you know, part of the team that brought schema.org out. I ran SEO at MSN prior to that. I used to work in gambling work for Caesars Palace, wrote a couple of books, build guitars for a hobby, all that kind of crap. So I got thoughts, things to say.
Jon-Mikel Bailey: We want thoughts. That’s a good thing.
Duane Forrester: Good news for you too, Jon. If you need to go in and extract anything from my brain, there’s plenty of extra space in there. So easy working environment.
Jon-Mikel Bailey: It’s like working on an old Ford, you can crawl inside the engine bay and go to whatever you need.
Duane Forrester: Absolutely. screwdriver, and duct tape. You’re all set. Good to go.
Jon-Mikel Bailey: Alright, well, naturally, when I think of SEO greats, your name is certainly on that list. And I’m excited to talk to you about SEO, both present and future. And, you know, maybe get into a little crystal ball stuff here. So Whoo. Let’s do this.
Gen Z and Their Impact on Search
Jon-Mikel Bailey: So first and foremost, I want to talk about Gen Z for a minute, you shared an article that discussed what they’re not interested in, which was alcohol, cars, and razor blades, which I thought was an odd combination. But I’ll let you read the article and make your own determination. I’ll post the link.
So do you think that Gen Z is fundamentally changing how search engines operate? In that their search and other behaviors seem to differ vastly from previous generations?
Duane Forrester: I think yes is the short answer. But it’s very complex, yes. Because if you look, generationally, I think what that leap from Gen x, Gen Y, and Gen Z actually showcases is it’s incremental but it was a faster curve than what we’d previously seen. Like you, I’m Gen X unabashedly. And if you look at how I grew up, I remember when our first computer entered my life. And I remember how my parents approached it, and how I approached it.
I remember sitting in an office on a weekend, where I happened to be pulling a shift as a manager, having access to an internet-connected computer, and exploring the bookshop called Amazon, and feeling compelled to buy a book to see if it would actually be delivered to my address. And low and behold, like nine days later, it showed up there and mind blown, right?
My parents, it took them years to get to that point. And so like the leap then was already starting. And I think that this, this article, what it really talks about is like the kind of Nth dimension manifestation, and it happened during Gen Z’s time. Because you had a lot of those influences where I am not a digital native. I grew up pre-digital, and then I adopted it.
And then you know, kids, if I had had children, they would be digital natives, because they would only know that world. Now my kids would probably have suffered greatly because I would be like, “No, you can’t have an iPhone until you’re 18 years old. My house my rules,” like all that kind of stuff, right? Tell your friends to stay off my lawn and the whole thing. I’m easing into this these days and I’m loving it right? Because my lawn is all desert landscaping. It’s full of cactus. So I’m like, “No, no, come sit on my lawn, please,” so taken in a new direction.
But the Gen Z thing is, I think it’s just like maybe the most manifest form that we see today. And it doesn’t, it’s, it’s attached to Gen Z. And there are some times, it feels like we’re calling people out in a negative way when we call out generations, but like, but I feel like they are the ultimate manifestation of something that happened or started after the Boomers. And in fact, you could argue that the boomers had that from the greatest generation and like, like there was that change all the time in the advancement of technology.
But if you follow all of the rules of Moore’s law, and the advancement of technology and the speed at which it advances, then ultimately you see it compressing as we get closer to our near term, think about what this is going to be like for Gen alpha coming after Gen Z.
Jon-Mikel Bailey: They scare the crap out of me.
Duane Forrester: Dude, they’re going to look at Gen Z, and go “here, hold my drink.” And like, they’re going to be a generation of geniuses that we have never beheld in our lives. And I don’t necessarily mean geniuses, like, they’re all gonna sign up at Mensa and you’d be super smart. They are going to be digital problem solvers in a way that we could only dream of.
Because we’re sitting here, we’re living through the growth of machine learning, and they are going to get the manifestation of matured, machine learning, not fully realize machine learning and AI, but they’re not going to be dealing with a three-year-old, they’re going to be dealing with a nine-year-old, which terms will make them geniuses.
So their ability then to roll up to a problem and go “here, solve for x” is like, it’s irrelevant. Those types of things simply won’t be roadblocks, they won’t have to put time into thinking about it. And so I think, you know, I think that the behaviors that we see, I think, if you look at the behaviors of Gen Z, this article calls it out and kind of like a funny catch your eye way, right?
But there’s the larger conversation we’ve all seen going on, and there’s a knock-on effect from this, you know, like, Oh, sure, we make fun of man buns and beards and whatever else, and that’s why the razor blades are in the thing, right? But like, really, that’s the funny part of it, that comes and goes, you know, but I mean, look, you and I were the balance, right? Like you’re not using razor blades, I am using razor blades, right. So we’re in balance here, you know.
But when it comes to Gen Z’s larger outlook, like, take the pandemic out of this situation, all predictions were that our society was moving toward a more urban environment. Because you had generations, two generations, you have millennials and Gen Z, who grew up in school, and then went to their first jobs, a lot of them in cities, they grew up in that environment, they found themselves in that environment, they started relationships in that urban environment, they’re comfortable in that urban environment, they start families in those urban environments, they don’t want to move to the suburbs.
So a 4000 square foot five-bedroom home is lost on them because they don’t need it. They don’t want it. Take the home out of the equation, insert automobile, it’s the exact same thing because I live downtown, I don’t need my own car. That’s added money. That’s insurance. That’s everything. Oh, wait. Now we’re talking about a knock-on effect that hits the insurance agencies and the insurance world as well.
And so then what change does that force into insurance? Do I need an agent in my community? Or can my agent be virtual? Well, if you can pass the savings on to me, and hey, look, we’re having a baby now. So I do kind of need my own automobile, it makes sense to me to have control overcoming and going now that I have a baby. But I still want that cheaper insurance.
There’s so much of that and, and being digital natives, their ability to pick up a phone and vote with their thumbs and show where their wallet is going. It’s unprecedented, and it’s caught businesses unaware just like this pandemic has. It’s forced this real broad change. And I think thankfully because millennials were at the front end of this and kind of forcing that work from home that you know, delivered to my doorstep like all of these things. Could you imagine what we’d be living through right now if these services had not come to be in the last five to seven years?
I honestly don’t know, we would have a very different world today. It would be tragic. I mean, anything from batteries from your remote to food delivery like and yet because of the change in direction, all of these things are here. All of these things continue to move forward. All of these things are in some ways, the savior of businesses today.
I’m not gonna name names, but there was recently an employee of a company who was posting TikToks, who got fired for posting those TikToks. And I sincerely wish them the best of luck. Because I’m scratching my head, I’m like, “you don’t get it, you really don’t get it.” Like you just alienated an entire generation of people looking to enjoy a little change in their life, and they would have given you their money. And now, you look completely detached.
Jon-Mikel Bailey: So, here’s a strange follow up on that. So do you think that Gen Z is informing AI? Or AI is informing Gen Z?
Duane Forrester: Yes. It’s funny, you say that, right? Because like, that’s, I think, the right way to ask the question, because machine learning is constantly learning from all of the inputs, right? And so if you give someone direction…
Okay, I’m gonna pause this here for a second. Jon, have you seen the Social Dilemma? The documentary. So if folks are listening to this, and you have not seen the social dilemma yet, you need to watch it not because it’s going to blow your mind with stuff that you don’t already know. Because if you’re watching this, you probably already know these things.
But it does a really good job of framing it for the layperson, the person who’s not as well-tuned in our industry and in social media in general. Also, a fair few of you are probably going to learn a few things because you think you know, some stuff, but really, this is going to crystallize it.
And so to your point, there’s this cycle that happens, right? Like my behavior reinforces the action that the AI put in front of me, which trains it and trains me. And so when you look at the scale that these entities operate on, the scale is mass learning very quickly. And then they can produce models that are mimicking, but then they can iterate on those models.
So if you can train a million people in a day, because you have that reach, I’m looking at Facebook, I’m looking at Google, Amazon, I’m looking at, you know, Netflix, like all of the large brands, if you can reinforce your model, based on some kind of common sense thoughts that your engineering and teams put into it. The actions of me, the general person in the population, reinforce, corroborate, or refute what those thoughts were that were put into the system that trains the mechanism.
The machine learning then backs away and says “Hang on a second, I now have a model.” The model then can be segregated into an infinite number of variations of itself to test an infinite number of models in the background, just with data modeling. So you need one run with real people to get the input, start the flywheel, then the system can be taught to teach itself with variations. And the variations that start to become statistically viable and useful, are then brought forward with more human input to be then put in front of you and I again, and this is why you’ll go into Netflix, and you’ll be like, “why the hell are you recommending that to me?”
Because you’re thinking there’s, like, I have zero interest. And then the third or fourth time you see it, you pause and you let it pre-roll? And you’re like, “Ha, I’m not. I’m not abhorred.” And then you read the description, and you’re like, “wait, is that Oh, I thought it was as an actor?”
And now you go in, and then you give it time, and you put it on your watch list. And then you watch an episode. And the reason that that happened was because you watched the movie twice in the last year that had an actor in it. You would never have connected those dots, you would never have thought because you had no idea they were in this vehicle. And yet, the system is saying, “Let’s try it,” you know, or I go on a run where I watch a whole bunch of us. Oh, my Lord, his name is escaping me right now. Duke, the Duke?
Jon-Mikel Bailey: Oh, John Wayne.
Duane Forrester: So I watch a bunch of John Wayne movies and then all of a sudden, you know, three months later, a bunch of Harrison Ford movies are being recommended. Because in the meantime, there have been articles written that have connected Harrison Ford as a mature actor to the way he kind of leans into a role to the way John Wayne leaned into his roles, you know, generationally separated, and yet there’s a similarity.
And so they put those things together. And that’s the whole point of machine learning. I mean, like at some point here, we have to talk about knowledge graphs, because we can’t be touching on this crap and not dive into Knowledge Graph like it’s, it’s extremely important. So whatever else you do, john, your task as a host is to get us knowledge graphs today.
Jon-Mikel Bailey: I think I can do that.
Duane Forrester: So back to the original question. Yes is the answer. Yes, the generation is training machine learning and machine learning is training the generation. And let’s be clear, generations before that were trained by Google’s search box. Give us a word and we’ll figure it out for you, and they got actually pretty damn good, if we gave them a word, at bringing back the right answers.
And then there were variations on that word, sucked a little bit, and then it got better and whatnot. And what we were seeing in there was, we were seeing machines learning, that’s what we were seeing. I don’t expect a two-year-old to tie the knot properly every time. It’s gonna take them a few months, right? To truly get proficient with it. The first few times, you know, mom or dad are like, “wow, that’s impressive.”
And they’re nailing it, you know, but when they’re on their own, at the end of the hallway, and you realize, holy crap, it’s too quiet my house, what’s going on, you walk around the corner to find a knotted mess, you know, like, that’s the algorithm learning, and that’s the reality we’ve lived through.
Jon-Mikel Bailey: Or you’re like, “hurry up and tie your shoes, we have to go.”
Duane Forrester: Yeah, and then you don’t recognize that that’s the seventh knot, they put in the string, right? And you’re like, wow, I just created a huge problem for myself. That happens with AI a lot, too.
Understanding the Knowledge Graph
Jon-Mikel Bailey: So Dwayne, I’m gonna give you a little gift here. I’m gonna dive right into Knowledge Graph. So, first, if you could, please explain to the layperson what a knowledge graph is, and maybe a little bit about why it matters in SEO?
Duane Forrester: Okay, well, it’s a graph that is filled with knowledge. Right? Okay, kidding aside, right, there’s a couple of things we have to call out here. First off, there is a product that Google calls the knowledge graph. So you will have heard it in relation to that, right, which is generally that data that they hold on the right-hand side pulled from Wikipedia or whatever source.
It’s literally a graph of knowledge based on whatever your query was, often applied to person/place/thing, because these are finite elements, with the exception of like, people may still be alive, or they may be dead and those kinds of things will fluctuate. But generally, you know, Jon is Jon, Duane is Duane. And if we are an entity of note, then we will have data about us, and if somebody is searching on us, then you will see that data.
And so the flip side to that is all modern search engines are built on an autograph. And a knowledge graph is comprised of a couple of key components. Okay. Mostly, it’s comprised of entities. So that would be, you know, Duane, Southern California, Camarillo, educated in Canada, born in Canada, American citizen today. All of those pieces of Duane the entity, all those pieces are the attributes for the entity, name, address, phone number, date of birth, shoe size, hobbies, you know, education, like all of that stuff applied to a person, things they’ve done that were noteworthy, all of it.
And so the idea of the knowledge graph is really the relationship between the entities and attributes, and all of the attributes and other entities themselves. So the way you need to think about this is it’s always three dimensional. Prior to having a knowledge graph to pull from and Google introduced it, I want to say, back in 2012, but prior to that search was largely two dimensional, it was just a very complex lookup table.
Give me this word here, a bunch of synonyms, I’m gonna guess that if I think that word is what you actually meant, if I understand it, and then I’m gonna go bring back a bunch of stuff. And I’m gonna give you all these blue links. And when you click on a link, you will reinforce whether I was close or far or what I was, and we can refine it, right?
And you can get really, really incredibly smart results by pursuing that path. But it doesn’t scale. Well. It’s a massive amount of uptake, upkeep, and work. And when you move to something like a knowledge graph, basically what you do is around the same time, Google went on record saying they wanted to map the earth. They wanted to literally send people out into the world and map the entire world and all the objects in it.
What they were describing as building a knowledge graph that powers search into the future. So they sent out automobiles with the spinny things on the roof, and they went out and took photos all around the neighborhoods and on the back roads. And they hired a company, actually bought a company, that built these things into a backpack.
Then they sent these people to Machu Picchu, into base camp at Everest, and all those places to hype those places with a 40-pound backpack on their back, carrying a spinning camera above their head to capture all of that. So you and I can go on Google Maps and walk through it right? But in order to answer questions about Machu Picchu, you have to go do that work. So Google invested in that, that built the knowledge graph.
If you look at Amazon, every product, and every service, and all of the attributes for all of those things, the size, the weight, the shipping, the shipping dimensions, the colors, the sizing of clothing, like all of these things are the attributes for them. They’re all built on a knowledge graph.
And if you look at Airbnb, Uber, they’re all built on a knowledge graph. They also share something else that’s really, really key to their success. They all start their experience with a giant search box, front, and center. Because if you tell me what you want, I will get you the answer through my machine learning.
So if I do my job, right, feeding all of the entities and attributes into the system, the machine learning is what goes through and takes millions of instances and says, “Here are all the possible relationships between them.”
And it can connect an Airbus A 380 airplane, to a doctor, and you and I might be saying, what the hell did those two things have to do with each other? Well, what they have to do with each other is it was that airplane that landed at Heathrow after that doctor did an emergency surgery on board that airplane, and saved a celebrity’s life, that checks off a lot of boxes for a lot of people around the world as being an important item of news.
And so suddenly, when you’re having a conversation about that airplane, incidentally, it’s retired now, it was a pretty impressive airplane, I’m disappointed. I never got to fly on one. But there you go. Now you see the connection between the airplane and the doctor. Whereas if it were just you and I walking down the street having a coffee we would never get from airplane to doctor like just, it would never happen.
And that’s not even a unique example, because, you know, doctors have been saving lives on airplanes since airplanes lifted off the ground. But that’s the capability of machine learning is to tease out all these different things. Okay, to answer a question like a “doctor near me, that takes Cigna insurance open now, taking new patients,” like that type of query is something that we see happening a lot more because of voice, which, again, is natural language processing, which is machine learning.
And so if you have a knowledge graph, and I’m going to loop us back very quickly here because I want this to be like, I want to take us from the abstract to the practical on this. So it’s all well and good to know that these large companies have this. And this is why they have this all right, your average business, what are you doing with a knowledge graph? Well, what you’re doing with a knowledge graph is you’re identifying every single entity and attribute that you have for your business:
Name, address, phone, number of products, services, dimensions, your C suite, their education, philanthropy support that the company does in a community, how long you’ve been around, absolutely anything and everything that is attributable, that you want to be public knowledge about your business, is going into your knowledge graph.
And your knowledge graph doesn’t need to start as anything more than a spreadsheet, because until you put it into a machine learning-powered environment, it just needs to be in a container. Put it in a spreadsheet, put it in a Word document, it doesn’t matter, you just need to identify all of those things, because those become digital assets for your business.
And then when you put it into machine learning, and then you say to the machines, “figure out what the relationships are.” That’s when you build your knowledge graph. That’s when you now have the ability to say “now let’s put a search box at the front of this and answer crazy questions.”
Like things such as, “can I use my fishing license from last year because the store I get them from is closed because of COVID-19?” That was an actual question that came up in the back end of the system that my company powers for the state of New Jersey last March when COVID shut them down. We powered that system.
We put it up and I saw that question come in in real-time. And I thought it was broken. Because I thought why is there a question about fishing in here? Right? Like, it doesn’t make any sense. To the person asking it made perfect sense. The fishing season was starting the following week. They knew that they couldn’t go get a fresh license. They wanted to go fishing. As long as they could launch their boat, they wanted to know if, in good faith, “I show you my license from last year, because my store is closed and I can’t get a current one, will I be safe legally.”
The state of New Jersey saw that come in and the next day they had answers written and put back into their system. And then suddenly it was “well look, we’re not really going to answer that question. What we are going to remind you is that all the state parks are closed and you shouldn’t be launching your boat.”
Yeah, like well let’s just stick to the facts of this but it immediately answered the question for that individual and then for every other individual, and it turned out in that upcoming five-day period, they had 1000s of questions asked from people about the fishing season last year. And the team there in New Jersey even said this, they said, “we never would have thought to answer those questions.”
And let’s take this back to SEO, you’re gonna do keyword research. Oh, good for you. How very 2010? Like, I know, I’m poking fun out of here, right? It’s still all your keyword research, it is still fundamental. It has to be a part of the equation. But it is not the anchor, it is only a part of the chain attaching the anchor to the ship. So do you skip it?
Well, you could you just end up with a, you know, chain, that’s one link shorter? I wouldn’t skip it, I’d still do it. There’s lots of data to be pulled out of that. But I would be looking at my search box and my log data to find out what questions people are asking me because that gives me a very clear understanding of intent, which is what the search engines are focused on.
Because if I can understand your intent, then I know where to put a product or service on your path. And you will most likely convert, that’s the power of all of this. Because keyword research is kind of like, at this point, it’s kind of like loading your own shotgun. And you’re like, “Okay, you know, I’ve got the gunpowder in there, and I’ve got the little pellets in there, and I’m gonna pull the trigger. And I’m gonna hit something.”
That’s kind of what keyword research is. Right now with the rollbacks in Google’s data, the combining of data and lumping under a single keyword or just here’s a number, we don’t tell you what’s under it, like that type of thing is taking us from, you know, the laser-sharp focus to a broader like, this is where we used to be back in the 90s, in the 2000s, when we didn’t have the data.
And being able to turn to your own systems. Imagine you’ve got a system that is powered by a knowledge graph that can answer all of these random questions the same way Google answers questions with the power of knowledge graph. So now you have somebody who comes to your website instead of bouncing over to Google, because you didn’t answer their question, they just stay with you.
And that whole experience, you’re showing them the funnel, buy now, contact us, sign up for our email, whatever it is, you get multiple kicks of that can instead of what five and a half seconds of attention span, and then they bounce. And let me be crystal clear about this, a person who comes into your website from Google. If you don’t answer their question almost immediately, they will bounce back to Google.
Now they have trained Google that you, even though they might have asked for you as a brand, you are not a good answer. So the next time that query comes, the system has been taught you’re not a good answer. So then it brings up a third party, maybe a competitor, certainly some random answer. But if the question is about you, or your product, you are the best source of truth for the answer.
To me, it’s a very tight circle and a very obvious thing. And so I’m a huge proponent of start by building your own knowledge graph, get your data together, it’s hugely important.
I mean, I know SEOs around the world are sitting there wondering what to do because they got so much time on their hands. They’re like, I need to fill in the blank here, you know, I need to look busy.
So Seriously, though, joking aside, this has got to be a priority, right? Because there is, if you will permit me, a wellspring of information and data to be culled, from your own search box, your own search experience, even when it sucks, and I know it does for most companies, it sucks.
But at the very least even if you fail them on the answer, you will still be able to see what they asked, the question? And that then you hand to the team who’s doing your content, the same team is taking in the keyword research to base content around and now they’re looking at it not saying “Well, here are random keywords, let’s try to infer intent.” What they’re seeing is here’s the entire question I was asked, what do I think the intent is?
Along with that keyword research data, it paints a very clear picture of… “Oh wait. So we’re seeing growth month over month on this keyword term. And we have five instances in the last two weeks of people asking a question that could all be bucketed together to be essentially the same answer. Okay, this tells me this data is accurate. The growth we’re seeing is real. We have no content. Let’s dive in on that.”
And now you’re not deploying based on keyword volume coming from paid ads triggering things, you’re actually basing this on “we have compelling evidence that this is exactly what our visitors want. And if I gave them that, they would stick around longer.”
That’s massive power right there. So, knowledge graphs, are the current kind of leading edge of Search and information retrieval. They’re the core that powers machine learning. Like, this is what you feed machines so that they can learn from, are knowledge graphs, right? So massively important for businesses.
The Sneaky Question about Schema
Jon-Mikel Bailey: So, you touched on this a little bit, and I want to get a little bit deeper. But I also kind of want to ask a quick follow up. So is schema, in a sense, a way for you to cheat that Google condones.
Duane Forrester: I don’t like the cut of your jib. I don’t condone cheating. Let’s dive into this because I know where your head’s coming from, right? And like, I get it, right, because I’m like, I run into this all the time I end up on client calls, or whatever, and people are like, “what are the two things I need to do this month to outrank all of my people?
Dude, it’s not a monster truck show. Calm down, right? Like, it’s slow and steady. And it takes a while. And it’s a lot of work. So here’s what here’s why schema is important. Schema is a language that allows you the owner of the content to identify that this is what I tell you. These are the official hours of operation. This is the official name of the business.
And schema when it’s implemented, it gives the search engines an easy way to say, “well, Jon certainly has done the work, he must own the content because he took the time to put this in there. And only the owner of the website has access to the code. Therefore, I am going to trust it at a higher level than I would if I just crawled it and found it.”
And part of that is because injection attacks on WordPress websites, where all kinds of crap get loaded up on your website, you don’t even know about it, you’re completely unaware. And Google’s looking at it going “well, hold on, Jon’s normally a pretty good guy. This doesn’t look like the type of website and content that he would normally have. I don’t know if I can trust this or not. But he’s still publishing on it every day. And yet, this is still there.”
That happens in the millions of iterations a day and Google sees it. So they can’t just blindly trust the website, because Jon shows up to work every day and Jon publishes new content to it because maybe Jon’s an idiot and doesn’t know he’s been hacked. I don’t know, not saying anything, just putting a placeholder there in case we want to come back and examine that later.
But the flip side to that is, look, if Jon’s taking the time to understand what’s applicable to him, he’s gone through the library, identified the elements that match his content, and then he’s gone into his content and put them in correctly, clearly, this is someone paying attention to details.
Think of it as giving off meta signals, okay? Not metadata. But like, when you give off good vibes, people feel that from you. When you show up, and you’re a Debbie Downer, they feel that from you, you know. And so the algorithms, the machine learning, the systems, the humans, who ultimately are the final arbiter, when they’re called into action, the system can’t make a decision on A, B or C, it tries again, it tries a third time.
At that point, it goes into failure mode, which is pump this out to a human, have the human look at it Tell me is it A, B or C, because I can’t determine. Here are my projections, right, I’m leaning toward B, but I need a human to sign off on it. And that’s training the machine another way.
When those instances happen, they sometimes are just avoided completely because you have schema in place. And the system is looking at it going “we have schema we’ll trust and we’ll go from there.” There are plenty of other checks and balances that if you screwed up or you’ve been hacked, or there’s something else that they won’t put you into a search result, but it’s not an automatic pass through. It’s not a get out of jail free card. It’s not an automatic ranking signal.
But I will say this, wouldn’t it suck to wake up and find out that you’re the only one of your competitors, your peer group, that doesn’t have this on your website? And you’re wondering why everyone else is outranking you. Then it’s not so much a ranking signal as the new table stakes. And if we’re not doing it, then you’re clearly not as invested.
So if you’re not as invested there, now, I’m also looking at “are you doing review management? Are you responding to people? What do your reviews look like? What’s your response rate? What’s your frequency? Oh, I see all of that is lacking. And you didn’t do this. I don’t know if you’re the answer I want to put in front of people.”
And so then Google just goes and takes a different one and says, “look, these people respond to reviews in less than 24 hours. Look, they’ve responded to the latest reviews, look, they’ve got schema on their website. Look, they don’t appear to be hacked.” It’s instantaneous for the algorithm and makes the decision.
One becomes four, three becomes one, and your competitors suddenly has gapped you, and you’re like, “what the hell happened? What is going on?” And you’re freaking out. And it’s like, that’s because you can do basic work. And that’s what it is today.
So, shortcut cheat now, I mean, I wish, right? Like, I wish we had silver bullets, you know. I think really and truly at this point in our history, we way run out of silver bullets as an industry, like they’re just, you just can’t get there.
You know, when I talk to folks at the search engine, I ask them, “what if I called you up one day and said, ‘Look, I need you to find this object in the database.'” And they’re like, “yeah, that’s not gonna happen.” I’m like, “why not?”
And they’re like, “picture, you open your car’s hood and it’s just like a black box. And you open that, and it’s just a black hole, and you manage to survive that. And on the other side, you’re in another black box. And when you open it, you’re under another hood. And then when you go through that, you’re back out into whatever that reality is, but you still haven’t seen the object yet. Like, that’s what it’s like.
And, and I won’t get into exactly, we can do this another time, how a search engine pursues a page of content, what it does with it, but it doesn’t make it easy to find stuff, like even for itself, right? It’s about deconstruction, it’s about efficient storage. And to be completely clear, how you store things is very different depending on what the object is.
How you store a video, how you store an image, and how you store text, how you store a quote, review caption, all of these things. Efficiently storing and retrieving those things requires different types of coding, different types of hardware, different types of processes. And so it’s not like a web page is in Google’s index.
The notion of webpages in Google’s index, that it can bring back the URL that points you to the web page that’s on the internet, if you went in from Google’s back door, you wouldn’t see a list of pages and be able to recognize that it’s a page, right?
You’d open the back door and go, “what are these? Oh, these are individual sentences from random locations on random web pages around the internet?” Yes, you found the passage storage space. And none of it would make any sense to you, because they’re just like snippets of statements. And you’d be like, “well, I’m gonna recognize the words, but I don’t have any context. I don’t know where they’re from.”
And the identifiers that are uniquely assigned to them, there are all of them, they are unique, and they are all secure. So you can’t even ask it. Like, “where do you come from?” “I don’t know” Because I’m just sitting here sucking on a lollipop waiting for somebody to call me into existence? You know, and like, that’s it.
Jon-Mikel Bailey: I have zero desire to ever peek inside that black box because I think I think my brain will explode.
Duane Forrester: See, so here’s my thing. I think, rather than wanting to peek inside of the black box, SEOs need to find ways to hack consumer behavior. And when you do that, now you start to capture more of your market share, the market share, you should have naturally.
And you’ve got to pay attention to stuff like if you do a brand query right now on Google, you’ll see down below the results or sometimes closer to the top, you’ll see the “people also asked” section, okay, right, not the autocomplete that comes up. But there’s separate sections people also asked, and almost instantly in there, you see that they stripped the brand out.
And so this is them trying to learn. “Did you actually mean the brand? Or did you mean any brand, but their version of that product?” Right? Like “did you actually mean Lulu Lemon workout gear, or are you OK with Nike workout gear?” And like you’re using it like Kleenex, right? It’s like an interchangeable moment with brand names.
It’s trying to understand that it’s trying to disambiguate at another more refined level. The downside to that is if you are Lulu Lemon, then okay, first off, when I land on that page, the first thing I see are a ton of ads from my competitors trying to drag me away. And so I’m Lulu Lemon, now I need a bigger ad budget to ensure I’m sitting at the top of that so that I can get traffic that was supposed to be mine anyhow.
Then I come down the page, usually below the fold at this point, I find an organic listing, I really, really have to have a good experience because when that person clicks through, if they don’t see exactly what they’re looking for, then they’re going to ask me the question.
And if I fail to give them the answer, they’re going to bounce back because now in their mind, they’re thinking, “maybe I didn’t mean Lulu Lemon. Maybe I’ll go take a look at this other one. And that’s okay because I couldn’t find what I was looking for there.” And it’s just a massively lost opportunity.
And the system has been trained, right? “Well, you know, we have evidence that says they didn’t really mean Lulu Lemon when they asked for Lulu Lemon,” when in fact the intent behind it was that it’s just that they couldn’t show it to them. They didn’t have a good UX on the other side.
But if you can, if you understand that these things are happening, you can build a user experience that answers those questions. I’m gonna take us back to Amazon and Uber. And you know, Airbnb and Google, there’s a reason they’ve got the giant search box there. Because they’re literally saying, “just tell me what you want.” Just tell me what you want. And if you are clear about what you want, I will give you what you want.”
That’s their be-all and end-all. And we need to take actual lessons when we build our websites from exactly that.
Jon-Mikel Bailey: Craziness there. So I wanted to I want to talk a little bit about…
Duane Forrester: Drop the mic, but I own the mic.
Jon-Mikel Bailey: Yeah, there you go. You don’t want to, you know, damage the mic.
The Often Untapped Value of Vertical Search Engines
Jon-Mikel Bailey: So I wanted to shift gears a little bit and talk about an article you recently shared about vertical search engines. So we’re talking about like Zillow for houses or AutoTrader, for cars, you know, things like that. So can you speak to their popularity, and then talk about the future of these search tools? And, you know, their evolving relationship with Google? Because it seems to be kind of this…
Duane Forrester: Oh, yeah, it is.
Jon-Mikel Bailey: You know what I mean?
Duane Forrester: Like, it’s a very frenemy type situation. I want to rank well, and you, but really, I want people engaging with me, you know, and that is, you know, the opposite of what Google wants. Right? Google wants you to engage other pages so that you’re clicking on the ads, and then they’re making the money.
Both sides are right. Both sides are not wrong? Okay, so let’s look at it this way, a lot of us will recognize the concept of vertical search, because we all kind of grew up through that in the original days of search, right? Everything was verticalized. You know, if you wanted weather you clicked on weather.com, if you wanted autos, you clicked on autos.com.
It was just a pre-sorted pre-filter that gave you access to an index of predefined things that fit within that. That was very simple, very straightforward. It’s a bit different today, right? So when somebody thinks of shopping for a home, a couple of big names come to mind, right?
So I’m from Canada. So MLS kind of has a lock in Canada, but there’s also an MLS version here in the US. We got Zillow, we got Redfin, you know, you got realtor.com. Like, there’s, there’s a fairly, it’s a fairly pointed group at the top right. It’s not singular, but it’s still a pyramid shape with a very small plateau. And there’s some major people on there.
And if you ask the average person, you know, tell me the name of realty company, right? These names that I just mentioned would come up. And of course, they’d also say things like, you know, REMAX and you know, all of these other company names, but in terms of “Okay, well, who do you think gives you the largest selection of homes for sale?”
Okay, you know what any brand name that’s associated with, like actual selling homes is out the window. Because everybody knows that REMAX only has REMAX homes, they don’t show anyone else’s homes. And so that’s where the Zillows, and the Redfins come in, and all these other services.
So we’ve kind of been trained for probably the last two generations of searchers to expect better. Google started it, you know, give us a word, we’ll get you really close to the target. And, now these services have popped up. And now when people are just even browsing, you know, like, they go to Zillow, and they start just, you know, here’s the zip code I’m curious about. What are the home’s cost and what are the schools like in the neighborhood? And all this data is filed on there.
And then that allows them to then say, “oh, okay, well, you know, maybe Redfin has different ones.” And they go, and they see, like, you know, 80% of the homes are the same between the two of them. But there’s a few differences in there, and it keeps them engaged. But they’re not engaged on Google because they feel that the real estate search is probably a better place to start.
Because what you’re not getting in Google, or what you do get in Google are things that are related to real estate. And when you start going deeper, it’s not just more homes in the area, it’s actually more results that are irrelevant. So you know, and even in that list, the first page, let’s be honest, right, you do a search for real estate in a zip code on that first page, you’re gonna see everybody that we mentioned, listed.
And Google telling you, “you should go to Redfin, you should go to Zillow, you should go to all these places,” and people naturally go there. And so we’re starting to see when you look at millennials, and you look at Gen Z, and certainly going to happen with Alpha, there’s less of an appetite for looking around and finding stuff and more of an expectation of “just answer my question.”
We all kind of feel that right. And so I feel what’s happening here is, we’re seeing the rise of this because the proximity to the topic means that you can focus resources on getting better answers for me. And so if I come to you and you are known to be the repository of all things coffee, why would I go to Google and search for coffee? When you’ve proven to me to constantly exceed my expectations?
Now? is Google better empirically? Yes, Google, we better in every measure than that vertical engine. However, if that vertical engine continues to impress me, they’ve met my bar as a consumer. So I’m not as a consumer constantly going between the two services saying “is Google better today? Is this better? Did I miss something?”
No, I’m satisfied with the houses Redfin and Zillow show me, I’m satisfied with the coffee that the coffee vertical search shows me, and therefore I’m a happy searcher. So there’s a lot of this. And if you extend this out, where I think this logically ends up at is, look, if you’ve got a search box on your website, you need help. You need to fix that crap, because I’m telling you right now, and almost 90% of the cases that I see, they’re just busted, they’re broken. I’m not kidding.
Oh, I went on a website, brand name, national household name last week, on a call with them and I’m talking and typing in keywords that are from their website, product names. And their search is coming up with zero hits. I’m like, okay, just tell me your search box is not connected to a database, because I’m having a really hard time understanding that the product name of the things you sold, were not included in the database of answers for this.
So then I thought, okay, the ultimate test. Below the search box, they had their five newest items that they brought to market, current television campaigns around these items. With a quick search link, click, zero results.
I’m like, all right, seriously, I’m almost willing to put on a double mask, get on an airplane, go to your office and have a stern talking with someone. You had to go edit the page to put the link in? How did you like, it’s, the test is a click that click, fail, oh, that didn’t work. Maybe we shouldn’t include that. Or maybe we should put it in the database.
And that’s the world of Site Search today. And it needs to change because there’s a massive amount of opportunity hidden in an average website’s Site Search. And if they don’t take advantage of it, well, then people come to them. They don’t see what they want. They bounce back to Google, you just train Google that you’re not the right answer, even when it’s your name.
And they’ll give it to a third party, maybe a competitor who will then throw some Fudd up on the page to convince the consumer “Oh, you didn’t want theirs, take mine.” The consumer doesn’t care. Okay. The consumer does not care whether it’s Lucky Brand jeans, or Levi’s they’re got their jeans, they’re happy with the look and the fit. Okay. Are they tied to Levi’s? Yeah, some people are.
I am so I will wait, I will find a way to buy the Levi’s product. But that doesn’t happen all the time. You know, you mentioned at the top of the hour, how amazing the guitars I build are. I may have put some words in your mouth there, Jon, I apologize if I overstepped, but I use some CA glue, super glue for a lot of the different things that we have to do and building them. And I don’t give a crap about what the brand is.
I care about the viscosity, I care about the shelf life, I don’t care about the brand. It’s all the same product, like wouldn’t surprise me if there’s a single factory on planet Earth, making all of it, just putting it in different bottles. It would not surprise me, don’t care, doesn’t matter, it does the job. And I just know I need it.
So, that’s the general consensus with the consumer. But when I sit down and I say no-no-no, it’s got to be Krazy Glue brand, or it’s got to be Gorilla brand. Hey, you know, that becomes a problem when everybody’s sold out of it. And an opportunity, you know, because now when I go to the Gorilla website, and I’m asking for those questions, and I want that product, you have clear intent that someone who’s a fan wants a product. Maybe you should create a program that actually protects the loyalty of those fans.
So our next production run, if you’re a member of our fan program, that’s dedicated to fulfilling your order if you place it directly with us. Well listen, man, that goes against everything I’ve taught myself over a decade-plus of being a Prime member like really it does. But I’m going to do it because nobody else has it. And if you’re willing to take care of me, I’m willing to attach to you.
And so then at that point, you know, for the rest of my life, my super glue needs are taken care of. Anytime I think of superglue, I think of the program I’m involved in, I think about a problem solved. I know what to do. Here we go, and you know, failing that, I’ll put on a hazmat suit, go to a grocery store and buy some crazy glue.
There’s always a fallback to it, but consumers are looking for that kind of path of least resistance, and Google is happy to give it to them. And if you’re not answering the question, they’re happy to take something else that looks like an answer. And for the consumer, look, your intent is to solve the problem not to be the brand loyalist. You would like to be the brand loyalist, but if that’s failing you, you’re going to jettison, you’re just going to go solve your problem. And that’s it.
And so there’s a massive amount of opportunity there for businesses to just kind of hack that. And along the way, you get all this data, you get all this information from the systems that are telling you exactly what people are looking for.
So forget keyword research that tells you 100 people search on this in the last 30 days because it triggered an ad in Google’s system. Well, that’s awesome. But where are they in the funnel? Are they at the research end? Or are they at the buying end? Your content will have to be worded differently depending on where they’re at, and you don’t really know that.
The data you get is coming from a paid system. So you know, but if it’s coming from your own search box, you know exactly what they asked. You know when they asked, you know how long they engaged with the website, you have the full query string, you have all of it. So now you can start pre-empting that and you can do stuff like preload a bunch of FAQs that you never thought of. And that could dramatically reduce costs that your call center because you’ve been able to directly answer the question.
And hey, what a great moment to ask them to leave you a review of the experience they had and drive your reviews up. Because “I came, I asked, I got what I wanted. I love you.” This is how you hack your way forward today. As an SEO, in fact, SEO, let’s start thinking of it more broad terms, replace “engine” with “experience,” “search experience optimization,” because that’s really where we’re at today.
And “engine” is a part of it. But there’s a broader context I think. So there now that I’ve bastardized SEO and said, you know, it’s not dead forever. Here we go.
Jon-Mikel Bailey: Fair enough. And I couldn’t agree with you more.
Does Bing Matter?
Jon-Mikel Bailey: But we’ve been talking about Google this whole time. And I have to address the elephant in the room. So you helped run Bing’s Webmaster Tools program for nine years. So I have to ask the question, two questions, really, one, is Bing still relevant? And two, do you think that there are opportunities with Bing that businesses might be overlooking?
Duane Forrester: So to answer question number one, I have a phalanx of flaming arrows converging on your office right now.
So yeah, so listen, Bing is still very relevant. Okay. And there’s a very important reason why right? So again, let’s take our thinking beyond a traditional SEO mindset. All right, what you want to be looking at is you want to be looking at what does Bing represent to Microsoft and Microsoft’s plans?
Okay, you start a search engine, you’ll notice that when the earnings reports come out, right, nobody’s bragging about how much money Bing is making, right? Bing is making money. It’s covering its costs, the shareholders aren’t complaining about it. And the reason that Bing exists is data, it’s all about data. And what does that data do? Well, that data feeds Azure, and the machine learning systems, that cloud component that makes Microsoft one of the most valuable companies on planet Earth right now?
That’s what your search engine is capable of doing? So is it relevant? Oh, hell, yes it’s relevant. It is massively relevant. Because at this moment in time, Microsoft Azure is one of the premium, probably, it’s either one or two in the list of the top cloud services. So when you roll in there, and you want those services, it’s premium, right? And it’s because of that data.
And oh, by the way, there is still direct traffic that comes in, and there are still millions of users a month that you can tap into. Oh, and by the way, if you run ad campaigns on there, the cost per click is demonstrably cheaper than it is on Google. And still, to this day, consumers convert at a higher rate from Microsoft properties than they do from Google.
So there’s a lot of compelling nuts and bolts reasons to be considering this. Okay. And yes, you can argue very clearly, market share, volume. I’m totally with you, I get it. That’s fine. Here’s the upside. The upside is Bing Webmaster Tools still provides more and deeper data than any other toolset that’s available from an engine.
So you want to have that because if you take what you consider to be a less than ideal market share, and apply that data, and extrapolate it, it gives you some indication of those missing things from Google. At least it gives you a trend and direction to look in if not an accurate number. So there’s a lot of value to pull from that area.
Plus, you know, let’s not lose sight of the fact, the thing is global, they have popularity in some markets that rivals or exceeds Google. So depending on what you’re doing, there may be, I think of it as kind of like a squishy ball of reasons why, right? The good news, of course, is that there’s no extra work that you have to do, like the work you’re doing for Google is applicable to Bing, and it’s all well and good.
So, you know, I think there’s still a compelling reason to be focused on it, Jon. And at this point, I’m going to claim COVID brain and I’m gonna ask you to repeat question number two for me so that I can be accurate on that.
Jon-Mikel Bailey: Fair enough. So do you think there are opportunities that businesses might be missing by overlooking Bing?
Duane Forrester: Yeah, so here’s what I think about this, right? I think that in a lot of instances, we tend to want to dumb things down to make it simpler. So we can consume information faster, get to a decision quicker, and move past whatever it happens to be a moment in time.
Unfortunately, that approach trains us, and sometimes with negative consequences. So when we apply it to this, a common refrain is “Google is most of my traffic, Bing is inconsequential. I’m not going to focus on Bing, therefore, I ignore Bing.” Okay, that’s totally fine.
But if I told you that inside Bing Webmaster Tools is an actual SEO tool that will scan your web page and tell you the things you need to fix? Does that sound like something that’s valuable to you? “Oh, hell, yes. Absolutely.” Oh, so now you want to pay attention to Bing? Well, yes, absolutely. You know, that’s been there since I launched that tool half a decade ago.
So, yeah, there are good reasons to pay attention here. Because if you’re an SEO, and you’re gonna argue with me that you’re all about the data, then I’m going to argue with you that you’re not all about the data unless you’re taking in data from all of the toolsets that are the official toolsets.
So not just Bing. But what are you looking at from DuckDuckGo? What are you looking at from Yandex and Baidu, and these other areas, because, hey, look, man, there are a lot of topics that you’re going to be into as a company that you may think are a national thing, but are actually a global thing.
And trends in other areas may be indicative of things you’re seeing here. So, it’s not that moment to narrow it down, to scope this down to make it so easy. That’s the only thing I have to focus on. And, unfortunately, your company will follow your lead as an SEO, when you’re the voice of knowledge on a topic, and you tell everyone, nothing matters more than Google, they’ll believe you, and they will support you. And that’s dangerous.
Because when you no longer have access to reliable data from Google, or I don’t know, your GMB account gets hacked, or some features were taken away that you relied on for your reporting. Even innocuous things like they do an update and you’re affected, not badly, but you lose 30% of your traffic overall. And that has a direct effect on revenue.
It’s a really dangerous position to be in to be the person saying Google should be our only focus. That’s a really, really dangerous position. So I’m not saying you’re going to offset that. Or you’re going to find an extra 30% revenue over in Bing.
But I am saying, be aware of this, put time in your week, to go to the tools, to look at the data, activate the account, claim your listings locally, all of those things, do that work, this is your foundation. And if you’re not managing the foundation, how do you expect Google, your number one, to look at you and say I should really take you seriously?
Because clearly, you’re indicating that you’re not doing everything you can for the consumer, you’re focusing only on Google. And as much as Google likes that they don’t like that, that you have to focus on that customer. So to me, it’s as broad a footprint as possible. That’s what it is.
Luckily, in the world of search, again, it’s not that broad, you know, like, it’s not like DuckDuckGo has so many queries, that the data you’re going to be seeing you can’t wrap your head around it. Of course, you can. And their toolset isn’t that deep, that you’re going to be like, oh, wow, I’m going to spend a month in here and still not understand at all? No, of course, you’re gonna be able to understand it.
But you do have to be very careful because SEOs will lose mountains of time looking across a sea of different tools, most of them third party, and wondering “how do I translate that data into this? And how do I make these things match up? And is this data reliable?”
“Hey, has everybody who has this tool suffered an outage of no missing data? Why is it gone?” And suddenly, you’ve lost half a week just trying to get data from a tool? That’s a third party that I don’t know, maybe Google just said, “screw you. We’re blocking your crawler. We’re not going to scrap this anymore.” And that like, what do you do then? Well, I’m asking you, Jon, what do you do then?
Jon-Mikel Bailey: I don’t know? Right? I’m frightened.
Duane Forrester: It is a scary world, right? You know, if it’ll make you feel better, go watch the Social Dilemma again. And you’ll just like, that’s like a ray of sunshine.
Jon-Mikel Bailey: That is sage advice.
Duane Forrester: Yeah, just so everybody listening. So you know, I was born in Canada and that was a healthy dose of Canadian sarcasm on ray of sunshine attached to that movie.
Jon-Mikel Bailey: No, you ruined it. I wanted to go watch it and have their day completely ruined.
Duane Forrester: The hate mail on Twitter right? Duane, you killed my Saturday night funday with my kids, they hate me now and I’ve been on Twitter nonstop since then. TikTok is taking over my life. In fairness, I’ve been on TikTok for years and it has taken over my life.
Jon-Mikel Bailey: Well, I could literally talk to you all day.
Duane Forrester: I did mention four hours this morning in our email.
Jon-Mikel Bailey: You did, you did, and, you know, if I can raise enough money to to be able to afford an episode two, I think I might have you back for sure. So I’ll start a GoFundMe, bring Duane back.
Duane Forrester: Well, look, I’ll tell you what, when you as the money comes in, let me know how the fundraising is going.
Jon-Mikel Bailey: And I’ll have one of those little thermometers.
Duane Forrester: Yeah, I’ll see if I can make sure that my invoicing matches that number.
Jon-Mikel Bailey: Perfect. Perfect.
Duane Forrester: Let’s you know, like there’s I think maybe there should be an appearance fee next time.
Jon-Mikel Bailey: Oh, yes, fair.
Duane Forrester: Kidding. I’m happy to come back and be with you guys, Jon. Happy to you, guys are awesome.
Jon-Mikel Bailey: Thank you so much. It really has been a joy and and I’ve learned a lot and I know people watching this will learn a ton as well. So with that, thank you very much and stay safe out there and I will see you on the other side.
Duane Forrester: Absolutely, man. Happy Holidays, be well. Hi, mom. She’ll never see this.
Jon-Mikel Bailey: No.
