Aleyda Solis, Cindy Krum, Ethan Smith, and Kevin Indig joined Bernard of Clearscope to share their thoughts on Google’s “Helpful Content” update.
Here are our biggest takeaways from roundtable:
Topical authority is only getting more important
Create useful unique content to stand out from the “copy cats”
Each of the panelists share what changes in strategies they’ve made (if any) and how they’re guiding their writers
Watch the full webinar
About Aleyda Solis:
Aleyda is an International SEO Consultant and founder of Orainti, a highly specialized, boutique SEO consultancy.
She’s a popular blogger, speaker, and author of “SEO, Las Claves Esenciales”. Her writings have been published in Search Engine Land, Search Engine Journal, and Moz.
She is also the host of the Youtube video series “Crawling Mondays” where she shares actionable SEO tips.
Follow Aleyda on Twitter: https://twitter.com/aleyda
About Cindy Krum:
Cindy Krum is the CEO of MobileMoxie and author of Mobile Marketing - Finding Your Customers No Matter Where They Are. She speaks and trains people around the world in mobile marketing and mobile SEO, and is credited at being able to accurately predict changes that Google will make, before they are made.
Follow Cindy on Twitter: https://twitter.com/Suzzicks
About Ethan Smith:
Ethan is the CEO of Graphite, a premium growth agency. Graphite focuses on SEO and content strategy, that builds scalable growth engines for companies like MasterClass, Robinhood, and Netflix.
Connect with Ethan on LinkedIn: https://www.linkedin.com/in/ethanls
About Kevin Indig:
Kevin is a strategic Growth Advisor, creator of the Growth Memo newsletter, and host of the Tech Bound podcast. He ran SEO organizations for companies like Shopify, G2, and Atlassian, consulted for big brands like Ramp, Eventbrite, or Finder, and is an active angel investor.
Follow Kevin on Twitter: https://twitter.com/Kevin_Indig
About Bernard Huang:
Bernard is the co-founder of Clearscope, the leading SEO optimization software for high-quality content teams. Before Clearscope, Bernard started an SEO consulting agency, was a growth advisor in residence at 500 Startups, and lead growth at a YC startup called 42Floors.
Follow Bernard on Twitter: https://twitter.com/bernardjhuang
Read the transcript
Bernard:
Hello. Oh, welcome everybody. People are just trickling in. I guess I'll start with a brief little cookie cutter thing. What's everybody's favorite book that they've read so far? Keep it to one to two sentences real quick as we have people trickle in. I'll start to say that I've been reading this fantasy novel series called Clearscope Archives. Really nice to finally take a break from reading so much productivity and business stuff. Kevin? Oh, Aleyda. Go ahead.
Aleyda:
I started actually reading House of Dragon the book, after I watched episode one because I realized that it was good and it was a proper book behind the story. But oh my God, 700 pages. I'm like, "It will take a while. Let's see if I finish the book before the series actually ends in a few years."
Cindy:
I'll say all my reading is business oriented or very serious. I was trying to think of something that was lighthearted and I can't come up with anything. But the Hostage at the Table is a business negotiation book that's very good.
Bernard:
Oh, interesting. Kevin, Ethan, any books?
Ethan:
My most recent favorite book is of course Product-Led SEO by Eli Schwartz.
Bernard:
Of course.
Ethan:
But my other favorite book is 15 Commitments of Conscious Leadership, which is this book that's actually very good about different patterns in the workplace, like don't gossip and how to give difficult feedback and things like that. It's super useful. That's my favorite recent book. And then Product-Led SEO.
Kevin:
Yeah. Ethan, Double on 15 Commitments of Conscious Leadership, probably one of the books that changed my life the most. And the second one that also a profound impact on me is Principles by Ray Dalio. Classic. If you haven't read it, then you should.
Bernard:
Awesome. Cool. We have over 125 people here now joining us for the round table. Want to just give a great thanks to Aleyda, Cindy, Ethan, Kevin for joining us. We want to discuss everything that's been going on with Google's most recent... Oh, I'll go update. Definitely a lot of chatter that people have been speculating on, and we figured why not get some of the best and brightest minds in SEO together to think about and discuss what's been going on with the helpful content update and also forecasting what this implies for SEO and content moving forward.
A brief bit of introductions, we have Aleyda. Aleyda is an international SEO consultant and founder of Orainti, a highly specialized boutique SEO consultancy. She's a popular blogger, speaker, author of SEO Las Claves Esenciales. Her writings have been published in Search Engine Land, Search Engine Journal, and Moz.
We next have Cindy. Cindy is the CEO of Mobile Moxi and also an author of Mobile Marketing, Finding Your Customers No matter Where They Are. She speaks and trains people around the world in mobile marketing and mobile SEO.
Next we have Ethan. Ethan's, the CEO of Graphite, a premium growth agency. Graphite focuses on SEO content strategy that builds scalable growth engines for companies like Masterclass, Robin Hood and Netflix.
Lastly, we have Kevin. Kevin is a strategic growth advisor and creator of the Growth Memo Newsletter and host of the Tech Bound podcast. He's run SEO at organizations like Shopify, G2, Atlassian, and consulted for brands like Ramp, Eventbrite, or Finder. He's also an active angel investor.
Brief bit about myself, I'm the co-founder of Clearscope, the leading SEO content optimization software for quality content teams. Before Clearscope, I was doing SEO Consulting where I did companies like DoorDash, Strava, Compass. And I did a lot of growth advising at 500 startups. Round of applause for everybody here.
To kick us off, I wanted to ask Kevin and Aleyda to get their initial thoughts on what the Google helpful content is, and when it initially rolled out, what were your first impressions? You had a lot of things that y'all put together in terms of the resources and your opinions that got shared pretty widely in the community. Just to start us off, what your initial impressions were?
Aleyda:
My initial impression is that this update could have been call the authenticity or originality update, rather than helpful update. And I have to say, I actually launched this poll over Twitter to see if people were seeing any impact. And most of the answers, and I got 629 votes, most of the answers, 58% of people didn't see any impact. Then 16% saw positive impact. 9% only, mostly negative.
And it makes sense because when you go and take a look and see which are the websites that tend to have seen an important meaningful impact, these are all scrape aggregated, non original content type of websites. For example, manual, dictionary, lyrics type of websites, there are tons. If you go to C street movers and shakers and go to those websites that have drop the most in this time interval. These are very spammy type of websites or not necessarily real valuable businesses or businesses that have a real business model or real product or service beyond ads. It makes sense that all of us who are SEOs haven't seen an impact on our clients, that tend to have an actual product or surveys or value behind in it.
Kevin:
Yeah, I want to double that. And to just take a step back, I think one context of this update fits into is that there's more and more content on the web. It's super easy to create content. We haven't even spoken about AI content. I'm going to not dive into that for now. Maybe we can later. But in general, it's just super easy to publish. And the challenge that Google is in front of is that it's getting harder to crawl the web.
One thing that wasn't mentioned in the context of this update is that every August for the last couple of years we see indexing issues on Google side and infrastructure problems. There's a challenge of just reigning in all this content. And then of course ranking it is also more and more challenging. For example, one of the types of sites that seem to have been impacted the hardest are lyric sites.
And a lot of these sites are just copy pasting lyrics. There's no additional value to the user. Is pretty much duplicated content. And the sites that actually one, like genius.com and azlyrics.com, they all provide an additional value. Like in genius.com, you can comment and annotate lyrics, which is an additional value.
Google just sets the bar higher and higher of what acceptable content looks like. And this is another example of an update that does that, that facilitates that at scale. Now one thing before I pass it on that I want to highlight real quick is that I think we all, and I certainly contributed to that, I think we all expected more regular content to be affected by this update and what the very early data shows, it's that it's really the bottom of the barrel type of contents. Almost spammy, no value, goes after some high search volume, get a quick answer. I saw a website for example, that teaches math and they go after queries like "What's 10% of 73?" There's super basic type of stuff. And that's I think what Google means with content that's credits for search engines rather than users. I'm going to stop there.
Bernard:
Ethan, Cindy, anything to add?
Cindy:
Total agreement. I think that may be that we'll really get a stronger sense in the next couple of weeks. It does seem like this was a plan, that they launched the helpful content update and then the September update. It may be that helpful content added some kind of new evaluation or some kind of new waiting and then the core update expanded that out or applied that in a wider way. Yeah, I think it's TBD, but definitely uniqueness. Does uniqueness and value add are being given more weight, however that plays out?
Ethan:
Yeah, the way that I think about it is, social media is filled with fake information and Google is filled with useless information. They're generally trying to solve this problem of useless information. And I think I inferred five themes from the update. The first two already existed. Theme one is "Did you find your answer? Did you go back and get it somewhere else?" That's the back to Google update is a very bad signal and finishing on your particular URL is a very good signal. That's been around and they're reiterating that.
The second one is "Are you making content for Google? Are you making it for users?" Which is also been around for a while. They're reiterating that. There's three new themes that I see. The first theme that I think is very interesting is topical authority, which is something that we've been thinking a lot about at our company.
And that means that rather than domain authority, which is what's your authority overall, rather what is your authority for a particular topical area? For MasterClass early on, MasterClass had authority for the names of the celebrities like Gordon Ramsey, but not for the food category overall. We would focus on things like Beef Wellington. Topical authority is a great way for sites that are smaller or newer to rank for things.
Google repeated in several of their points, "Are you known for a particular thing?" And the inputs for that are things like branded search, a masterclass and Gordon Ramsey, or when somebody types in Beef Wellington and they landed on your site, do they have good engagement? And then you'll basically build topical authority that way. That's about one new theme.
The second new theme is rewritten content, which I think is probably the biggest problem around useless content, which is everyone's rewriting each other's articles. And they do that because it's cheaper, it costs more money to write a brand new article and to have somebody with expertise. You just say, "Hey, go rewrite the first result," and then somebody else rewrites that result and everyone's rewriting each other's content. And then you have basically thousands of the variations of the same thing. In the last-
Cindy:
Oh, sorry. It's not just cheaper. It takes original thought and that's hard. And I think Google wants make things a little bit more fair where if you have original thought that's valuable, that you get rewarded with traffic even if you're not a great SEO or even if you're not paying a content farm or whatever.
Aleyda:
I think that actually Google did a good work commenting the type of criteria they will take into consideration, because in the original dot that they publish, they actually mention they will be able to identify if the content seems to have little value, low added value or is otherwise no particularly helpful to those queries, for the searches. They actually have a very straightforward system, let's say.
And the low added value or little value, especially for queries, as Kevin mentioned before, we can see broader queries, ambiguous queries that can have a diverse type of intent. For example, I see queries, for example, white cutters or toy guns, for which aggregators of reviews, were ranking before in the top 10, not anymore. I see more whether actual best type of list well written and research and backed, and then commercially driven, category, pages that are rich in their content, ranking now instead.
I do believe that Google has done a really good job in this case, specifying that this automatically generated content or that you pretty much use whatever tool out there, or very cheap writer without not knowledge at all in order to be able to rank for this broad term or query like, yeah, you're not going to be able to do that anymore that easily.
Bernard:
That's here in agreement that... And this is one of the questions that have already popped up. Is that the helpful content update is like panda version one, where it's very clear that certain sites, lyric ones being the example that we're all familiar with, which are all more or less copy and paste examples from one another, are taking great hits.
And I think we're all trying to talk right now about this idea of original, relevant, high quality content, where the content itself has information gain or some people call it originality, uniqueness, a special point of view. I want to ask them, to make this kind of a more actionable point, what does that information gain or unique and relevant content mean to each of you? And how would you really start to think, okay, if helpful content, at least in its current form, and it's very likely Google continue to build on top of that, how are you thinking about adding information gain where you're not just simply just rewriting what is being written?
And also similarly, you can imagine information gain is a potential way that Google might think about combating AI content, where AI content, at least with the current models, is simply regurgitating a lot of what was already written as well. How are you thinking about information gain and how are you thinking about creating unique relevant content moving forward? What are your thoughts there?
Kevin:
Can we pass it to Ethan? I want to hear what the rest of his thoughts were about the previous question and then maybe you can answer the next one too.
Ethan:
Yeah, well the last thought is just around, I would say a misapplication of SEO tools, like SEO tools will say "In general, the word count is this." Or "In general people are talking about these things," which is useful. And then you would say, "Okay, well so in general, maybe people want this particular length or these are themes that people are interested in." And a misapplication is, "Okay, well I have to have 1,057 words and I have to include the word lemon six times."
I think that's a misapplication of tools. I think that that was my last theme. And then, Bernard, your question of how we're thinking about... I think the main thing is that the reason why there's rewritten content and regurgitated content, number one is because it words. Number two is because it's expensive to have premium unique content. I think you need to change the business model.
And the reason why you do that is because if you write a hundred articles, five of them are going to work. And you don't know which five, so you have to spend a small amount of money on the 100. If you can be more thoughtful about selecting those five that are going to work and spending more money on those, and then Google changes their system to incentivize uniqueness, then you can justify spending that money.
We spend a lot of time just being really thoughtful about picking those five so that we can spend that money. And then, on our end, we basically hire domain experts who are journalists. If we're working in finance, we would hire a journalist who's written in finance, who has some domain expertise. Now we have a business model to write really great content for these five niche topics. That's how I think about it.
Cindy:
I'll add in a unique thought, at least I think. Is that it does seem very much like in this update and a couple of the previous ones, Google is directly going after affiliates and aggregator sites. And that is something that they've done for a long time, but there's a unique reason why they would be doing that now. Because you see they're going after the affiliate sites on this side of the page, and then on this side of the page, they are becoming more and more of an affiliate site themselves.
Google pushing everyone hard towards merchant center and of course flights and hotels and rumors of car stuff and insurance stuff, and all the other things where Google wants to be an aggregator, it doesn't want competition either. There's possibly some double dealing.
Kevin:
Yeah, I do agree with that very much. Because one type of sites... We spoke about content that doesn't add a lot of value, but there's also of content that just answers very shallow questions. And that's something Google does already themselves. They do it pretty well and they don't need sites to do that necessarily. I do agree that there's an angle of... Maybe the primary intention is to just say, "What else can we answer ourselves?" But of course that has an impact on stuff like affiliate revenues, data that companies can gather, even from shallow questions and users that they can track.
Cindy:
Well. And the other thing, Kevin, and this is something that I've been talking about a lot, and I say double dealing like it's bad and maybe it's bad, but maybe Google does, I think, have the best interest of the user in mind. But when they can pull in a feed or host something in their cloud, they can know that it's live and good and up and they can serve it more quickly and they can transcribe it, transcode it into audio or whatever. And I think that that's where they want to go with the future of connected devices, internet of things, multimodal search. When they host the info, they can understand it faster, better and serve it faster, better.
Aleyda:
Yeah. I agree the two things. On one hand, raising the bar for the position zero type of queries that they were looking to answer anyway and keep the traffic to them. And then on the other hand, the money making ones. At the end of the day, they're also shifting to become an eCommerce type of portal too, competing with Amazon. And we just saw earlier today, like an hour ago, yet another new release from them facilitating for anybody pretty much just using structured data, to be able to make the most out of the product carousels. And a lot of the updates that they have been doing also in the affiliate reviews since last year or so, have been very focused on that, that those actually focus on those queries, have original content, originally review this product.
Originality and authenticity here, and getting the bar higher in order to grab this traffic, there's certainly pointing to that. Then on the other hand, definitely the way to go will be to hire experts to write the content, but if we think about it's not that easy. The real expert will be actually doing that on a day to day, rather than necessarily only writing about it.
For example, I get emails from time to time asking me to write something and they would pay to me. No, thank you. I'd rather focus on doing actual SEO with my high paying clients and brands, and then whenever I want, I will share my insights about that experience rather than being hired as a writer.
However, I think that a good approach that we can see in our own sector is the one that, for example, Content King has with their research, they have a very well developed content hub. They call it SEO Academy. And what they have done is to cover every area or topic, the major ones in SEO nowadays, and they have their own SEO writing the content. And then this SEO will ask for input of people specialized in that area so they can actually have this added value of real experts.
And the effort for the expert is actually minimal, because it's just a little quote, or an idea, or an experience in that particular area. And then of course you incentivize the person with a link back. The person will also share with their own audience and the community. And usually these are influencers too. There's a win-win type of approach. Of course, it's very easy when the topic is SEO and we are a willing, and are very knowable of the value of participating. But I do believe that there are many different ways nowadays like with Harrow or all the similar services out there to get input from experts too.
Kevin:
Another question or dimension of thought that I want to add is that information gain looks very different based on the keyword or at least user intent. If one piece of information gain could just simply be looking for the results of a football game or a soccer game for the fellow Europeans. And the key piece of information is the score of course, but maybe some information gain is the different statistics that a team got and I don't know, in their steps or whatever.
But it's very different. From an eCommerce perspective, there's certainly the price, but then there's also stuff how does it fit? How does it fit for a certain body type? There's all sorts of secondary, tertiary information that companies don't think about to quantify what information gain really means. And the most fuzzy aspect is when it comes to very complex topic that's covered mostly with text.
When you maybe write about why, I don't know, Napoleon failed in Russia or whatever, make something up. Something that needs a lot of explanation, that's where in my mind, information gain is the hardest and it's the easiest when it just comes to providing users data that's very useful.
Bernard:
Question here from Amelia who asks, "Does Google take into account the expertise of the writer or just the originality of the content itself?" I'm going to add a follow up question to this. How does this all factor into expertise, authority, and trustworthiness and what you've seen a lot of websites do, where it's medically reviewed by X or trying to infuse the content to make it look more trustworthy? Are you all thoughts on this?
Ethan:
Oh yeah, I've never actually seen an impact from having a specific expert listed as the source of the information. I've only seen writing content around where your topical authority is. You'll rank better at that. The MasterClass example, you’ll rank the Beef Wellington because you have to have topical authority for Gordon Ramsey, but I've seen no impact from listing Gordon Ramsey as author versus MasterClass editorial team versus no author. That's my data.
Cindy:
I was just going to say that I think that that is what Google is trying to do, but it's hard because, like Aleyda was saying, the topical authorities are busy being topical authorities and aren't necessarily also being authors. There are going to be people who write that finding up whether they're the expert or not. It's a tough thing to do algorithmically.
Aleyda:
I do believe that is very hard to isolate, because for example Lidy, she's very focused on EAT, she has shared quite a few examples in her experience where she mentions how doing a set of improve improvements on the articles among them listing a clear author and having... But these are usually also your money and your life type of topics where we know the sensitiveness is a little bit higher than other type of topics.
And also, at the end of the day, it's also very hard to, let's say, at a granular level, isolate what is that the authority of the author, the authority of the publication itself in general. Things like the authority of the back links pointing to the piece, things like that. However, I have to say that the clear seniors you can provide being an authority in that field that you're writing, the better will be one of the things that we need to keep in mind.
Also, I believe that Google has done a really good job in this case and the commanding update. One of the questions that they actually added their focus on, people first content in that section is like, are you keeping in mind our guidance for core updates on for products review? They all overlay and at the end of the day, the intent and the goal of Google is pretty much the same.
They want high value content that is expert written, that is actually comprehensive, that actually adds value, et cetera. I actually aggregated all of the questions in Google sheet that I shared around to make it easier for us. Because the difficulty that I can see from this is many, many times explaining to copywriters in a way for them to double check and validate beyond the tools. Because indeed, I think it was Ethan who mentioned before, if we start adding the recommended number of keywords and the length in the piece, well we can always work around and just have a listicle of keywords and that's it. At the end of the day and it will pass the ratings or we'll have good ratings that will pass the validation.
But we need to think about, is this really comprehensive? Will answer in a better way the need of the user searching for this versus all these other players that are already ranking for this query. At the end of the day, this is very context focus. What I actually will think about is, if there are many articles already ranking the top position, all clearly being written by doctors, but very visible credentials, I want to ensure that I have that too. I don't want to be missing that for sure.
Kevin:
That's a good point and I really wonder how close Beef Wellington is to... What's his name? The chef. Help me out real quick.
Ethan:
Gordon Ramsey.
Kevin:
Thank you. Gordon Ramsey, on Google its entity graph, I'm sure they're not too far away. This is my take on authorship and how important it's to have authors. I think that Google can understand the connection between entities and if an author is an entity in Google's knowledge graph, then they can probably understand that. But I do think the bar is incredibly high to be recognized as an author or an important person in the eyes of Google. There's probably some credentials or some publications or some media attention that you have to have for that.
Two examples I want to have really quick to make this practical is there are two sites that do this that highlight authors and they do really well. One is Healthline. Very, very successful side in the medical space and the other one is Trucks Advisor, which has gotten a lot of attention from SEO recently because they rank for best truck tires or whatever. They've been crushing in general lately.
And I don't think that they've been crushing it because they highlight the author that wrote the article, but because they, first of all, go to links to show how often the article was reviewed and by who it was reviewed. And those are typically subject matter experts. By the way, Chewy does it... Well, they have, I think it's a pet disease content hub or something that has content reviewed by veterinarians.
And third, and this is I think the most important piece, is just that the content is better because you believe that it was reviewed by experts. Lets say that Google says, "Oh, it was reviewed three times and oh, these people have a doctorate title." But I think it's much more that the facts are correct and the information gain and depth of information is pretty high on these articles. That would be my take.
Bernard:
Yeah, I agree with that. I think that obviously as SEOs, a lot of people are trying to understand the ways that the algorithm works and look to expedite gaming the potential systems that the algorithm is looking for. And I think a lot of that time, what SEOs might end up doing is chasing the wrong signal.
And I think in this particular case, this is one of those ideas where you see a lot of top websites, whether it be Chewy, Healthline, where they do these medically reviewed by and people see that and say, "Wow, that must be the rationale of why this page is doing so well," and they look to basically replicate that experience. But I think to build off of what you were saying, Kevin, it's really that because it's been reviewed by a lot of experts, there is a lot of good evidence backed suggestions and statements that are being made.
And oftentimes people just want to eat the cake and not really do the work. And I think that's what you see a lot of people then chasing the wrong signal and saying, "Okay, well, I need that." But really a lot of the best articles, they don't necessarily need that reviewed by this expert. They just need to have that right structure and that right brand, and that right user experience where, when I land on an affiliate site, I want to know why I can trust you and what was your evaluation criteria that you took to make these decisions.
I want to see photos of you actually using each one of the different products and really showing me that you did a good job in this evaluation. And a lot of people in affiliates just slap on a, "Well, Kevin Indig reviewed this," and they're like, "That should be good." And it's like, no, no, you got to do the work. And I think this update is Google's way of currently putting some PR behind what they want to influence us as content creators.
And I think it's very similar to what Ethan has stated in his initial thoughts on helpful content. It's very likely that this is version 0.1 and we're going to see Google continue to build on top of this framework that they've had. And I think just lots of good thoughts going around here.
Aleyda:
I think that it is really worthy here to be proactive. If we know that Google intent and goal is to try to replicate human assessment of value and satisfaction, let's try to identify as much as possible all those signals that we use and we certainly want to integrate within our work and effort in order to be best in class.
I don't want to wait until Google confirms that they use X or Y as a signal just for me to start doing it. I know that I want to give the best experience to my clients, to my users in the website. This actually goes most of times well aligned whit UX, and with a lot of other, let's say, good practices in general to provide the best expectations or to fulfill the expectations of users.
I will try, as much as possible, to align my work with the product team, with UX team in order to whatever we can highlight and show the users that we know that will improve the conversions anyway and help to achieve results, I'm all for it.
Cindy:
There's also this concept that's going around with blaming the internet and social media and Google and all the big players for dumbing down content and rewarding the lowest common denominator or just rewarding engagement that is not necessarily with high quality content.
And you see that even with guidelines that come out about people who write complex sentences, "Oh, Google doesn't like complex sentences." And yeah, Google's algorithm has a hard time parsing a complex sentence, so that makes sense. But at the same time, if I'm Google, I don't want to just reward sites that write "Jane likes Beef Wellington, Jane wants to cook Beef Wellington, Jane needs 10 ingredients."
I think that it may be a backing off of that kind of simplicity or a more rewarding content that is a bit more nuanced. And if I were Google, I would want that because I would separate myself from maybe Facebook or things like that. Because you do see, like in Google Discover, some really lowest common denominator content coming in and it's because of just engagement. And if you want to separate yourself from just rewarding that kind of thing, you would do what they're doing now or the way they're communicating it. That's what I would be going for.
Bernard:
I want to circle this back around to a lot of Q&A that's that's been popping up. I know we're dancing around this whole information gain topic, but Justin asks, "Okay, but how is Google going to determine original thought?" Back to this information gain thing, really to say, "Okay, if we're going to put on our SEO like crystal ball hats here, how are you all thinking about practically speaking, demonstrating to Google that you have content that has information gain, since a lot of it could be very subjective and measuring originality is something very difficult, at least potentially difficult for AI to accomplish.
Ethan:
I have thoughts on this. I think the answer that sounds good is good engagement. You found your answer, you did go back to Google, but those have already been signals. I don't think that's the full solution. I think the solution is around uniqueness and useful uniqueness. The way that the algorithm currently works is there's labels that the machine learning algorithm trains for, so these three results are great, so anything that is similar to these three results is also great. Which means everyone's rewriting each other's articles and therefore the system incentivizes everyone to rewrite each other's articles.
Then you want uniqueness, but you want useful uniqueness. If you think about a camera, if you have on a camera and you have the word tangerine and nobody else has the word tangerine, that's unique, but it's not useful uniqueness. But if for that camera, you're the only one that talked about lowlight photography, that's useful uniqueness.
The way that I think that they would do that is they would know that lowlight photography and camera are close to each other in their knowledge graph and tangerine's way over here. My guess is that they're going to be looking for, "Did you mention subtopics and themes that others didn't but it is closely associated within a knowledge graph?" Is my guess.
Kevin:
Yeah, I agree on the knowledge graph. I don't think this is so difficult for Google to be very honest and blunt. I think that one of the big questions in this whole conversation is also what's the stack rank of the signals or how do they relate to each other? And I think in the past, even though it was often denied it was a big impact from back links. I'm not saying that's necessarily going away, but for originality, to make it bigger impact, you can also just turn other signals slightly down or that originality slightly up. But Google should already be able to understand what information a piece of content has, especially after transformers and MoM. This is text mining on steroids if you will. That should be very obvious to Google. And then another signal to pair this with, because not every signal lives in isolation, is the behavior of users.
If you go back, if you start with a very high level topic or a very short query, you read something and you go back and you enter a longer query that is more targeted, Google can understand that perfectly fine. This is how a lot of personalization works. And they can measure very well what that looks like after people have visited a certain piece of content. I think there's several signals that balance each other that help Google to make a right call. And from a purely technical perspective, I don't think that this is something groundbreaking the new, I think the signal that Google has decided to end up, because the content quality is often so bad.
Cindy:
I'll just throw one more thing in and that is the content frequency. Even though this is another thing that Google denies, having an aggressive publishing schedule where you're publishing every day, Google says that doesn't help you. But I think that in some cases it does create some benefits or has created some benefits. But if you're the original thinker, the original creator, the original author, you might not be publishing every day and that's okay.
I tend to see this, and perhaps differently than the other panelists. I see it a little bit as an easing off of some things that have been maybe not direct signals, but secondary signals and rewards in lieu of things that are harder to create with AI or create with a team of regurgitating writers.
Bernard:
Speaking of AI, that's been a topic that's been highly speculative, highly sought after. Let's talk about AI. What are your thoughts about AI content as it relates to SEO, this concept that we're talking about information gain? And I'll start with Kevin. You had a really great blog post recently that talks about basically the pros and cons. It's not a do it or don't do, it's a it depends on the situation. AI is really good for certain classes of content, more specifically definitional things, things where I would say... And this is back to that content uniqueness idea, AI's really good when us as human beings have all decided that X is X. The first president of the United States was George Washington and AI's really good at that. Anyways, people's thoughts about how AI, specifically content created by, stuff like GPT-3 is going to influence SEO content moving forward? We'll start with you Kevin.
Kevin:
Yeah, thanks for setting the stage, Bernard. As you already mentioned, I think we really have to talk about the types of content and what the role of content is. It is not always the product or the thing itself. It can sometimes just be adjacent or supporting the main role of the show or the main piece of content, which in the eCommerce for example, is the product. And then you have a lot of content around that describes the product or gives you helpful secondary information that helps with information gain.
And for things that are not already complex, the AI already does a good job. And this is something that I changed my mind on a lot before I saw more practical examples. I actually thought that when companies have an AI, they all create the same content and then Google just devalues it like they devalue links.
And I don't think that's true. I think what's actually happening is that some companies are figuring out how to do this and do this well for certain types of purposes and certain types of content, and it ranks pretty well. In the light of the helpful content update and previous statements by Google's spokespeople, there was this notion that Google hates AI content.
I paid a lot of attention on sites that lean heavily on AI content to drive organic traffic. And that traffic has only increased during the helpful content update. That definitely hasn't tanked. To me that goes to show that if you do it well and if you do it right, then there's no problem with AI content. Why should it matter who creates the content? It doesn't matter if it's a machine or a human. All that matters is information gain and is it useful.
Aleyda:
Well, I was going to say, is how you use it. Actually, I want to add here an scenario that I see a lot where clients that tend to go through the internationalization process, most of translations tools at this point leverage automatic translators like DeepL, amazing quality already, or Weglot, for example. They integrate into the human translator process or workflow to facilitate their task.
And they can have that as an input. That doesn't mean that the content will be completely translated by AI completely, by the tool, by the translator, but that it will be much easier for the translator to go ahead and do potentially, I don't know, 10 pieces per day rather than five or four. I believe that there's this good sweet spot, as Kevin mentioned, it makes me remember a lot with templated content back in the day.
Like, "Oh, is templated content good or bad?" Oh, it depends on how you actually leverage it, and how smart you are, how creative you are, what are the criteria that you take into consideration to generate it. And do you use stats? Do you use your unique insights? Et cetera, et cetera. Exactly the same is how you use it, how you include it, also what type of content it is in what type of pages. Is this content actually valuable to answer the doubts at that stage of the journey independently if it is 100% original or not?
There's a lot to be save there. I will highly, highly recommend people to be creative about it, to think of how you can leverage it, rather than live it on out pilot, which is definitely not great, how you can use it to accelerate the process of content optimization, content translation, et cetera.
Ethan:
Yeah, just building on that, the way that I think about it is that there needs to be underlying wisdom. Especially if you have structured data, and an example would be a sporting event. You talk about the score and somebody made a basket at this particular point in time, that's all structured data and that can be generated. You can generate sentences based on underlying wisdom, which is that structured data.
We invested in this one company called withpower.com in their extract structure from clinical trial information and generate sentences based on that and that has underlying wisdom. We then also tested generating product descriptions for products. And I remember GPT-3 said that a particular soap was oxide free, but it wasn't oxide free. And if you need it to be oxide free, that would be a big problem.
And the more nuance it is, and the more it's based on a personal experience that Google mentions "Did you use the product?" If you need to use the product to see if it's usable, you don't want GPT-3 telling you whether or not the product is usable. You want a human talking about that. I think it needs to have underlying wisdom and some things have structured data that you can automate. Having underlying wisdom and a lot of stuff does not.
Bernard:
All right, we have quite a few open questions. I'm going to make this more of a rapid fire type thing, throwing this to who I think might be the best fit in answering. But if you have additional thoughts, chime in and let's make this happen. Anonymous asks, "On the topic of expertise, what's stopping organizations from hiring journalists to do journalism? i. e. doing deep research, chasing experts or quotes and context or publishing long explainer features like Vogue's and other journalistic publications." I'm going to kick this over to you Ethan, since you're really deep into the content space right now.
Ethan:
I think it's the business model. That costs money. And again, if you have a hundred articles and five work and you don't know the five, then you cannot afford to pay that writer. If you do know the five, then you can afford to pay the writer. The business model and selecting the topics that are likely to perform is number one.
And then number two is that a lot of times journalists don't want to write content that is search friendly. Suggesting to the journalists, "These are the general subtopics that users are interested in" versus "Hey, please stick these keywords in," guiding that journalist to answer questions that users have but still have journalistic integrity, so the process of working with that journalist, those are the two main things. But I think that both of those can be done if you have a better process.
Bernard:
Awesome. All right. Another anonymous asks. "Think Google should bring back authorship tags to determine the value of writers for determining expertise. If someone who is an expert at the CDC is writing on a new site about virology, should it be promoted more than an intern writing in the editorial column on that site? Should the power be in the author's topical authority expertise or the power to be in the domain for what it's labeled as?" I'm going to kick this over to you, Cindy. What are your thoughts?
Cindy:
Yeah. Number one, I think that the author tag was educating Google about where to look for author information. And now that it's changed and not as necessary, that doesn't mean it goes away. I think Google used the information and now they're just better at finding who is the author and ascertaining information about that author and connecting it to the topic.
I think that every time Google does something like that, it's just asking us to code stuff so that it can train their algorithms to know where on the page and the site to find that information. Authorship still matters. The one thing I'll caution is that I think that we on the internet throw around authorship quite loosely, that anyone can be an author and publish things on the internet. But if we think about separate being an author online and writing stuff that anyone could publish online, versus writing a book or writing something that has to be published by a publishing company, that kind of authorship may carry more cache in Google's mind.
Not just lumping all the different authors together. And yes, I do think that Google has the goal of connecting author's credentials with what ranks where. The example earlier that Aleyda gave from all the PhDs writing about medical stuff, yes you want to have the right credentials and you don't just want staff writer transcribing what the actual expert basically said.
Hiring the experts when you can is always ideal. It's just like what Ethan said, the model doesn't allow just... There are only so many experts and they only have so much time and they're expensive and hard to get. Yeah.
Bernard:
Can you speak to the relationship between creating original content and using keyword or other SEO best practices? Are both still important? I'll tack on like a follow up thing. This idea of subtopic search intent, zero keyword search volume, there's a lot of confusion now around, "Are you targeting keywords, are you going after the long tail?" What are we doing? I'm going to kick this over to you, Aleyda. What are your thoughts on this?
Aleyda:
It depends on the goals of the business at the end of the day and what's their business model. For some type of businesses, it does make sense to develop topical authorities, and at the same time that they are targeting more commercial queries to also develop the content hub for their top topics for example. And indeed, the challenge here many, many times is to, on one hand, show the value of this type of content that will play a role through the customer journey indeed, but won't necessarily generate conversions right away. It won't be at the end of the journey, necessarily.
It's important to a little bit educate and evangelize of the need of, "Okay, if you want to establish yourself in your sector and you want to be thought as a leading brand etcetera, you need to develop this content." And then on the other hand how you talk with writers in a way that you don't alienate them.
I think that we have had a lot of that for a while. And SEOs in general, I'm afraid that in certain sectors with copywriters, developed this reputation. I remember quite a few years ago when I was at this content conference in Minneapolis and I was literally the only SEO in the whole conference that was pretty much focused on content people, they did made a joke at the expense of SEO. "Ah, yes. The SEO comes and tells you 'Add keywords,' hahaha." Oh my god, that was terrible.
Anyway, I think that we are long past that with tools, with content tools. Of course, they provide us which are those terms that are identified to be semantically connected, relevant related. But whenever I start with a new process, I do this, let's say, a little bit of trading and worship with the writers and we go through practical sessions showing "Does this make sense?" Not because all the articles that are ranking the top 10 already are mentioning this.
And there are quite a few times that they mention it. It actually makes sense for you because maybe there's diversity on the nature of the articles are ranked and you are not like these articles necessarily, but these other ones. It's important that there's a good communication, best practices, yes, but always with the final goal. And this is something that I have found that writers appreciate a lot, that at the end of the day they are the ones who are writing, they know about the topic, they are going to research that you can recommend and it's highly advisable whenever it makes sense for the goal of the articles to include certain terms.
But it's not like a red flag to, "Oh you need to add 50 keywords. If you don't add them, it's bad." No, sometimes it doesn't make sense. Sometimes it's not doable because of the role of that particular content or that particular piece. And it's okay. Is important to have this conversation and I do believe that copywriters will highly appreciate you clarifying this very common doubt, that is not back in the days like "Keyword here, yes or yes." And we need to move away from this reputation, indeed.
Bernard:
Couldn't agree more. A couple more. The highly voted ones that Rico asks, "How much of a reaction to TikTok being used more and more to solve queries do you think this update is?" I could not think of a better person than Kevin to answer this because your recent blog post is more or less talking about this. Kevin, your thoughts?
Kevin:
My thoughts. Quick answer. I don't think they're too related to be very straightforward. A signal that would've convinced me otherwise would be to see a lot more short video carousels in the search results or see just a lot more TikTok content in general. But I think it's more like the anti TikTok update, if I really think about it. Because if we would transcribe the majority of content on TikTok and rank that high search results, there would be a lot more conspiracy theory believers and a lot more fake news. I think actually TikTok really needs the helpful content update right now, we're fact checking content. I don't want to hog the mic too much, but short answer is, I don't think they're related and content on TikTok is disastrous.
Bernard:
... mentioned that you write a hundred pieces and five will work. How do you pick the five? Is it based on volume, intent, difficulty to rank? Garrett then has a follow up question. "When you get to a more focused approach, through 10, maybe 20 out of the hundred begin to work."
Ethan:
Two things. It's search volume at the topic level and it's topical authority, versus difficulty. When I say topics, I mean for something like butter lettuce. The MasterClass article ranks for 500 different keywords, and the sum of the volume of those 500 is the volume of that topic. That's the topic volume number one.
And then it's topical authority versus general keyword difficulty, versus domain authority. To the Gordon Ramsey example. If you're known for Gordon Ramsey, then you're going to be very topically authoritative and competitive for Beef Wellington. It's a combination of total volume times probability to rank equals expected visits. And then also ideally, if you have some conversion intent and you can factor that back in and then you can have expected traffic times conversion equals expected conversions.
Bernard:
Okay. To add on that, because topical authority has been thrown around quite a bit in this discussion, how do you know you have topical authority? And where would you go to figure out what those very related subtopics that you have a very high chance of placing for quickly look like?
Ethan:
For the first one, all you need to do is go to Google search console, download all your keywords and then do end grams on them. By that I mean chop them up into one, two word phrases, and that's your topical authority. The inputs to that are branded search, the MasterClass plus Gordon Ramsey. That's like a back link with rich anchor text Gordon Ramsey, back links with rich anchor text. Those are the inputs, shares of popular posts and what the post is about. Those are the inputs. But the short answer to how do you know, literally just download all your keywords and that's the answer to topical authority.
How do you find adjacent topics? That's a little bit harder. That, you need to use keyword tools and use some intuition and stuff like that. There's not really an automated process. It's mostly art and some science. I think Google has their own knowledge graph. Ideally you would just be able to download their knowledge graph, but in lieu of that, you need to make some guesses and use the tuition. Somebody should build that tool though.
Bernard:
Somebody should. All right. I'm going to then conclude with the final question from anonymous. "I'd love to hear what new directives or expectations you've given to content writers on your team, to your clients, to the people that you're working with, other SEOs, everybody." Let's close it out with this. Maybe starting with Cindy, what are some new directives or expectations?
Cindy:
Well. With the helpful content and then September update, we've always assumed that helpful content is something that is going to be impacted by September update and I think it will. But also looking backwards in the launch of pros and cons, I wonder if having nuanced perspectives is also considered helpful. Marking up your pros and cons, that's where I'm going right now, and making sure that you're giving balanced information with pros and cons, that usually seems helpful to me. That's where I'm going.
Aleyda:
Yeah, the main goal that I have already for a while, and I re-state every single time is like, is this the most comprehensive and best in class piece of information that deserve to rank all of the other players that are ranking at the top at the moment? That should be the north start. And whatever it is required, depending on the intent, on the type of the format, of the query and the servs. Let's try to leverage that and make the most out of that.
Many times it will be to ask for testimonial to power users, ask for reviews or integrate reviews even to act. All the times it will be to ask for input from experts, for example. It might different depending on the nature of the content, of the type of content. But I think that as a good approach, it will be, have data is an north star to be the best up of those that are already ranking. Add something extra, add something more on one hand.
And then on other hand, understanding what are the opportunities within the servs, which are your competitors that already ranking the top positions, not only in the organic ones, but also in the video carousel, if there is any. People also ask, pay attention of what is in the servs to leverage and learn and integrate as much as possible in order to perform and to be the best answer out there.
Bernard:
Yeah. No. I guess I'll go next. I think that to keep it short and simple, really think about you're above the fold experience. Tons of people get this wrong. You see it in "Oh, how do I do this?" And then you have an H one that's "What is this?" And I'm like, "Ah, come on. That's not why the users here." Or the concept of hero images. Lots of people just put random hero images in the top spots and it's like, "Well, the user doesn't really care about that." Really just hone in on, "Okay, when a user clicks onto your site, can you give them the most useful relevant information as quickly as possible?" Pass over the mic to Ethan. Thoughts on? Directives?
Ethan:
The main directive is to cover all the subtopics and to be comprehensive. For the butter lettuce example, try to understand what 500 keywords you're actually targeting and what the themes are. Like health benefits, comparing with other kinds of lettuce recipes, and make sure that you cover all of those subtopic themes. Subtopics versus keywords. Don't over optimize for an exact word counter adding these specific keywords. Understand the general themes and subtopics and talk about all of them and fulfill the intent of the whole keyword cluster of the topic. But in general, we don't suggest to writers anything at all about SEO. It's more about this is the outline than these are the subtopics and themes and questions that users have, write a really good article.
Kevin:
That's a good question. I haven't changed any guidance so far at all, because the data is still not out there. The September update is still running. We still don't know what's punished and what's changed. Imagine just on Google's announcement that would've given guidance to companies to go to extra lengths, invest more budget, do all the stuff, and then we see just some, like lyrics side being penalized. It already suck. I'm really cautious. But the one thing that weighs on my mind a little bit is, if it comes to stuff like content briefs, to add a question along the lines of how is different than everything else? What's the unique thing about this piece of content or what's the unique insight that people take away that they cannot get anywhere else?
Bernard:
Kevin, thank you so, so much for joining us. I got a kick out of just hearing all of you all's opinions, thoughts, perspectives, and I'm sure that our audience did too. There were so many good questions that were asked. So much good discussion. We definitely need to do this again, make this happen again. We'd love to have you all on for an encore, whether that's when Google announces random stuff or just regular programming. We would love to have you here. Thanks so much for joining us.
