BC: I don’t know how many of you have attended these sessions. We’ve done several of them, including as a keynote at the Affiliate Summit. Ask Us Anything is actually ask us anything but there are slight rules. don’t ask us to solve the major website problem you have if it’s specific to your website. We ask that you come up and talk to us later. If it is generally something that the audience would care about, that’s what the question-and-answer period is all about.

There are three of us on the panel. I’m Bruce Clay, and I run bruceclay.com. They’re Stephan Spencer and Duane Forrester. We’re going to have everybody do more of an introduction of themselves.

We’ve got something like nine books over 50 years in the industry. We think we can handle almost any question, and we’d be happy to have you try to stump us.

DF: We’ve done this in the past. It works beautifully.

BC: It works great. Everybody circles 10s.

We all win.

DF: And we deliver the goods.

BC: By the way, we are all very good friends, and you will find us going to kid each other. We’re going to have fun up here. Duane, introduce yourself.

DF: Hi guys, Duane Forrester. I used to work at Caesar’s Palace, moved into sports betting, wrote a book, was hired by Microsoft, ran SEO there, launched Webmaster Tools in 2012 for Bing, left there, was working with Bruce for about a year, and will be starting with a new company at the end of this month. That’s me.

And it’s a secret. You don’t even tell me where you’re going.

DF: Yeah, we haven’t announced it yet.

His wife knows.

DF: And the company knows.

Okay, I’m Stephan Spencer. I have a few books under my belt. They are O’Reilly books. This easy one is Google Power Search— how to find anything in Google, like confidential business plans and Forrester Research reports that cost thousands of dollars. I’ve even found credit card number files containing the actual expiration dates. I didn’t teach you how to do that in the book, by the way. 

This is Social eCommerce, which is how to drive sales through social media. This beast is the bible of SEO (The Art of SEO). It’s a thousand pages. The 3rd edition came out in September 2015. A few things are missing, like AMP, RankBrain, and a few tools and things, but it’s still very current. 

DF: You have to carry this home.

BC: That book is very good for reading just before you want to go to sleep. 

My friend, Amy Africa, says it’s better than Ambien. I’ll take it as a compliment. I don’t know if it’s one. And I have two podcasts you must subscribe to, The Optimized Geek (now Get Yourself Optimized) and Marketing Speak. The Optimized Geek is not an SEO podcast, even though it sounds like one. It’s a biohacking and life-hacking podcast. 

Who’s heard of The Bulletproof Diet or Bulletproof Coffee? Dave Asprey, anyone? Yeah, he’s awesome. I’ve had him on my show. I’ve had Dr. Daniel Amen, who wrote Change Your Brain, Change Your Life. I’ve had Byron Katie, a huge self-help guru. Some really amazing guests.

I’ve had Tim Ferris and David Allen, who wrote Getting Things Done. He’s a great guest.

If you want to jot that down, The Optimized Geek or you can just come and get a podcast card to remind you after the session. The other one is Marketing Speak, which is also amazing. I’ve had some incredible guests like Jay Abraham, Dan Kennedy (the godfather of copywriting), and many great folks who have spoken here. In fact, the co-organizer of this conference, Shawn Collins, has been one of my guests. 

BC: Okay, I’m Bruce Clay. I have a website which is bruceclay.com, which makes it easier for you to remember. I started doing SEO in January 1996 as a pure-play SEO company. I’ve been given credit—it depends on who you talk to—for inventing the term Search Engine Optimization. I wrote a book not quite as big as his. It was bigger when I did the first edition, but he had to pass me. It is Search Engine Optimization All-in-One For Dummies by Wiley. It’s a tabletop book. Hopefully, it helps you.

Tabletop book? Like a coffee table book?

BC: It’s a little bit bigger than that. But it’s a reference book, so that helps. I have offices on some five continents. We do SEO and PPC stuff, so any questions that bridge over, we should be able to answer up here.

DF: The rest of you, get ready because this one might suck. We don’t know.

Audience: I have two questions; they’re related. I’m with Penguin 4.0. We obviously saw the granular devaluation of links, so the change in the Penguin algorithm. There seems to be a school of thought, and we support that Google has introduced as part of the Penguin 4.0 algorithm more of an important filter on topical relevance of the location of links. I’d like to hear your opinion on that.

This time SEOs seem to think—I’m talking about published SEOs—it’s doing it at an individual link compatibility level. Some think that it’s going deeper. It evaluates the topical relevance of where the link is located, the content, the website’s theme, et cetera.

The second part of my question is, do you believe there is any link between that, if you accept that hypothesis and RankBrain?

BC: Okay, I’ll start. On your first question, you’re absolutely right. Google has added to the algorithm the capability to determine that your inbound links are inorganic. In other words, the site that links to you is not about the same thing you’re about. Its theme is based on the sending side.

That was to fight a lot of spam sites that we just published blogs of any kind and linked to you. While it may be contextually relevant, each blog across the site was all little islands. The way to fight that is to make the site that links to your theme similar to yours.

Also, your site has to be more themed. It’s more of a site-wide analysis than a page analysis. That’s the first part. RankBrain is more of a content-centric thing trying to figure out what you’re about.

The official line from Google is that it’s all about analyzing the query. But machine learning is going to take over all aspects of Google and their determination of what search results should get to the top and which one should get buried. It’s not just about the topical relevance of the link. It’s about the geolocation and other aspects, like location on the page. 

If you are trying to rank in Las Vegas, you want to get local links from Las Vegas—convention, visitor’s bureau, tourism bureau, and all that stuff that is local to your city. 

If you think about topical relevance in terms of entities and neighborhoods, there’s a really amazing tool coming out from Searchmetrics. It’s Topic Explorer. It’s handy for you to do your keyword research or more topic research. You can even use their content evaluation tool, paste in an existing article that you wrote or write an article inside their content editor. It will tell you if you hit those topics adequately or not. 

It’s an amazing tool. You can add topics to your topic and so forth. That’s the Topic Explorer from Searchmetrics. It’s part of their content suite. 

DF: I’m going to second that. I saw a preview of the Searchmetrics tool; It’s fabulous and does a very good job of this.

A couple of things are rolled into this—the concept of inbound link context, which is essentially what Penguin’s dealing with. Google has been saying for years that they were capable of doing this, and finally, with this update, they are now capable of doing this, having that granular definition of what local for a link means. It may mean local, within a community, city, region, or ZIP code. It could mean within a neighborhood online—a section of a vertical, this kind of idea.

That’s what they’ve progressed to. They progressed to what they claimed to have been able to do and wanted to do. They’re now factually there and able to do that, which is why we see differences. It makes it very hard to chase links then because that broad spray-and-pray approach doesn’t work anymore. You have to be good at this.

The other thing about this is RankBrain. By its very default, this stuff is a part of RankBrain. But let’s not get too confused by this RankBrain. It’s just a new fancy term for redefining the black box that’s always existed at Google.

Before that, we have the black box, and you won’t tell us the algorithm. You won’t tell us the pieces we should be looking at. So then we figured out some stuff on our own—things like content and quality—and then Google came out and said, oh, the three most important things are content, quality, and RankBrain. Okay, what’s in RankBrain? Oh, sorry, I can’t tell you that.

They don’t even know what’s in it. 

DF: That’s part of the problem now. It has gotten smart enough that if you kidnap all the engineers, they wouldn’t be able to answer the question for you. At best, you can recreate the earliest version of it with those human beings. So, yeah, it’s in there, and ultimately I know Google and Bing, having been there, want to get to that point where it’s all algorithmic, where it’s all trust machines, and they generally do a good job. Does anyone here watch the NFL?

What’s that?

FD: Sorry, National Football League, Stephan, that is an almost national pastime that people enjoy. Bing’s prediction feature pissed me off last weekend because my Seahawks lost. We’re using these technologies to mine the data and come up with information relevant to a step in the searcher’s journey.

As Bruce alluded to there, that’s what RankBrain’s job is. It’s less about, “Hey, let’s manage all these signals and give you something. Let’s manage you all and then create a series of individual signals and give you what is much closer to each of you.” I wish they’d apply it to what they’re recommending to me on YouTube because it’s way off the map.

BC: Okay, another question.

Audience: I’ve been blogging for seven years, and 99% of my traffic comes from social. We’re trying to get into SEO. SEO is our goal for 2017. There are so many agencies out there to help us. There are so many people promising this and promising that. What would you recommend? The top three things we need to look for in SEO are.

That’s a great question. The first thing is honesty. I have this thing called the honesty test. I can’t claim ownership of it. A friend of mine came up with it. It’s a really brilliant idea. 

In the interview process, you ask them a pretty innocuous question and say, “If we do hire you, what would you think the most important quality or attribute would be to do this job well?” Is it attention to detail, honesty, creativity, technical acumen, dedication, or perseverance? What would it be? 

You just let them answer the question, and the only right answer is honestly. Because if they’re not honest, if that’s not their number one priority, or that’s just not baked into their character, you’ll get screwed. There are so many charlatans out there that it’s tough. 

Another thing you can do to weed out the charlatans and posers is to go to my website (stephanspencer.com) and download a free white paper called the SEO BS Detector. A handful of questions seem innocuous, and you work them into the interview process. 

They won’t know it’s a trick question because you’re not an SEO expert. So when you ask, “What’s the process for optimizing meta keywords?” This is a trick question because the only right answer is meta keywords are never counted in Google. They’ve always been a waste of time. If you care about Google, they’ve never counted. 

Now, if they say, “Well, they don’t count as much as they used to, there’s the door because that’s not the right answer.” They need to know that they never counted in Google. That’s on Google Webmaster Central’s blog record back in 2009.

Audience: And you said that’s found on your site?

That’s on my site. It’s called the SEO BS Detector on stephanspencer.com, under the resources section. That is really helpful.

Also, there’s an SEO Hiring Blueprint that will be quite helpful. That’s another white paper in the same area on my site. It has a seven-step process for screening, hiring, and onboarding. Whether it’s a contractor, an agency, or an employee, it’s applicable. That’s my recommendation.

BC: I’m going to take a different stance. I agree with everything you just said. The problem is that we’re commonly their third or fourth SEO agency when people come to us. That happens a lot.

What happens is you hear promises. If anybody talks to you and says the word promise, “I promise you ranking,” run. If they say, “The way we’re going to give the ranking is we’re going to buy links for you,” that’s a known spam tactic, run. If they come to you and say, “We’re cheap because we’re offshore,” run.

All of those things diminish their ability of them to communicate, to be able to support you and understand your community. You want people that can do it. The right SEO company sells value; they don’t sell rankings. Rankings aren’t a good measure of success for SEOs in general. Everybody should care that your business is prospering. That’s the partnership of a true SEO company.

Audience: And then I have one other question. This may be in Searchmetrics. I’m not sure about that technology. What is the best site that you recommend for finding long-tail keywords?

DF: Do you have access to your server logs?

Audience: No.

DF: You need access to your server logs.

Audience: Honestly, I’m up here, and I feel like an idiot because I don’t know SEO, but talk to me about Facebook. Like we all have our strengths and weaknesses, and SEO, I’ll just come out and say, “I don’t know.”

DF: I’m actually going to roll back your first question. No one will take a more vested interest in your business and your success than you. A part of this is learning some basic things. These guys have great opportunities and great resources.

I started my journey in all of this because I didn’t want to have to pay anyone. Yet, here I am 20 years later, and I’m the one being paid. For the record, I don’t offer these services, nor will I be in my next career step at the end of the month, but I spent a lot of time.

I founded the in-house committee with Sambo. So I am very well-connected to everyone who runs in-house search groups for all major brands and all the large companies worldwide. I know what it takes to build teams of SEOs. I’ve made them and trained myself.

A couple of things stand out to me. One, you have to learn some stuff. There’s no question. You never hand your future success over to someone else who talks their way through your request for the proposal process.

Honesty. Having worked beside Bruce and having known him for a long time, honesty is the number one answer. I’ve seen it in action there. I left Microsoft because I chose to join Bruce Clay because I know there’s an ethical standard. I also say the same thing about Stephan because he is a friend.

But there does come the point here where the price starts to play a role in this. This is a skilled area, and the people who know what they’re doing command the price. I’m not saying you can’t find good value at lower price points, lower being relative to everyone’s budgets. But do not move forward with something because you feel desperate, and it fits your budget. That is a dangerous combination of hope, and hope is not a business model.

Audience: Agreed.

DF: If you want to reach out to me on Twitter, I can answer your questions. I can look at people and see whether it’s real. I know a lot of people as well. I certainly don’t mind supporting folks in that way as well.

Yeah, likewise for me, too.

DF: It’s one of those things. We’re here. We’re a resource. We’re making ourselves available. Use that.

One of the things that I used to do when I was looking to hire a company to do this in my startup days created an interview just like I would for an employee. When I interviewed the company, I went through a list of things. First, they had to give me past businesses that they felt confident would give a recommendation. Then I have to do the legwork to talk to at least three companies.

I’d have various questions for them. It doesn’t matter because there’s only one question you want to ask that company. Knowing what you know today of that service provider, would you hire them again to do the same work? If the answer is anything other than instant, yes, there’s a problem. Something is not working.

Nine times out of 10, it was okay, but it wasn’t that great. It wasn’t what I had hoped for. That could be that the expectations are out of whack. Or it could be that these people tried everything they knew, which was limited. It didn’t move the needle, and everybody drifted apart. That frequently happens as well.

Get recommendations or referrals from the company on to who to talk to. And on their website, find client logos and their contact numbers because they’re going to give you their very best referrals, not the ones that went sideways.

BC: One of the things that apply to everybody in SEO, especially do-it-yourself SEO, there are a couple of levels. You think you know SEO, so you only want tools. Then, there’s the “I don’t know SEO,” so you want training. And then there’s, “I’m capable, but I can’t answer the tough questions.” That is a consistent thing.

In our company, we offer tools, and we have a three-day and a five-day training course of training people on what SEO is. Because if you cannot speak the language to your SEO company, they will snow you. You have to do it. Read the books or study enough to be dangerous, at least if you’re talking to an SEO company.

Take some training online. I have a 10-hour course on SEO auditing your site, like do-it-yourself. You don’t have to do it yourself, but you’ll know the right questions. 

There are some amazing tools if you ask which tools you should use for long-tail keyword research. Some of them are free, like AnswerThePublic, which will tell you questions people ask. 

You can answer those questions with great content and get a featured snippet, Position Zero. That’s another topic we should discuss at some point in this session. Also, Semrush and the Topic Explorer from Searchmetrics. 

There are so many amazing tools out there that are not common knowledge, and they’re not the Google Adwords Keyword Planner, which is problematic. There’s a great post on The Moz Blog [SEO] about how bad the Keyword Planner is from Google. 

For example, they’re buckets, and you won’t get an actual number. You’ll get a number you think is real, but it’s a bucket number. It’s 216,000 or something, and the next bucket is 241,000. It’s either that number or the next number. There’s no in-between.

It’s like, wow, how can you have something so imprecise? I’m just scratching the surface of the problems. If you’re not spending money with AdWords, you’re not even getting numbers. You’re getting massive ranges, like between 100,000 and a million. That’s useless. It’s a useless tool. 

I’ve got a free webinar on advanced keyword research. There’s a signup page on my site on stephanspencer.com. So you at least watch that. It’s like an hour of training on some of these tools.

Audience: Sounds good.

BC: One final comment and this applies to everybody. How many of you have phones? How many of you are doing voice search? Anybody?

DF: Focusing on it as something you’re building your business toward.

BC: You need to understand that a year from now, two at the outside, the keywords you’re picking today are not a keyword somebody would say. You have to anticipate they will say, “What is?” or “How do I?” So you’re going to have to start looking at whether or not your website ranks for questions. That is a big part of keyword research. I wanted to throw it in as a final thought.

DF: I’m adding a final thought.

No, I got one too. 

Audience: I’m going to sit down so I can take notes. Thank you.

DF: You don’t need to be an expert in these topics because I’m sure many of you feel this way. They’re not strong in SEO, but I need to learn more.

When you’re approaching the service provider, the only things you need to know are what your goals and metrics are and what the focus of your business is. You need to know very firm things about your business, like exactly who your client is, your customer, and your target market. You never want them to define that for you. They can help you refine it if they’re working with you, but you want to look for a company that can do more for you than just one thing.

Can they do content? Can they give you guidance on pay-per-click? Can they provide you with guidance on organic? Can they give you advice on the technical aspects of things? Can they pinch it and write for you? All these different factors because you need to apply this across the various areas.

Just take one or two of those areas and get some basic education so that when you come in and start asking those questions, you have a clear question and a clear answer that you know and trust for. And then you see their answer to the question because that will help you understand whether they have strengthened the area or are just trying to snow you.

The problem is that the jack of all trades is a master of none. For what was it 15-16 years, I ran Netconcepts, which was an SEO agency. We were not a paid search agency because it is such a different set of skills. If you spread yourself too thin, you cannot be an expert in all the other areas. 

I know a lot about paid search, but I don’t manage paid search campaigns, and I will never do that myself. I’m an SEO. If you try to hire an agency that’s supposedly an expert on all that stuff, you’ll get bait and switch. Some guru is going to sell you.

DF: That’s not true, Stephan.

It totally happens, right?

DF: I ran the operations of Bruce Clay. There’s expertise in every one of those areas.

The owner or the thought leaders, the one who’s going to talk to you and get you all excited, and they’re going to blow you away. Who’s going to write the SEO audit? Who’s going to do the link-building strategy? Who’s going to do the keyword strategy? Not he or she. It’s going to be the team.

You’re going to be assigned an account manager. You’re going to be given an analyst and a consultant. It’s not going to be the thought leader that sold you. It’s just that’s not the way it works. I ran an agency for 15 years. We were in three different countries. You cannot scale the thought leader across dozens of clients. It’s not possible.

I’ve got a different model where I only work directly with the client. I’m the one who’s presenting the audit, who’s gone in there and done the work. You can’t scale that. I can only do five clients at a time. That’s an ideal situation.

When I hire expertise, whether it’s Facebook advertising, paid search, or whatever, I want the subject matter expert, and I want that person to do the work. So I’ll use a bunch of different point resources and assemble a team, but they’re all contractors. That’s my two cents. I know that’s very different from the agency world.

DF: But it’s still valid.

Yeah, thank you.

BC: Okay, we have another candidate.

Audience: Do you find any negatives to shopping cart solutions like Shopify or big commerce regarding SEO? Second question: normally, when working on SEO, we look at Google, with Alexa and Siri, using Bing. How do we optimize our sites for Bing? And what do you see as the importance of that?

DF: Oh, I’m going to get on this one first. As Bing’s former public speaking person, let me clarify one point for you. There’s no work you will do that’s different for Bing or Google. The big buckets of what you have to focus on are the same.

I will say that Bing is more focused on the quality of the site, the answer, and these kinds of things because it has a smaller index than Google. On the other hand, Cortana, from Microsoft, does power Alexa and Siri with those kinds of ideas. This is where Bruce’s point of voice search becomes very important.

Did anybody watch what happened with Alexa over the holidays and the sales explosion? They are by far the market leader in the home hardware space. It was very interesting to see the news media talking about Google’s uphill battle with their Home device. It’s a good device, but Alexa is out in front now. That is becoming very prevalent, and people just ask out loud.

This is a first-world problem to solve because is it too much to ask that we press and hold a button on our phone to be able to speak to it? Because we all have to move past that and just speak out loud to the device. The cool thing about it is that it allows the consumer to get exactly what they want. The nebulous part is, is the consumer getting what they want? The choice is gone. It’s one thing that they’re being told.

It’s a very refined answer, which is where things get complicated because you have to be able to mark up your content, you have to be able to use schema and play in that very technical space to be seen to be that one answer that gets included.

I’ve always been a fan of focusing on the consumer. That is the message Bing and a long way around Google says. So you focus on that consumer, their journey, and where you’re in their journey.

Stephan mentioned this idea of understanding what you have and where it fits into that customer’s journey. When you know that, now you can predict where else they’re going to show up, and then you can position yourself there. And that is a winning formula. It’s a little pedantic, and it’s a little nonsexy. It takes away a lot of guesswork, but it’s much more successful.

BC: To answer your question about shopping carts, most of today’s shopping carts have facilities for SEO. Many of them, you can implement incorrectly. Faceted search will kill you because you’ll end up with 200 pages of the same content indexed if you aren’t careful. There are things to be careful of at an SEO level when you’re implementing some shopping carts, but in general, most of the shopping carts aren’t that hard to deal with from an SEO point of view.

There’s a big difference between search engine–friendly and search engine–optimized. If you get a shopping cart that is search engine–friendly, you are not search engine–optimized. All the configuration settings and so forth and all the different extra options you can pay for, like search-friendly URLs, plugins or options, or whatever, seriously, you have to pay for that. That’s crazy, so buyer beware.

Also, do your homework to make sure that you’re going to get search engine–optimized or optimal because search engine friendly just means you’re not repelling the spiders. Googlebot is not having a freak-out attack looking at your site and saying, “Wow, there are no edges to this site.” It seems like I can go on forever. Maybe it’s misconfigured, and you’re serving up a 200 status code instead of a 404, and it still says, sorry, this isn’t here, whatever, we’re out of stock or whatever, but it’s not a 404.

Google doesn’t trust your site because it doesn’t see an edge. It just goes on forever. It could be a configuration error. It could be just this shopping cart is not built by SEO experts, and that happens a lot.

My previous company, Netconcepts, developed our shopping cart software, Gravity Market. We sold it to another company. There weren’t very good options at the time. So we needed to build something with SEO from the ground up. Because if you end up with some solution that is not optimized and after the fact, you hire an SEO, and they’re like, oh, wow, this is a train wreck, they got this wrong, this wrong, and this wrong.

Think of it like this. You’re building a house, and you didn’t know to ask the builders to wire it with electrical. Now there’s no wiring, the house is almost finished, and you’re like, how do I turn on the lights? Oh, you didn’t specify that. That’s going to be expensive because we have to tear out the drywall, and we’re going to patch up the drywall. That happens all the time to websites.

What about this aspect? What about structured data markup? What about an automated generation of XML sitemaps? You didn’t specify those things. That’s going to be expensive because now we have to retrofit. Oh boy, not good. That’s one thing.

Another thing I’m going to rain on your Bing parade. What is the number two search engine? Hint, it’s not Bing. YouTube, yup. If you don’t have a channel, you don’t have videos, and you haven’t optimized your videos, you haven’t done your keyword research on YouTube. Do you know how to do keyword research for YouTube? Anyone? I need an answer quickly.

DF: No.

Google Trends. It’s a free tool that gives you data specific to YouTube searches. Under web search, there is a pulldown list, a select list, or whatever. Web search is one, as is image search, and there’s YouTube. Select that and then do your keyword searches.

Google Trends is not a perfect keyword research tool because you’re not getting real numbers. You’re getting percentages, which is useless. But you can compare two keywords and say, wow, this one sucks on YouTube. Nobody’s searching for that versus this other one, so that’s awesome. There’s so much we could talk about on YouTube.

Just bear that in mind. Do not neglect YouTube. You’re missing it if you’re putting all this energy into Bing and not thinking about YouTube.

BC: Okay, why don’t we see? Any other questions?

While he’s walking up, here’s a super ninja trick for Alexa for the Echo. Does anyone sell anything on Amazon? Do you have any products to sell on Amazon? You will massively increase your desire for your product by having a few friends who have Echos buy the product through an Alexa voice command. That’s genius. Because if people are buying your product with Alexa, trust me, that will look awesome.

Audience: I have two questions. One is Google has said that user experience matters a lot in engagement, but they’ve also said we don’t directly use your analytics data. Do you believe them because it seems like that’d be a good source of engagement metrics? My other question is, are there any link-building strategies you see a lot of success with?

BC: Okay, the answer to the first one is no. Everybody’s good with that, right? Do you understand?

What? Say what you mean; mean what you say.

BC: Okay, I’m going to start with the link building, and you guys can take the rest. Link building has always been problematic. According to Google, if you get a link, it should surprise you. You shouldn’t have planned for the link.

Oh, you shouldn’t have. I’m so grateful. Thank you. I had no idea it was coming.

BC: But in theory, everybody arranges their links. They promote their content, get people to document what’s going on, or promote their blog post or website. Generally, the link should be earned.

That’s a complex way of saying the right stuff worth linking to and then telling the right people about it, but don’t go out and buy links. Don’t go out and try to make your blog post footer comment with a link. Things like that are just way, way old. Nobody’s going to do that anymore.

What about private blog networks? PBNs are the hottest things, right?

DF: I’ll weigh in on the first part of this. I wouldn’t assume that Google uses your Google Analytics data as an indication because that data is often inaccurate when you get it. So it’s not like they would feed themselves inaccurate data points, so no.

Now, they accurately see the data related to you versus their tracking system and see it before you see it. They see it raw, which helps them understand how the engines work. It’s a series of networked services in the backend of a search engine and the raw data that comes in. All services like Google Trends, AdWord planner, and Google Search Console would then apply to that raw data feed to take the pieces of it to use in tools that they show the data to consumers.

Analytics would be a customer of the raw internal data feed from Google. The version of the data that they are allowed to show would be sanitized. Generally, it’s close because we’re talking about a relatively secure space. You’re logged in, and it is you. You’re verified. Therefore, they can show you what’s relevant to you. Except there are times when Google wants you to take action so that the data may be skewed.

Witness Google AdWords planning and the tools that they have in there. They very clearly want you to give them your money, and then they’ll give you something that you think is valuable. Of course, it’s not, and it’s a red herring. But the advertising works, so there’s still a quid pro quo.

They will also get data from other sources because Google is a domain name registrar. Does anyone put up a website and notice that almost instantly, it’s been indexed by Google, you can search it, and you’re like, shit, “I was going to do some testing on that, and now it’s in there?” It’s because they see the domain name changes in the background.

As a domain name registrar, they have access to that data. They see it instantly. As soon as the domain is registered, they’re watching. They’re constantly pinging to see what goes live, and the instant something goes live, it’s immediately put in the index. If it’s a crappy experience, you’re judged on that, so be very careful.

Put a robots.txt disallow up immediately.

DF: That won’t stop them.

Right. Let me just clarify something here. If Googlebot never spidered your site, robots.txt disallow will stop them from spidering, and the pages will not get indexed. But if Google has gotten in there already, you have to use not disallow, but a noindex.

This is a crucial nuance, but it’s one that developers and the system admins get wrong. They’ll use a disavow to stop the pages from showing up in the Google search results, and it doesn’t cause that to happen. It prevents Googlebot from visiting and updating the content. So you’ll end up with these horrible search listings that say that no information is known about this result because the robots.txt disallow directive or something.

Don’t do that. You remove the disallow from your robots.txt and then add a meta robots noindex tag in the HTML so that Googlebot can discover that and Bingbot, and obey that and drop that from the index. Noindex means no index. It’s not showing up in the search results, whereas disallow says Googlebot doesn’t visit here.

DF: Yeah. It means it’s a very subtle difference. It’s arcane; we’re search geeks. To us, that’s like a black-and-white moment. It’s like asking a programmer, hey, can you put a redirect in place? And they just tend to default to a 302 redirect, which is temporary. So it tells the engine, hey, something was here, it’s been moved temporarily, it’ll come back shortly, keep this in your index and wait for it.

That said, counterintuitively, 302 is more beneficial nowadays, not the 301s from an SEO perspective.

DF: Bing would hit the 302 five times. After that, if you were still 302, they assumed you meant 301, made a mistake, and then started passing value. Google has now stated that they pass value through the 302 as well. That one’s matured to the point where it’s been solved for everyone, but you don’t want to wait because that takes years.

But the 302 passes context for longer, and the 301 now. The anchor text used, which has some great keywords you want to rank for, tends to continue getting passed with a 302 versus a 301 that drops off quickly within just several months. According to the research done at Link Research Tools, a very large data set and very credible study, a 302 is more beneficial than a 301 from an SEO standpoint.

DF: Don’t go out if you need to do redirects. Don’t go and purposely build 302 redirects for these things.

I don’t know about that because I just had a client who switched from their 301s to 302s, and they got a nice lift.

DF: And they picked that backup. Wow, okay.

Always use the noindex. If you don’t want the pages to appear in the search result, just use the noindex. Disallow will work if Google has never spidered the site yet. The safest thing is never to use disallows; always use noindex if you don’t want the search engine to put it in the index. If you use a noindex directive in your robots.txt, that only works with Googlebot and does not work with Bingbot. Do this as a meta tag.

BC: Just one more thing to answer that specifically. If you’re developing a new website, create the robots.txt before you create the website. You can use disallow then because you don’t have anything being spidered. We’re discussing if I want to stop a new section of my website. The disallow becomes more difficult.

In either case, we’ve seen people build a website with disallow, move it to production without removing it, and lose their production rankings. You have to remember that if you want it indexed now, you either remove the meta noindex or remove the disallow, or both.

There’s no downside, though, to always just using the noindex and not doing a disavow. But you’re not going to get your desired result if you’re using both. So it’s one or the other. If you disallow, then you don’t let Googlebot in to see that there’s a meta tag because it doesn’t visit the page anymore. That’s a common mistake.

A couple of quick things about the previous question. Does Google use your data against you? Absolutely. Quick example, in Google Search Console, you’re doing some sketchy stuff with one site, and they connect up your whole network sites, and they all get penalized. That happens. How did they find it? Because you claimed all those sites in the same Google Search Console account. That was a really bad idea.

DF: That example could be even more innocuous. You could have three or four related websites. That is fine. Then for some reason, half of them start dropping off. Consumers just aren’t engaging with them. Google is now paying more attention to all your websites to see if that pattern will pan out because they must be ready to flip a competitor into that spot to satisfy searchers’ needs. So yeah, they can and use that. Google gets data from many different sources beyond itself that help it understand the traffic coming to your website compared to any other website on the net.

BC: Okay, a question here.

Audience: I got one question in 27 parts.

BC: How many?

Audience: Just kidding.

I did ask him to heckle before this session started.

Audience: All right, as AI becomes more and more personal in delivering search results, how much do you feel confirmation bias will be a problem in search results, in that people won’t get results that are relatively new for them? And if you think that’s a problem, what do SEOs do about that?

DF: Give me an example of how you see it going wrong.

Audience: The way I’m looking at this: we’re looking at Google just doing everything so personal that you’re only going to get results for things that you’re intimately familiar with or very, very closely related to. And therefore, you’re going to be missing results that fall outside of that spectrum.

DF: That happens now with the state of things. I see it as a leg, so I go in and do something. It’s like I’m buying new tires for my truck. I purchased the tires, they’re on their way to me, and I see tire ads chasing me around the Internet for the next three weeks.

Google has a similar scenario, and AI has the same problem. You were interested in this five minutes ago, five days ago. You may still be interested in that result when you ask this query. I’ve learned, and I have refined slightly. With every piece of data I take in, I get smarter about what I’m looking for, but I’m the pointy end of that spear.

Even AI can’t keep up with that pointy end of the spear until it’s seen all of the pointy ends of the spear. So it says, “Well, these group together and are likes, and therefore I can skew in this direction and see how they respond to that.”

BC: Let me give an example. If you search for a hammer—everybody knows what a hammer is?—the number one result is the Armand Hammer Art Museum at UCLA. The number two result is a vitamin. The number three result is a bowling ball. And the number four result is MC Hammer.

DF: Thank God. I was getting culturally worried for a second there.

BC: If it weren’t for MC Hammer, none would be tools.

Oh, come on.

DF: Anyone who builds a good tweet out of that gets one of Stephan’s books.

You can’t touch this.

BC: If I had searched for pliers, equipment, tools, a screwdriver, and a hammer, it would always be right. I wouldn’t get anything other than the kind of hammer I wanted. Web history has been here for about nine years, maybe longer.

Search for dolphins. If you were searching for Steelers or something beforehand, you would get very different search results than if you were searching for Wales.

Audience: I’m not necessarily talking about the type of results. I’m talking about the actual specific sites that you’re going to get that fit within the type of results that you’re looking for, where confirmation like, well, you’ve seen these sites before, so we’re going to deliver those sites again versus these sites, which might equally match your query. However, you haven’t been familiar with them.

That’s a real risk. It’s like the filter bubble you experience with Facebook; it will get worse before it gets better.

Audience: What do SEOs do about that if that becomes worse?

Build great content.

DF: It will worsen because we’ve seen a version previous to the AI and RankBrain, where they were guessing what it was. If you engaged with something, they took you down that path, and then you were unaware of what the other path looked like.

The way around that was to clear all your cache, clear all your filters, reset everything, and suddenly, everything looked different in your search results. In addition, you’d see some things that weren’t quite there but related.

The AI’s learning curve on that will be much faster than an algorithm and how it would approach that filtering, but it will get a little bit worse before it gets better.

By the way, we have five minutes left.

DF: Okay, I’ll be brief on this. It’s worth researching the three levels of artificial intelligence and their meaning. When an AI starts, it generally has the cognitive ability of a two or three-year-old. In the hierarchy of organisms, that’s advanced. However, that two-year-old still needs to know what a hot stove means. And they will learn that by touching the hot stove.

We’re at an eight or nine-year-old level with the AI and RankBrain. We saw a year-and-a-half ago when the search results sucked. That was the two-year-old AI doing what it needed to do to learn very quickly. The results are more refined now, which will continue that curve. Will it ever get ahead of where it needs to be? That is doubtful because the human mind is far advanced in its ability to connect dots relevant to you, the individual at that moment, and not share that externally. Nothing can keep up with that.

BC: Okay, we’ll take your question, but I want to make an announcement. We’re almost out of time. All of us would be happy to follow up with you. Take a business card, write your question on the back, and bring it up. We have plenty of information to send out, plenty of answers. You can come up and rush to the stage afterward. You can do whatever you need because right behind us is this little thing called a keynote. Nobody goes to those.

If you wanted a book and didn’t get one yet, who wants a free book?

BC: You guys haven’t learned yet?

Okay, you don’t have to rush the stage for that. I’m going to give everybody a free book. Everybody in the room. I know, it’s like an Oprah moment. Here’s the URL you’re going to write down. It’s a free electronic copy, and you can pick which of the three books you want, so you don’t have to get The Art of SEO if you don’t want to have it with you.

BC: Okay, question.

Audience: How’s it going, guys? I’m with an agency, and we’ve got a bunch of dating sites. One of the smallest sites caters to the polyamorous community. To make it more popular  (I suppose), we made a glossary of 130ish polyamorous terms that weren’t defined well on the Internet, so we defined them. Then, we built the pages with AMP. They’ve got 600–800 words of good content, and we linked them all pretty well.

Are they all on one glossary page, or are each definition separate?

Audience: Each definition is on a separate page. On a few of the terms now. It’s been up for six months, and we are Google’s preferred definition for X polyamorous term. It’s openminded.com if anyone’s curious. How can we accelerate the growth on the other pages that might be tracking along slower?

Are you talking about getting featured snippets? Is that what you’re trying to do?

Audience: Yes, stuff like that. Five are featured snippets at this point, but there are 125.

Just really fast lightning round here. First, you want to get as high a click-through rate as possible because you could be number 10 and take the featured snippet, which is position zero if you have an extraordinarily high click-through rate for a number 10 result. 

You worked to have an incredible title and meta description. Hence, you have a very compelling and nice snippet that gets lots of clicks and delivers massive value, not just a definition, but helpful. That’s one thing.

Also, you can look at your top competitors. Use a tool like Semrush. You can see all the organic keywords they’re ranking for and set the filter to featured snippets. Display only your competitor’s featured snippets and then target those featured snippets.

The ideal thing is not to try and go after featured snippets with a paragraph snippet but to have a better type of snippet because paragraph snippets are weak. For example, if there is a paragraph snippet currently, it could get overtaken by an ordered list, an unordered list, a table or whatever. 

If it’s a better answer to the user’s query to have it done as a table or as an ordered list or something, then create it with that and use the ol tags for a numbered list. A bullet list would be ul tags and li tags in the HTML. What else do I want to say about featured snippets?

BC: Why don’t I do this? It is time. Remember how we started? Everybody, circle tens. I wanted to make sure you didn’t forget that part. We like coming back and doing this. That helps them want us back.

Be sure and write in there, Oprah moment.

Yeah, Oprah moment. As I said earlier, write the question on the back if you have cards and more questions. We give away things. I need cards, but I give away things. Stephan gives away things. Duane just talks you to death.

DF: It doesn’t cost anything, though.

BC: He’ll answer anything you want to ask. Come on up, rush the stage. Thank you for attending our session.

Yeah, thank you.

DF: Thanks, guys.