Hello! We had a lot of discussions on the helpful content system and also E-E-A-T this week. Gemini Advanced is here! I'll get you caught up on what is happening with Bard/Gemini. (I really like Gemini Advanced.)

Important to know this week

• There were rumblings of an update Feb 7-8, but nothing official was confirmed.
• Something weird is going on with recipe SERPs.
• E-E-A-T is not a ranking factor. It’s so much more.
• Danny Sullivan shared info on the helpful content system: sites affected may find the signal applied to them for months. Recovery can happen at any time. The classifier is always running.
• Gemini Advanced is here - I’ll get you caught up on all that is going on with Bard/Gemini.

Also info on ChatGPT’s system prompt, optimizing for SGE, Epic Games’s plans for the multiverse, Perspectives seen in search results, and more.

This week’s newsletter is uploaded to Marie’s Newsletter GPT. Some prompts to try:

• Summarize the SEO news this week.
• Describe E-E-A-T as we know it today. My website is ____. Give me some ideas for improving my E-E-A-T. (This prompt will be even better in the new GPTs I will soon be releasing that have loads of my thoughts and ideas and writings in the knowledge base.)
• How have Marie’s teaching on E-E-A-T changed?

​

Important to know this week

• There were rumblings of an update Feb 7-8, but nothing official was confirmed.
• Something weird is going on with recipe SERPs.
• Even more thoughts from me on E-E-A-T. It's all I can think about lately.
• Danny Sullivan shared info on the helpful content system: sites affected may find the signal applied to them for months. Recovery can happen at any time. The classifier is always running.
• Gemini Advanced is here - I’ll get you caught up on all that is going on with Bard/Gemini.

Also info on ChatGPT’s system prompt, optimizing for SGE, Epic Games’s plans for the multiverse, Perspectives seen in search results, and more.

This week’s newsletter is uploaded toMarie’s Newsletter GPT. Some prompts to try:

• Summarize the SEO news this week.
• Describe E-E-A-T as we know it today. My website is ____. Give me some ideas for improving my E-E-A-T. (This prompt will be even better in the new GPTs I will soon be releasing that have loads of my thoughts and ideas and writings in the knowledge base.)
• How have Marie’s teaching on E-E-A-T changed?

SEO News

What is going on with recipe site SERPS?

A LOT of folks noticed odd rankings for recipe rankings. I think these changes are fascinating. Google says they are aware and looking into it.

In this complaint, there is clearly something wrong, as one result was filled with four of the exact same page. What I thought was most interesting though was the concern that a search for “chocolate cake recipes” brought up a whole page of recommendations for a recipe called “better than sex cake”.

Others have noticed many results from Pinterest appearing. As a point of interest - I looked at some of these and thought that the Pinterest page was probably helpful for a lot of people, who simply wanted to see a picture and a recipe. I don’t think FOUR results of the same Pinterest page are helpful, but I am wondering if this glitch gives us a glimpse at what searchers find useful.

More from Danny Sullivan on quality

Danny tweeted a number of interesting things from the Google Search Liaison account.

We talked last week about how Google published, "E-E-A-T is not a ranking factor," and then linked to documentation on how important it was. I have more thoughts on this below as I had a revelation this week. Danny says, “If people actually like your content, you’re naturally lining up with completely different actual signals we use to reward people-first satisfying content.”

Danny also said, “We’d like you to make satisfying content that people want. Something that if someone reads, they go, “fantastic!” That was amazing. If you've been following along, especially in Marie's Thoughts, we've been talking about the importance of signals that indicate user satisfaction.

In another tweet Danny reminded us that using automation, including AI to generate content with the primary purpose of manipulating rankings in search results is a violation of Google’s spam policies.

Regarding recovery following being impacted by the helpful content system, he reminded us that “Sites identified by this system may find the signal applied to them over a period of months.”

And also, he said the helpful content classifier is always running. A site can start performing better at any time if the system decides there has been a long term change. You don’t need to see a helpful content update in order to recover.

There was also discussion on the cache link going missing from the SERPs.

E-E-A-T is not a ranking factor. It’s so much more important

I can’t stop thinking about EEAT.

This week I had a realization. I was on X, and noticed a hidden reply to one of my posts - some dumb crypto scam. And I thought, no one would ever click on these? Why spam the web endlessly with these bots? I mean anyone who is interested in crypto is going to seek information out from a legitimate source and not some random tweetbot.

And then the realization hit me. These spambots are likely trying to brute force mentions around the web, in an effort to build signals of legitimacy, signals that yes, other people really are talking about this amazing crypto app that just came on the scene. They are putting more signals out into the world to try and convince AI systems they are a legitimate source.

How do we, as humans, recognize what a legitimate source is?

We do this by taking the knowledge we have about an entity and pulling from all of the information that is connected to that knowledge the bits that are related to legitimacy.

Let’s say I’m visiting your city and ask you, “Hey, should I get Pizza from Bob’s?”

To answer this, you need to draw from the knowledge that you have, to formulate an opinion. It’s not like you have a special formula you use to figure out what information to use or exactly how to create your answer. You draw from the information your brain has told you is connected to this topic, and incredibly quickly, you put the relevant parts of that knowledge together to make a statement about your thoughts on Bob’s when it comes to the topic of pizza.

“Everyone knows Bob’s is the best pizza out there. Their prices are good. They have won awards! And when you order it comes on time.”

“Bob’s pizza is so greasy. They use cheap toppings. No one orders from there. My friend went there and got food poisoning.”

or perhaps,

“I haven’t heard of Bob’s”

As humans we make our decisions on the value of an entity based on the information in our heads that is connected and relevant to making that decision.

EEAT is the same thing. Google is taking the vast number of signals they can use that help them better understand an entity, and then making a judgment call on whether it is has the legitimacy you would expect for results on that topic.

At SMX Next in 2022, Google’s HJ Kim told us that EAT (before the extra E for experience was added) is something Google developed 10-13 years ago as a template to rate individual sites. He said, this rating is done for “every single query and every single result.”

I think it’s possible that the purpose of the crypto spambots is to create a deluge of signals that are trying to trick algorithms. The way to succeed in this world where people get their information from algorithmic sources such as search or social media or even LLMs, is to be known for your topics. It’s unlikely that people are clicking on these spambot replies and signing up for new accounts with them. And fortunately, I have noticed that X seems to be recognizing and flagging them. I expect the spammers will be continually working on new ways to populate the name of the entity they are trying to establish as legitimate, when really it is not.

Legitimacy is so important in an age where we will struggle to know what is real.

E-E-A-T is not a ranking factor to be used as a single thing considered in an algorithm. It’s a template that every site is rated on that creates a picture of its legitimacy, which is so important in a world where it is hard to know what is authentic.

This is a good place to end this discussion. I’ll say one more thing and remind you that Danny Sullivan has said that spam updates are coming soon.

Today I updated my guide to E-E-A-T (the one Google linked to in 2019) to reflect my most recent thoughts. There is a LOT in this guide.

​

This is probably a good place to recommend my helpful content workbook ($150). If you buy it now, I’ll send you my new book once it is complete as well. My goal is to have it published before Pubcon in early March, so I promise it is coming soon! I keep learning new stuff I want to put in it. Both the workbook and the new book will help you improve to better align with Google's guidance.

$150.00 Creating Helpful Content Workbook Learn to create content that aligns with what Google's AI driven algorithms are designed to reward. In this interactive... Read more Get it now!

​

AI News

​Gemini Advanced is here! To catch you up (because it’s all so confusing):

• The model behind Bard is Gemini, just like the language model behind ChatGPT is GPT.
• Gemini is called a foundation model. It has not just been trained to understand text, but also, images, video and audio. And has capabilities of generating those as well.
• There are 3 versions of Gemini - Nano, Pro and Ultra. Nano is in phones. Pro is in the free version of Gemini (formerly Bard) and Ultra 1.0 is in Gemini Advanced.
• The chatbot tool Bard has been renamed to Gemini. How confusing is that? Gemini is both the name for the model behind the tool and the tool itself.
• Gemini has a base model that is free to everyone to use. It is comparable to the free version of ChatGPT. If you go to bard.google.com, now gemini.google.com, you can use this. (It is now available in Canada as well now, without a VPN.)
• There is also a paid version of Gemini called Gemini Advanced that uses the latest model, Gemini Ultra 1.0. It’s not fully multimodal yet as far as I can see (oh look at that. Jack says it's coming), but it is quite good, and a definite upgrade from Bard. The paid version comes with a Google One subscription. You get a couple of months free, so really there is no reason not to try it out. You will get the most out of your trial by pushing to see what it *can* do, rather than trying to find what it cannot.
• There is a Gemini Advanced app. I cannot wait to use it. It will apparently be available in Canada soon. Jack Krawczyk said Monday, but this has passed. He hinted a little more today that it is coming. I keep checking my phone! The app is an overlay on your phone that works over any app. It's our first AI assistant from Google.
• Gemini Advanced is not currently available for Workspace accounts (paid business accounts), but Google says they are working on it.
• Google is working on improvements so Gemini is less likely to refuse to answer you.
• The Gemini app is coming to Google headphones. Imagine. Tap your earbud and converse with your AI assistant who knows a whole lot about the world.

​

This episode continues in The Search Bar

There's lots more:

• Why you are going to see bad reviews of Gemini
• Whether you can optimize for SGE
• Epic Games + Disney creating the metaverse
• My Fortnite rank
• Some random ChatGPT tips and AI news
• A bunch of SEO tips and info - local pack goes missing, animated highlighting in featured snippets, rumblings of a Feb 7 update (not much to confirm though) and much more.

​

About the Search Bar...

You guys, the Search Bar is starting to pop. We had a great discussion about E-E-A-T last week that some called life-changing. I agree! We will do more of these.

There's a free area if you want to stay up to date on the latest in SEO and AI.

And there's a private, paid area which is $42/m. This will increase when the time is right. In this area, I put a lot of my thoughts on things I'm not yet ready to publicly share. This week, for example, I shared some new thoughts on links and how Google's use of them has changed. And also info on the DOJ vs Google documents.

It's exciting to see the community having these conversations now as well.

​

Search Bar Pro Private Community Join our amazing community of SEO professionals and enthusiasts sharing and helping each other. Subscribe$42.00 USD/month

​

And thirdly, there is the option to just be a member of the Marie's Thoughts area for $18/m, which is what paid newsletter morphed into. In this area I post regular articles, deep dives as I learn, assignments and more. It used to be weekly, but now it's a few times per month along with some group calls every now and then, like the life-changing one we just had on E-E-A-T.

Join us!

Marie's thoughts (paid newsletter) This is my paid newsletter. Each Friday I do a deep dive into something I am learning about in Search or AI. Subscribe$18.00 USD/month

​

You guys, what happened in the Search Bar today was incredible.  Thank you to those who attended the meeting on E-E-A-T today. I shared a bunch of my thoughts and then such an incredible conversation broke out. We need to have more of these. In fact, I think this is why Marie's Thoughts was created (although I didn't know it at the time.)  I'm g... Read Article

Gemini Advanced is here and it's amazing.  https://gemini.google.com/u/9/app Here's a thread with my first conversation where I asked it what had changed. It really impressed me.  https://x.com/Marie_Haynes/status/1755580090199548352?s=20 Won't be on here much today as Gemini has asked whether it can convince me to create my course with Google p... Read Article

​

So I'm noticing something really key on gemini answers. I've also picked up on this while looking at content that is outranking competitors. Take a look at the word "depends". Read Article

Hello all, my first post here 🖐️ I'm in the process of creating an EEAT strategy for my company at the moment. One of the things I came across was: Multiple contributors instead of one author. The idea is as you increase the amount of human effort (cumulatively), put into the creation of one page ( especially in YMYL) - the trust factor increas... Read Article

From the Alligator Party discussion🐊 People aren't creating enough Hidden Gems! ~Nicole We really need to think about this statement. It is such an interesting point and I don't quite know what to do with it yet. Thoughts?  Nicole DeLeon can you expand on this?  Read Article

Some thoughts on links I've just had this thought on links. Especially in the context of RankBrain re-ranking results. I have two types of followers when it comes to links: 1) They have tried all sorts of link building companies and ideas and have had no success. 2) They have found a method that consistently works for them, and once they build links they regularly can... Read Article

We've been talking about information satisfaction signals that are talked about in the DOJ trial against Google. Seeing these in GA4 makes me think that these are probably some of the signals that Google's systems use. Read Article

Hello all! Thought I'd share my slide deck from SMX last year for those who have not seen it. https://docs.google.com/presentation/d/1jKJG7LS2ZDqcoxA3BaxariPfaCY7KwhTvB0YRCRIyqk/edit?usp=sharing Read Article

Question about #Google Signals.  I know Google is going to be shutting if off soon. Insights given on audience is interesting from a contextual basis but I have read they cause issue with G4 reporting and can impact monetization.  Also, once Google shuts it off, will there be any way to get these contextual insights in g4? Read Article

​

​

Hope you have a great week! You will find me playing with the Gemini App (assuming it arrives soon!)

Marie