Ready to take a bunch of notes that could help give your site a competitive advantage in the SERPs?
Tony Hill joins the Niche Pursuits podcast to share his experience recovering a website from a Google update.
Tony’s been building niche websites since 2005 and is very familiar with core updates.
For instance, he had some health sites doing big numbers wiped out by the Medic Update. However, it was the May 2020 core update that was particularly devastating, causing his main 18-year-old site to lose almost half its traffic in just a few weeks.
To recover from the update, Tony and his 10-person team formulated a bullet-proof game plan to recover the site.
The strategy included a comprehensive process to improve the site’s existing 1200 pages.
They started by using Google’s NLP API to analyze their content and improve its salience score. They found that word order is crucial to help the algorithm better understand your content. So at the sentence, paragraph, and article level, they made a deliberate effort to put the subject first.
An academic tool called Sketch Engine also helped them to analyze language patterns in top-ranking articles on some of the biggest sites in the world. And found that using second-person language and specific, confident wording can improve content quality.
So Tony and his team then reviewed their content to make it clearer and more confident, using specific and directive language. They didn’t delete any articles but rather cut out fluffy sentences and paragraphs and added more information when needed.
Other specific tactics included improving readability, adding more visuals, and updating outdated information.
They also added more categories and subcategories to their website to help users navigate content.
By focusing on updating content one category at a time, they continuously experimented and tested to see what worked. They also read competitor’s articles word for word to see how they structured their content to make improvements to their own.
They also worked with expert writers and editors to ensure quality and establish a persona for the site. They also created a contributor’s page, an Ask an Expert section on the site, and paid experts to answer questions on Quora and Reddit.
He also made a point to flesh out the schema of every article, going above and beyond the typical author schema and entity data.
They also made great use of internal linking, aiming to include at least one internal link in every paragraph of an article, sometimes even linking to the same article twice in an article.
Considering their substantial success in Google Discover, they also analyzed their most successful articles and found that related topics were being featured, so they created new articles on related topics and grew their success in Google Discover.
And after a year of consistent effort and multiple Google updates, Tony’s site not only recovered its lost traffic but experienced significant growth, now receiving nearly 8 million page views per month.
Tony’s interview is a high-level masterclass in site operations and site recovery that can help site owners at all levels. The way he thinks outside the box to improve the strength of his content will not only inspire you to do the same but also give you actionable steps you can start using today!
Topics Tony Hill Covers
- Building his first niche site in 2005
- Google Updates
- Prioritizing updates over new content
- The history of his 18-year-old site
- Importance of testing and taking action
- Updating one category at a time
- Comparing your content to competitors
- Google’s salience score
- Putting the subject first and foremost in your content
- Adding more categories and sub-categories
- Being direct and confident in your language
- Use of experts and personal experience
- Building out a complete author schema
- Liberal use of internal linking
- Google Discover tips
- And lots more!
Links & Resources
Watch The Interview
Jared: All right. Welcome back to the Niche Pursuits podcast. My name is Jared Baumen. Today we are joined by Tony Hill. Tony, welcome.
Tony: Hey Jared. Thanks for having me.
Jared: Thanks for being on board. This is this is gonna be such a good one. I was telling you before we hit record, just how excited I am because this episode is going to be, A really good mix of tactical, know-how tactical approach that you took and then just a good shot of inspiration.
I think a lot of people are gonna be really inspired by your story today and and, and kind of what you made out of a tough situation. So anyways, cool. I’m just really excited to have you on, on board today, so thanks for coming.
Tony: Yeah, for sure.
Jared: So a meat and potatoes of today is how you recovered a website from a from an update, from a Google update.
But before we get into that, like give us some backstory. Set the stage not only for maybe what you were doing prior to building websites, but also leading us up to kind of the situation we’re gonna have in front of us
Tony: today. Okay, cool. Yeah, so, you know, I launched my first niche website back in 2005.
I launched it with a coworker. He and I are working for a web design agency. And you know, we were tasked with helping clients with their seo and we had no idea what to do. And so I turned to seo book.com. I dunno if you’ve heard of that in Aaron Wall back in the day. I printed it out as a huge stack of papers like that and So just started going through it and learning everything I could about SEO and just fell in love with it.
And so my coworker and I were like, man, like why don’t we just try building our own website? We can rank it on Google, see what we can do, make, make some money through ads. I was aware of ads since, cuz I started, I. My very first website when I was a senior in high school back in 2003, and I set up absence on it and started clicking on it every day after school.
Earning, earning you know, maybe 20 bucks a day. I thought it was really cool until my account got banned a few months later. I got a lot of other people, I’m sure. So I learned my lesson there, but you know, so that’s how I got into it. And yeah, we started our, our first site together and we got into the medical niche.
And so we had a lot of health websites that we had built up. And, you know, back then it was pretty easy to rank websites. You just needed keywords sprinkled throughout. And not a whole lot of content, and it was easy to pop up as number one. And so, you know, I’ve been doing this for about 18 years now and I’ve had a lot of ups and downs.
I’ve been hit by all sorts of Google updates, from panda to penguin to the medic updates and core updates, and so, Yeah, I, I understand the, the peaks and the valleys of being in this business. And so I’ve had good days and bad days, and right now things are going well with my primary site and recovering from a core update.
And so I, I enjoy th those highs. I, I try to take time to enjoy it. It’s, it’s so easy. Just get to focus on the next thing and not to like stop and take a moment and celebrate some of the wins that come along the way.
Jared: Take a moment to talk about maybe the current nature of, and this is just from your opinion, but the current nature of Google’s updates that happen.
And then kind of going back to more of the more the bigger, more landscape shifting type updates that happened that you mentioned Panda Penguin, the medic update, like mm-hmm. You know, you’ve been doing this for, for a long, a long time. How do, how does your approach to these updates kind of change as they’ve changed over the years?
Tony: Yeah. For one There’s a lot more nuance to it. You know, they, everyone, it was really more rare for Google to throw out an update within a couple weeks or even two updates at one time, like what they can do these days. And so it was a few times a year they’d have some big updates that we would be, have to be prepared for.
And. And, and there were less people you know, reporting and diving in on the details and sharing what they were learning. And so, and there were some forms I was a part of and tried to learn from. But now there’s just so much information out there and people sharing information, which is super helpful to, you know, try to diagnose and figure out what, what’s going on.
But right now, I mean, it just feels like everybody’s just getting slapped left and right constantly throughout the year. Trying to keep up with all these updates is, is crazy. And You know, it’s, it’s a little more nerve-wracking these days than it was back in the day with, you just don’t know what’s gonna happen.
I’ve seen great sites just tank, and then like a month later they’re right back up. I mean, I’m sure you’ve seen some of those charts. It’s, they fall off a cliff and they just come right back up, fall off the cliff and right back up. And so my approach here is, you know, The, the long term for sure. I mean, already being, having one, one site in particular that I’ve been working on for 18 years, I just have a different perspective on how to approach the site as a whole versus someone who just started a site maybe in the last couple years.
That’s, you’re still so, so early in like that baby phase, that newborn phase almost. But I’ve got a grown adult now, eight, you know, 18 years and so. Yeah, there, it’s, it’s constantly evolving. If you were to go back to, you know the archive the archive org, I believe, and you can look up you know, screenshots of sites over time.
Yeah. And you can just see how my sites have evolved and, and so much since day one. And so it’s just that constant and never ending any improvement there that we’re focusing a lot on. And so just trying to find new, new things we can try and trying to find inspiration wherever we can.
Jared: It’s so true that man, you have an 18 year old site that is absolutely fascinating in so many ways that you’ve been working on the same site in in one regard for that long.
Well, okay. Let’s we have a lot to get into today, so I don’t wanna, we can probably talk about the last 18 years of SEO for. At least half the podcast, which I would actually enjoy. But too many good things to, to talk through. You’ve been great about generating a really long list of the, the processes or the tactics you went through.
But before we get to that, let’s set the stage for this specific website we’re gonna be talking about, because we’re talking about how to recover from a Google update. Yep. Let’s talk about maybe where the site was, some of the, the, the, the makeup of the site prior to it getting hit by the update.
Tony: Yeah, you know, the site was in a place where prior to the update we were focused a lot on new content. You know, I, I can also suffer from shiny object syndrome and trying to tackle the next you know, big keyword that’s starting to pop up. And looking at my competitors and seeing what they’re doing, and there’s.
There’s always that gap between, you know, some of the topics they’re covering and what I’m covering. And so I want to close that gap, but I also want to, you know, post new content that they’re not posting yet. And so I think I lost sight on. Keeping up with the existing content. And I was too focused, put too much, too many resources onto the new content.
Mm-hmm. And so that’s once, once we got by the, the core update in May, 2022 I took a step back and I was like, all right, there’s something about our content here that Google doesn’t like. I cuz the entire site site just dropped. I mean, there were some sections and some articles that dropped more than others, but overall everything went down.
So fundamentally I knew. There’s just some lack of, of trust and maybe some freshness that Google’s looking for, and that we’re just not meeting that a as well as users, right? At the end of the day, look, you know, people who are visiting the site, I mean, they’re gonna want up-to-date information.
They’re gonna want information that’s helpful, that’s accurate. And for some, Depend on the niche that you’re in. Some niche requires you to update your content more than others. And so that’s the position I found myself in. And so that’s where you know, we pivoted and we took our, our publishing schedule for new content way down to just one day a week.
And the rest of the time we dedicated all of our. Resources towards going back through all of our existing content, one section at a time, one article at a time, one paragraph at a time, one sentence at a time, and in some cases, one word at a time. We got pretty detailed in how we approached it and looking at our content through a new lens.
Jared: us about the. Tell us about the like where the site was at prior to May, 2022. I mean, in terms of traffic numbers. Mm-hmm. Number of pages, number of articles you had, just to give people an idea of the scope and if you can, you know, tell us a little bit more about the niche it was in, even if it’s even for, from a really broad standpoint.
Tony: Yeah, for sure. So the site was doing maybe around two, two and a half million uniques a month before we got hit plus a, a lot more depending on how we were doing with Google Discover. So, you know, a lot of people know Google Discover that algorithm is closely tied to the, their core update algorithms for their primary search engine engine.
And so we were getting a ton of traffic from Google Discover anywhere from, you know, 10,000 people a day to a hundred thousand people a day if something really blew up and, and could discover. We have about 1200 articles on the site. It’s in the lifestyle category, and so we we’re focused a lot on texts, but as well as media.
So it’s very visual. But also there’s still a lot of, of details that we try to communicate through texts as well. So with 1200 articles yeah, that was. That was a lot to process and, and get through and look at. And honestly we haven’t even finished it cuz it was so big and we were just spending so much time on it.
But these days you know, bec once we recovered from the core update I was, I would be happy just to get my traffic back, but we ended up, you know, at this rate, I think 50, 60% higher than we we originally were. So I’m pretty stoked about that. So right now we’re doing. Close to 4 million uniques a month, over 8 million page views.
And we’re back in Google Discover. It’s not quite as booming as it, as it once was, but I’m, I’m starting to see some patterns in Google Discover that I am going to be focusing on and seeing what What I can get to, to work. I’ve already ran one test already, and, and it, it popped in discover and, and did pretty well send like, you know, 30,000 people to it.
So that part’s fun too. So those
Jared: are some big numbers. I think that that’s one of the most important things to take away is I mean a larger site, 1200 pages is no simple feat to, to make updates too. But you were getting millions and millions of page views per month and then that just plummeted.
What kind of, what kind of plummeting happened? After the May, 2020, like how bad was it?
Tony: Yeah, it was, it wasn’t as bad as when Covid hit, if you remember, in March and, and April of 2020. That just tanked completely. And so that was probably the worst day. So this, so this one may wasn’t quite as bad.
You know, we were able to, to keep things afloat, keep everyone on the team. We were, we were scraping by you know, just. Just losing like almost half our traffic, you know over the course of a couple weeks is pretty brutal. But you know, the nice part about running niche sites is that they can be pretty profitable, especially with pro with your margins there.
You know I d how much you’re investing in, in content and whatnot, but. So we were prepared for it. I, I wasn’t shocked and I’ve, I’ve gone through it before. I’ve seen, I’ve seen worse days. You should see the analytics on my medic medical websites. Those went from, you know we were doing 20 to 30,000 visits a day, and they tanked, you know, to, to maybe a hundred.
So, oh my gosh. Yeah,
Jared: that one was pretty, I remember the medic update that was absolutely life changing from a an SEO perspective. And I mean, if you happen to be caught up in that medical space, who it was, unless you happen to be lucky and have been doing the things that they now required, you were just Yeah.
Swept off the, off the map. Yeah,
Tony: for sure. And you know, At that point, we were actually doing a lot of things that they were requiring before a lot of other health websites were. So we were getting our content, they were ghosts written, but they were then reviewed by a doctor and then reviewed by another doctor.
So peer reviewed or had another doctor just double checking everything. We had their names on the articles of, you know, who wrote it, who reviewed it. We had bios. We were doing all that back in 2015. It still wasn’t good enough. A lot of it, it just came to, you know, having that trust with your links and being one of those high level sites that, you know, you’ve got back links from the New York Times or WebMD, et cetera.
I mean, we just weren’t operating at that level. So I think that that was a big part of it too. I’d imagine the links are, At least back then and, and probably still now is a huge part of, at least the medical niche for sure. So, mm-hmm. We bowed out at that point, I knew there was just, there’s no way to, to turn this ship around if we were already com we’re doing a lot of things that Google was recommending and we still tanked.
Jared: That’s the tough part about updates though. It’s a good transition for us to talk about, you know, your site now and what you, the road you took to recovery. Mm-hmm. So let me use that as a transitional point. Before we go through every step you took, like how did you evaluate the site? That’s one of the hardest parts for someone, especially someone as you mentioned, who maybe is a couple years in on their journey.
And it’s very difficult to take a step back after getting hit by an update and say, mm-hmm what’s wrong and where do I need to go to go forward? Oftentimes you, you know, you. You kind of feel like the content’s pretty good. Mm-hmm. You see competitors that are ranking without many more back links than you, you, you know, see there’s all these feedback that you’re getting, whether from yourself or from third party data that would.
And make it kind of hard to understand how to set a plan and where to go. How do you recommend, or how did you analyze your site to come up with a list of things to do?
Tony: Yep. So beyond just the basics of, you know, looking at Google Search Console or looking at Google Analytics to look and see which pages dropped, which categories or sections of the site dropped When it came to analyzing our, our content and what we’re doing wrong honestly, there were a lot of little experiments that I’ve been doing along the way prior to this update hitting where I would learn something from a form or from Twitter or a podcast and I’d be like, oh, I should try that out.
And so, I would test it on an article or two and then look back if I remember to see, oh, did do anything with the traffic, you know? And and then I just kept moving forward and trying different things. But I wasn’t consist one, I think the problem is that I wasn’t consistently looking back to see if those if those things moved the needle are not, are not and then also just taking, you know, good notes and.
Then also trying to out more than just one article. Cause sometimes you just get lucky and make some tweaks to an article and it pops up. Yeah. And then you tried it again on another one and, and nothing happens. And so there were just a lot of little things. So, I mean, I did take some notes along the way of things that I tried and I just went back through them like, all right, which of these, these things that I’ve tried that I could really.
Expand this and, and implement it in more articles and see what works. And so there were, honestly, there were like 30 things that I, we we’ve done over the last year or so. And it was frustrating because. We, we tried so many and like nothing was changing. I mean, there was a core update in last fall and nothing happened with that core update, but we just kept pressing forward.
So, you know, my advice is, at least what worked for me is taking notes on. What you’re learning out there, like through podcasts like this and, and implementing it. Try it on at least a couple of articles and see what happens. Sick notes on that. You know, what, what you tried the date you did it. Set a reminder on your calendar if you need to, to go back and, and check analytics and or your rankings to see if anything happened.
And honestly, in some cases, my experience, it’s nothing’s gonna move the needle until the next core update. And so that part, you know, Google’s a bit of a, a black box, and so it’s just a lot of, a lot of testing and experimenting and not just buying an SEO course and just perusing through it and not implementing anything, or you just do it half baked, you know you’re not fully implementing what they’re suggesting.
And I’ve, I’ve done that all, all throughout the years. And so Just my encouragement for others to not do that and to actually take action and see it through I think is, is a big part of it. I mean, then the information’s out there, the knowledge is out there. I think just people really struggle taking action.
And where, where do you start? Like, I mean, honestly I just looked at what was the category on our site that was the, the lowest performing one. The one that we honestly neglected a lot and like, let’s just start there. That was our weakest link. So let’s, let’s focus on that category and then move on to the next one.
And so that’s what we’ve done. We’ve taken it one category at a time. Rather than like choosing random ones or the ones, even the articles that were the hardest hit, we still, there were some that we still haven’t gotten to and they’ve seen some improvements for sure, but they haven’t fully recovered yet, even from this past core update.
So, So we still have more work to do. And that was indicative to me that, alright, so there are these things that we’re doing that is tied to this core update because they’ve made Jurassic improvements in, in full recoveries versus articles that we haven’t touched. And they didn’t do that from this last core update.
I was gonna
Jared: ask you about that. Yeah. I’m glad you pointed out. For starters and if you’re new hearing that like oftentimes you get hit hard by a core update, it’s gonna take another core update to recover. Not always, but that’s by and large. Typically what happens, and like you said, sometimes it takes a couple core updates.
Don’t know why I could theorize, but that’s not what people are here to listen to. So sometimes you can take a couple core updates. So even if you make a bunch of updates after a core update and then another one happens and you didn’t rebound, it doesn’t necessarily mean that what you did isn’t gonna work.
It sometimes just takes more than one core
Tony: update. That’s right. Yep. That’s exactly what my experience was. So it’s just not losing sight of, of what it is that you want to implement. And you know, I’ve, I’ve got a team behind me. There’s 10 of us. And so it’s nice that we’ve, we set weekly goals. We hold each other accountable in that way.
And so when we’ve got a project, like we set it up as a project and we’ve, we’re tracking it and everyone has a role and like, We’re gonna see it through. So having that level of accountability has been huge for me personally. And so I know there’s a lot of people out there that are running their own site on their own, and they, that is one component that some people I’m sure are missing and that would really raise their game as far as, you know, taking action and, and holding themselves accountable.
And, you know, maybe that’s where they can join, you know, forums or mastermind group appears where, you know, they can have that accountability.
Jared: That’s a good call. Yeah. Yeah. You can sit down, you can sit down and look and stare and analyze and run reports and sift through age refs or sim rush and pour over Google search console data.
But at some point, I, I kinda like what you said, like I just looked at the worst, the one that got hit the most and I started there.
Tony: That’s right. Yep. You know, so I think. First and and foremost, the I’ll mention that, you know, our content has, was a mix of being written by experts, but also some freelance writers.
And so we, that was something else that I was looking at and, and reviewing our content, and I could see the weak spots in it. And so I know there are other, other site owners who write, just write content all on their own. And there are others that hire, you know, some writers to do it for them. And so it.
I think that’s where if you, either yourself or maybe someone else that, you know, ask them to read one of your articles and, and, and start there and, and give feedback. And honestly, you’ll start to see patterns. That’s, that’s what I’ve noticed. And so once I started seeing patterns where, like I sat down and I was reading these articles, I haven’t, I’ve never written and.
I’ve, I’ve scammed, I’ve skimmed them before. But they, before they were published, but I didn’t look at word for word. Mm-hmm. What exactly was being published, you know? And so once I took the time to read them word for word, and then I went to some of my top competitors, I, I looked at some of their articles and read them word for word.
I don’t, I mean, be. Before that, I couldn’t tell you the last time I actually read a competitor’s article. Every single word. Some of these articles are like 5,000, 8,000 words long. So I mean, it takes a while to get through, but it was an eye-opening experience, just reading their content and like literally I would have two windows open my, my article on the left and, and my competitor’s article on the right.
And just kind of going, going through and comparing and seeing like, okay, I, I see their approach there and, and, and how they, they. Put this paragraph together or structure the article or how they wrote, wrote this sentence. And, and they honestly, they did an easy, they did a better job at making content easier for people to read as well as I think for search engines like Google to read and understand.
And so that, that’s the trade off of when you have writers either who don’t have the full expertise or experience They, they can get kind of wordy cuz they’ve got, a lot of times they have a quota to meet. Right? And so they can put in some fluff sentences. So I saw a lot of that. And then we, we have experts who also write content but they’re not great writers, you know, and so trying to improve their writing as, as well, something that, that we focus a lot on.
Jared: Well, you published a really awesome breakdown of what you did on Twitter, and I’m gonna include a link to that specific tweet that you had in the show notes. So anybody who’s listening mm-hmm. Or watching, you can open that up. And I don’t wanna make you read it line for line necessarily, but maybe we could get into the tactics, like, and the exact set of things you did across the content to improve it.
And again, mm-hmm. I mean, it worked. It’s always hard to understand what of it was the, the key thing, probably all of it, or some collection of, most of it was the big, the big, the big driving force. But here we are, what, 11 months later, almost 12 months later. And you’ve had, you know a massive recovery.
But I, I’d love the detail and the depth you went into. Maybe we could start to, to walk through some of those different specifics that you did.
Tony: Yeah, for sure. So. One of the, you know, I made a lot of updates to the content or my team really, they did so much of the work. But looking at how our. How we were writing our content I don’t know if you’re familiar with Google, is nlp.
They’ve got a, their API has a demo and you can pay, send content in there and it will analyze it and it will give it like a salience score. And that really is, is looking at. It’s gonna break down basically all your entities or all your topics or subjects in, in whatever content you give it. And then it’s gonna give it a saline score.
And that just really represents how important that particular word or entity is in the content. And, you know, I went through and started pacing in not the, well, I tried it with the entire article, but that wasn’t really the helpful. So I started pacing in just paragraphs of my, my content into, into their free demo.
And it was showing me. Entities that really, from my opinion, in perspective, were not what we were primarily talking about. But Google’s NLP and what, what they were trying to understand if, if they were seen in a completely different light. And so what I’ve learned is that words matter and word order matters.
And so we, again, I just started seeing patterns. So I would just take more and more paragraphs, put it into the demo, click the button, see what, see what it thought was the most important. Topic or subject of whatever it is that I had, paste it in there. And so understanding that, like, just tweaking some words, something the subject structure or the order of things made a drastic difference.
So, you know what, one it tip that I, that was really helped me is putting what, what are the subject of whatever it is you’re talking about in a, in a paragraph or even a sentence. Put it first and foremost. I mean, even in your article, you want your most important information up at the top. But even on a sentence level cuz it changes its meaning and understanding, not only for people, but also for for search engines like Ako who’s trying to understand what it is you’re talking about and the words you’re using.
Jared: So, and, and I’m familiar with nlp. I, I, I mean, just to be perfectly clear, you’re basically saying like, Hey, if we do a paragraph on horses the ordering of the words actually affected the type of horse entities or related entities that Google’s n l p picked up on and, and highlighted and, and gave a higher salience score for it.
Tony: that’s right. So the, so the example that I’ve shared in the past is. So say you’re writing an article about hiking, snow covered mountains. If you write a sentence that says it’s important to have a waterproof backpack when hiking snow covered mountains. Like if it’s important to have a waterproof backpack when hiking snow covered mountains.
If you run that through the NLP demo, it’s gonna tell you the most important topic or word is the waterproof backpack. And not the snow covered mountains, but the article, the whole intention is, you know, you’re writing about hiking in snow covered mountains. Like that’s the primary subject and the most important part of the article.
And, and really the sentence, but. The Google NLP is seen, that it’s the waterproof backpack is the most important. But if you were to just reverse the order of that and you say like, when hiking in snow covered mountains, it’s important to have a waterproof backpack, then that’s when the NLP tool will show you that snow covered mountains is gonna have the highest salience and, and as the most important part of that sentence and what you’re talking about.
So yeah, the word of order Matters. It’s crazy. But that’s saying, I don’t want
Jared: to get into like a specific like pin your backup against a wall kind of question because I’m, this is one of those, I’m sure the answer is, it depends the classic, but is it typically the case where moving, I mean, kinda like what you said, moving the most important words and moving the most important subject matter, front of the paragraph and front of the sentence mm-hmm.
Will typically have a positive impact on the salience score and thus what Google interprets that paragraph to be about.
Tony: Yep, that’s right. And what it thinks it’s, and how it’s going to weight those words. Right? Cause it’s turning all those words into, into numbers. And it’s trying to figure out how they connect and, and what importance to give to it.
Yep. And you know what we also found that if you ever start a sentence with the word if I would actually move that to the backside of the sentence because generally you’re setting that up like you’re saying if, and then you’re setting it up for. Something important, and that’s at the end of the sentence.
And you wanna move that important thing into the beginning. And I think really that’s just helpful for, for humans as they’re reading and you’re saving them time and like you’re catching their attention with whatever it is that’s most important. Put that in the beginning of your article, but in the beginning of the paragraph, but in the beginning of the sentence, it’s
Jared: like, my wife always says when you say the word, but in a sentence, you can ignore everything in
Tony: front of it.
Ah, that’s a good one. That’s a good one. That’s a little bit more
Jared: marital advice there though, but. Well that is a really, really fascinating tip that I haven’t heard, and my mind is already racing for all the different ways I write content myself that have probably confused Google over the years in terms of what I’m actually trying to talk about.
How did that roll out in a practical sense? Like were you rewriting. Paragraphs and sentences with this in mind and, and, and was it really just, let’s go back and really try to make sure were you we’re gonna get the salient score higher for the, the words and the phrases we want. Were you cross-checking that back into an lp or we’re using, were you using some sort of software?
A lot of us are familiar that like a surfer, SEO will have an lp as a part of their reporting and stuff. Like, I’m curious how you kind of started to roll this out at scale
Tony: now. Yeah, I made my editor do it by hand. One paragraph at a time. They weren’t too happy, I’m sure, but they did it. They’re great.
Super helpful. Wow. But that was, that’s precise. They’re, they’re, they’re trained, you know, and they, they can spot it a mile away. And so sometimes it’s just a repetition of doing things for a while. And then at over time you’ll be able to spot that on your own. And, you know, we did the same thing with Hemingway.
I don’t know if you’re familiar with the Hemingway app. Mm-hmm. Great tool to help you really tighten up your sentences. Again, not make them too verbose, easier to understand. And so we did the same process. We took that paragraph from the Google NLP tool. We past it in Hemingway and followed all of its instructions.
And this was all be before chat G P T came out. And so now we’re saving a ton of time with a lot of this stuff because of a chat, j p t.
Jared: I wanted to ask you about some of the other updates you made. And so I might ping pong back and forth if that’s okay. I’m just kind of, yeah. I’ve got your tweet up in front of me and I’ve got kind of starred the ones I really want to talk about.
Cool. Some of the things you talked about are actually adding the number of categories. Mm-hmm. To your website, and then also adding subcategories. I don’t wanna, I don’t wanna say it’s heavily debated, but the, the idea of categories and how important they are is certainly something that a lot of people would put in the tertiary of importance, right?
Mm-hmm. But you seem to really talk about this and featured as something that was important. Why was that important and why did you feel the need to add triple the number categories? And then also it sounds like add subcategories.
Tony: Yeah. You know, one of my top competitors has been doing this for a long time, and I thought it was crazy.
You know, those mega menus can be overwhelming. And honestly, I don’t know if how many people actually use the menu navigation. Like they’re usually, they’re landing on, on the article mm-hmm. From Google, social media, et cetera. And so they’re there to consume that particular content, but are they really using the navigation?
I, I don’t actually, I, I know and it’s, and it’s not too often. But what I found was if they do decide they’re using the navigation, then we’ve structured an order to the entire site in a very logical way. As far as how we wanted to break it down, depending on the different types. Of people that are coming to the site.
So depending on like, say their location or their, their age or their gender or their marital status, like there’s just all sorts of different aspects or attributes of people who just have a different background and, and maybe a different lens or perspective that they’re looking at your site through.
And so we started to break down and create all these categories to help really. Narrow it down our content so you can, you can find it in various ways. So, you know, we have, we have articles that are in multiple categories, and so you’ll, you’ll see, you’ll, you’ll see them, let’s see, yeah, several places.
So yeah, we probably went from maybe 10 categories, like primary categories to, we might have like 50, 60. I mean, it’s huge. The menu’s huge. And again, we use subcategories as well. You’re just further breaking it down. And so a lot of it is redundant. And, and some of this I was afraid of like cannibalizing some of the articles because there are some articles.
Yeah, there are some articles that are basically it’s targeting the, the category as well. And, but they’re not really competing in Google. Google knows better that the article is gonna be more helpful for, for people than say the category. So it, we’ve kept those index in Google and they, it, because I think they serve the user well for those who are trying to navigate the site and, and cuz people can approach something from different paths.
Kind of like, you know, hiking, snow covered mountains. Like you, there are different paths you can start and there probably just isn’t one path. And so being able to be a guide for our readers depending on what path they come from to get them to where they want to go. Mm-hmm. Mm-hmm.
Jared: Yep. Okay. Very good.
Wow, that’s interesting. Let’s see. So we, we talked about a lot of the specific nuance updating you did to the articles. Was there anything else as a part of your article process and your article update process that you were doing beyond the NLP type of
Tony: stuff you talked about? Yeah, there are a couple other things that were very eye-opening.
So I, I got curious. I learned about this tool called Sketch Engine, and it’s primarily used academically to understand language and words. But what I was able to do is I got curious and I took some of the biggest. Keywords out there. I’d say like you know, best mattresses or like best CR credit card you know, some different medical conditions or like, you know, laser eye surgery.
I took some of these big keywords where some big sites like New York Times, Forbes you know, Healthline, et cetera. That were ranking number one for these huge keyword that, you know, a lot of ’em, they mean they’re gonna require a high eat from Google to rank. And so I took their, a lot of their number one rankings.
I took their content from their number one rankings, and I ran it through this tool and created basically this cor corpus of all these words. And it what showed me. The words and phrases that they were using the most, and I started to see some patterns to it of specific words and phrases that I realized like, you know, I think an expert like this is one that these are phrases that an expert would say.
And then, you know, two, the. The, the language that they were using was more like second person language than say first or third person language. So it felt like they were, as you were reading their articles, they were talking to you, not just to people, to a group, but they were talking to you. And so that was eye-opening to, to see some of.
The, the words that they were using and the phrases they were using and how confident they were in their language. So saying something like, you know, you can probably do that a few times a week is a little vague. And, you know, if, if that’s something that your doctor is telling you and they’re prescribing you some medication, like that’s kind of, kind of sketchy right.
But instead they would say, no, you must take this at five milligrams. Two times, at least two times a week. Right? So they were very specific and clear on how much and the frequency. And so that was another pattern that I noticed. And so that’s where we were going back through our content and looking at how can we be clear how can we be, be, be more specific and use more confident words instead of saying you you can try or maybe, or consider.
And instead using words that overall just came, come across as more confident and be more directive you could say. And
Jared: you’re in the lifestyle niche, so you’re talking about being very directive from a lifestyle perspective. Right. This doesn’t, you’re saying it didn’t just apply in your analysis to say a medical query or a scholarly query, but really that mm-hmm.
That was consistently coming out of Yeah. All of these big publishers.
Tony: Yeah, for sure. I mean, and even if you’re like, you’re in the travel niche, if you are, if you, you had a great experience you know, at a, a tea shop in, in London and you’re gonna say like, you have to try this particular tea, right?
Like, just even that excitement. And passion for it and just how confident they were at a sense, a strong signal just to the, to the human of like, okay, this is, this is something that’s worth me trying. Just how excited and confident that they were in it. So I think that could apply in all sorts of niches.
Jared: Let me ask you about your evaluation of the content and if you, if you ended up pulling any articles. And deleting them. If you ended up combining any articles or on an article level, like did you delete swaths of an article and rewrite it or, because so far what we’ve talked about is a lot of like real updates to the way stuff is written.
Mm-hmm. Which is, I’ve loved it. I’m just wondering like actual content wise, did you delete any, remove any, whether it’s U R L related or specifics inside of an article?
Tony: Yeah, we didn’t fleet any of our articles. There were a few that we had redirected cuz we thought that they were a little redundant compared to another article.
But for the most part we kept everything. And so we just went through and yeah, we, we cut out the fluffy sentences and the fluffy paragraphs, like just looking at asking ourselves the question like, is this really helpful? Or are we just gibber gibbering. You know and so yeah, we deleted a, a lot of paragraphs and a lot of sentences.
But then we also added in more, more information. And, and again, we work with, with experts that produce most of the content these days on our site. And so Once I had to figure out what it, how we wanted, what kind of information we wanted to include, how detailed we want to be. And then I like created examples to send to our, our expert writers of like, here’s, here’s a way to approach the writing going forward.
And here’s some examples and could give them that template so that they can, they can run with that. And then we still have an editor who receives all that and just double checks everything. And so yeah, it was just painstaking of just going through one, one sentence at a time, you know, but I, I think, I mean, you know, this, this is my livelihood, right?
So and supports 10, 10 people. And so, you know, we took it pretty seriously there. And we knew like this could take a couple years to get through. And so it was, I was pleasantly surprised to see that, you know, we maybe got halfway through this, through this process to start to see a recovery there.
Jared: How did you know this was the right approach? You know, because it would be so much easier to you know, build more back links to it you know, delete a lot of the content mm-hmm. And start writing new content with a better perspective. It would be like, those are easier solutions. Right. This is so, like you said, painstaking.
Mm-hmm. And. Expensive, probably because of the painstaking nature to it. Like what, I’m just curious why you were so confident that these things would have this positive impact, which clearly they’ve had.
Tony: Yeah. You know, there’s always that mom test that you can apply. Like if you were to show your mom one of your articles on your site, would, would you be proud of it?
Would you be proud to show your mom? And so as I was reading through some of these, my answer was no. Like, I, I wouldn’t and so. That was a big, just a big part of realizing you know, I want to publish stuff that I’m proud of. And that just trusting my gut on what that means and what that looks like, you know and looking at competitors and like, how can we do it better?
And so really just using, I mean, It was taxing just on, on, not only on the soul of losing so much traffic and revenue, but taxing on the brain on just going through all these details. And yeah, just looking at through that lens of, am I proud of his work was a big part of it.
Jared: Yeah. To some degree, the more painstaking the process, probably the more you do really wanna consider getting help.
You know, because if it had just been you for a year now doing this mm-hmm. You know, I mean, you might not have made it. You know, the human nature is, you know, you, you might stumble along the way, but getting help kind of helps that process get done because until you get it done or until you get enough of the articles done, you know, you’re probably not gonna see that recovery.
Tony: Yeah. Yep. And there was just so much that I learned through that process. I just, I challenge anyone to go through and read one of the articles word for word. Maybe take it one paragraph at a time, put it through the Google NLP demo tool and I, I, and do it through for several articles. There was just something that, it just started to change my mindset and how I was seeing the content.
And that just, just comes through practice and repetition.
Jared: I wish I didn’t have two meetings right after our recorded, because I would be doing that right away maybe tonight. There you go. So going back to, to some of the things that you shared in your tweet that you updated four or five of them certainly seem to have a a Google e a t.
Or e e a t Now that switched halfway through your update process. But you know, like just going through some of the things that you mentioned you created a contributor’s page. You created an ask the expert section. You paid an expert to answer harrow questions. You paid an expert to answer questions on Quora and read it on behalf of the site.
Let’s talk about some of these things. They, they kind of point back to this e E A T concept and, and kind of how you analyzed and, and how you went about doing these
Tony: things. Yeah, so the site has primary persona for it. It’s, it’s not me. My face and name isn’t on it. You know, this is a site that I think we’ll be going for a long time and relatively low-key person.
And so don’t really put myself out there. But, you know I felt it was important to work on the authorship of. That persona. And so being able to, to get the name out there beyond just our site, and that’s something I’ve learned along the way of just doing this for 18 years is like I’ve hired writers.
I’ve put their names and their pictures on, on the site before, but they come and go and what if I need to go back and update that content? And they’re not available. To update it for me. So then what do I do? Do I like have to scrap the whole article or do I have a new one come in and just put their, and now put their name on the article?
Well, if there’s a lot of weight to that person and that authorship and their credibility on the web, then I just lose it if I take down their name. And and so that’s where I’ve decided that I’m gonna have a persona that can la last forever to, to a degree or at least for a long time, for several decades.
And so everything, you know, essentially is, is ghost written. But it’s, they’re written by, by experts there. So yeah, but I think, but, you know, setting up the profiles on, on Cora and Reddit under this author’s name and Payne and Expert to answer those owner answered questions, questions There was a lot of good questions that stuff that, and those, those sites that I just haven’t seen before.
Cause I’m so focused on the keyword tools, right? And, and, and, but it’s a little different coming from, from people. And that’s actually one of the things that my to-do lists for this next year, is that have our own form submission people to, to ask their own questions. And we had to ask the expert section, but those were just questions that I got from like, people also ask or keyboard tools like sim rush But really I, I, I think questions are getting longer, more nuanced over time.
For a lot of niches. And so I think we’re gonna, we’ll start to see those when we just open up the doors and have users reach out to us directly with their questions. And so, yeah, you know, finding experts, I was able to find ’em on Upwork and, you know, I’m gonna niche where, you know, they’re blue collar experts.
They need to be licensed for a lot of what they do. And so that’s where I just can’t find anyone off, off the street to who, who, who claims to be an expert. And so there is a level of information that they’re going to share. That just comes from you know, that educational background that they had to go through.
Yeah. Does that, does that answer That one was there, were there some more?
Jared: No, that’s great. Yeah, no, that’s really great. I you know, a lot of people like you said, that they’ll, whether they put their name or not on it they don’t necessarily go to that nuanced of a level to make sure that Google has a, mm-hmm.
We’ll call it a trail of breadcrumbs, I guess, about this person, fictitious or not on the web. Mm-hmm. So that was a
Tony: key part to that as well, is. Then I updated the schema on every article. For, I have a, I have an author’s schema and it shows author’s name. It shows it has a link to their bio on the site, has a link to their LinkedIn, but also it has, it links to their Quora profile and links back to their Reddit profile.
So Google can make that connection that this is the same. Same.
Jared: Good. Oh, very good. Yeah. We had cow roof on a while back talking about, Hey, you know, you gotta have all this stuff dialed in. You can’t. Yeah. Yeah. Give it to the chance that, you know, Google hopefully connects all the dots for
Tony: them. Yeah, yeah.
You really gotta feed them to, to help, to remove the ambiguity to it, you know to really make things extremely clear. And so there’s no question. And so, yeah, I went crazy with, with schema over last year. We probably have three times the amount of schema on every article than any of our competitors.
Even when we have you know, contributors, like we interview a lot of. Experts for our articles. And so we feature their, their name in, in there and we link to, you know, like their LinkedIn profile. But we also do that with Schema. There’s a co-author. Schema field that you can fill in. And so we put it in their information as well, just again, to make that, that connection of course.
And we ask that they share the article from their end, and so they can, they can have that connection back from like, say like LinkedIn or social media platform. But yeah, I went crazy with, with Schema. I learned a lot from Kyle Roof as well. So he was a great, great source for that
Jared: information. On the schema note, I mean, not to get too into the weeds, but were you adding more types of schema to your articles or, or maybe and were you filling out the schema more in depth?
You know, because there’s a lot that say a generic SEO plugin will pull in automatically, but it mm-hmm. Really often doesn’t go into depth in terms of all the different fields you could add to the schema.
Tony: Yeah. You could really go crazy with it. Another important one that we started to add to it was entities.
So there’s a cool tool called Text razor. Text razor.com. And you can paste in an article and it will show you all the entities that it finds, and it will give you at least the Google ID for that entity and Google’s entity database, as well as a link to the Wiki data. And so we would figure out for each of our article, like what were the top two or three entities for it, and we would add that entity data into Schema.
We would add the link to the Wikipedia for it if there was one to the Wiki data. And we’d also put it in the Google entity. I. ID for that. And so that, and is all based upon. The WordPress tags that we use. So we basically set up WordPress tags for all these entities and we just went through and started tagging all of our articles and then we’ve got custom code in there that then is able to figure out, alright, this, it’s tagged with this entity, so show this entity code and the schema.
Jared: Feels like the old days of actually telling Google what your keywords.
Tony: Remember that. Yeah, for sure. That’s right. There was that keywords meta. Yeah,
Jared: that I was, well, I started SEO right when that was actually a thing, you know, and it actually had input. That’s great. Thanks for, thanks for going to detail that.
That’s, I’m gonna, another thing to add to my list to check out when I get done with my meeting later
Tony: today. There you go.
Jared: We covered most of the things that you did from a broad perspective that you outlined. Any I’m missing though, you know, I mean, I’m looking over my list here and there’s a couple on there that.
You know, we could go into detail but basically could get swept up with a lot of what we’ve already talked about. Mm-hmm. But any big kind of high level things that really important you think to the update process that we haven’t tackled? In the interview
Tony: I. Yeah, so I went crazy with our internal links.
I said, let’s link liberally. I was inspired by reading SEO like help documentation on Google’s website. And I noticed in the article that it linked to another article twice within that single article. And I was like, well, that’s a, a little redundant, you know, and it’s, and it’s. Has, there’s a perspective, an SEO of like, you know, link flow and link juice and you don’t want to dilute your links and whatnot.
But as a user, I found it helpful that Google link to the same article twice. And so I was like, you know, if it’s helpful for our users, like let’s go ahead and add an internal link to a related article. If we mention it a couple times an article link to it every time, I mean, unless is a, they’re like within the same paragraph or or two, we won’t.
But if there’s enough distance, then honestly I think the most helpful thing for the users provide that link for them if they’re, maybe they didn’t read that first, that first paragraph that contained the internal link, but they read that fifth one. And so they’re gonna click on that. And then, so I set a goal for my team.
Like every paragraph should have an internal link. We’re not there yet, but I was just wanting to, to show them like I’m, let’s take links internal, internal links seriously. Because they’re helpful for users at the end of the day. And so that’s something that we, we were also going through as we were going through article by article.
It’s just flipping through and trying to find an option for where we can place an internal link.
Jared: That is really interesting. Yeah, well we see that modeled with Wikipedia, that’s for sure. I don’t know if they do duplicate internal links, but they add plenty of internal links. They make us all look like in comparison.
Tony: Yeah, for sure.
Jared: Hey, I wanna ask you before time comes up here and it’s a little little bit of an aside, but Google Discover. You had a site that was getting quite a bit of traffic on Google Discovery. We don’t talk about Discover very much here, but I’d love your insights onto you know, how to get content into Discover.
I, I realize that’s like a whole podcast episode, but just from a high level, like, like what caused you to get removed from Discover Traffic and then what you did to get that back. Like what are kind of some of the how-to steps to kind of regain or get that discover traffic back?
Tony: Yeah, you know, Google Discover is somewhat of a black box.
You know, we just popped up in it one day a couple years ago and we rode that wave and then yeah, the core update hit and we were completely removed from Discover. It’s actually on my agenda to do another big Twitter thread on my experience with Google Discover. You know, from a high level I think it’s, part of it is just randomness of figuring out.
You know, the random articles that do show up, pay attention to those. I’m doing a lot of an, I’ve done a lot of analyzing on those particular topics. Like what were they exactly? What was the headline exactly? What was the featured image for it? Exactly. What did that first paragraph say? And something, a pattern, I’m noticing right now getting back into Google Discover is that There are related articles that are starting to pop up and, and Google Discover.
So if they say one, say, say we have an article on dog toys for toy golden doodles. Then maybe in a, in a couple weeks we might see one on like the best treats for toy golden doodles pop up. Even though we’ve got a whole site on different animal and dog breeds. But something about the, the toy golden doodles.
Is what popped up. And so it, there’s, I’m seeing a pattern of just, of similar topics. And so I’ve, I, one of the tests I recently did is that I noticed, okay, here’s a, here’s a topic that popped in Google Discover. What’s a, what’s a close, what’s a closest related one that we have to this article? And I took that and I changed the url.
To it. And I reset the publish date, and so it was like a brand new article and published that. And like within a couple days it popped up in Google Discover because Google’s figured out that, you know, the, the people who originally saw the first one about the, the dog toys, well it’s the same, same breed, but now it’s about treats they might be interested in, in this article as well.
And so if there’s, so just following that pattern of the topics and what’s currently working right now is something that I found is working
Jared: well, when you publish that thread, we might just have to do a part two on yeah. I’m
Tony: happy to come back on. Let’s do it.
Jared: Yeah. Hey, so remind us again as we close out, you know, this site was getting about if my notes are correct, about two and a half million page page views or, or sessions a month prior to the May, 2022 update.
Post recovery after Google’s co March, 2023 core update hit you, re you experienced recovery. Where are you at right now in terms of, you know, roughly how many pages for a month you’re at now?
Tony: Yeah, so we’re getting close to doing 8 million page reviews a month at this point off of so not just
Jared: recovery, but massive growth.
Tony: Yeah. Cuz e even the the review update that happened not too long after the core update, we saw another bump in that even though we’ve got maybe two or three articles that review products we, we saw an update with that. And so I, I think that comes down to We, we have experts who are writing from experience, from personal experience as well.
And that’s a big part of what Google’s looking for, is people who have personal experience as well as ex expert expertise to it. And so that was, that was just part of our, our newer continent strategy. And so I’m not surprised that we were rewarded for that. So I think it was more than just.
Categorizing looking at products, it’s, it’s what are people what, what is it that they are looking for? They’re looking for an answer. They’re going to Google. They’ve got a problem, they’re looking for an answer. And so Google wants to show one that’s that’s helpful and that people have experience with.
And so whether that’s the answer is a product or the answer is information, it’s kind of all the same. In many ways.
Jared: Yeah. Yeah. It’s so many examples of sites that don’t review products going up or down or pages on a site that don’t review products going up or down. So something to it. Something to it.
Yeah. You’ve gotta be on cloud nine. Congratulations. Yeah, sure. The recovery and the growth. I mean, in many ways this process sent you down a very. Very troublesome path of, of having to get your hands and you know, really dirty to fix things, right, as the, as the analogy goes. But I mean, not only did you recover that traffic, but the growth is phenomenal.
So huge. Congrats for sticking to it, for putting together that quite a process and for sharing that here. Thank you for sharing that with us. Yes. I think yeah, like I said, the outset, super tactical. If you were listening, you probably have five or six pages of notes, but not only that, just the inspiration that, hey, you know, a site can get.
Really, really nailed in an update. And recovery can happen. You just got to sit down and and buckle
Tony: down. Yep. Put in effort. Stay committed. Surround yourself with others who can hold you accountable.
Jared: Yeah, that’s right. The accountability. So I have teased your Twitter thread, which will include the, in the show notes, but share with us how people can follow along with, with what you’re doing.
Twitter, maybe beyond Twitter, just anywhere people can get in touch if they need to. Yeah.
Tony: Twitter’s the best way. Tony and t Hill, it’s my Twitter handle. Yep. Perfect. I think we can all
Jared: remember that. Yeah, pretty simple. Hey Tony, thanks so much for coming on board. Really appreciate you sharing everything today.
Tony: Yeah, my pleasure. Thanks Jar.