How could Google program EAT into the algo?

[goodroi]

Google has been strongly encouraging webmasters to focus on E-A-T (expertise, authoritativeness, trustworthiness) and they talk about it with their internal search quality guidelines. But that is more for their internal quality control process and not about their secret algo. So how do you think Google is (or could) implement the EAT mindset into the algo?

They could be looking at readability scores. I personally pay close attention to those scores but more for usability and branding reasons. It probably is too noisy a data point to use.

Maybe they are looking for certain trigger words to be present on the page. Like if the page has "MD" in a byline then it could signal to Google's algo to trust it for health queries or if it has "IANAL" than it could suppress the page from showing up for legal search queries. That would be too easy to abuse or falsely trigger, so probably not the case.

Or maybe they are have a TEALrank for backlinks (aka Trust Expertise Authority Link rank). Similar to how Pagerank functions, find a handful of E-A-T sites & seed them with a TEALrank of 10 and then the first circle of sites they link out with become TEALrank 9, and the cycle repeats. Google knows how to do Pagerank, so repurposing it for EAT would be very feasible but then again Google also knows links are already a high target of abuse.

So how do you think Google is (or could) implement EAT mindset into their ranking algo?

[Abaros]

Imagine that you have to solve a problem: how to determine the quality of millions of publications.

The solution is relatively simple if you have sufficient resources:
1 - You hire thousands (maybe millions) of experts.
2 - You build an expert AI in everything.
3 - Look for patterns in the highest quality publications.

I don't think it's necessary to explain why options 1 and 2 are not viable.

Google uses the word "signals" very often and that has been Google's strategy throughout its history, starting with "anchor text". There are also "negative signals".

Having the largest database of information ever created, it will not be difficult for them to find patterns.

So make sure you don't give "negative signals" and that you give "positive signals" to Google. I'm sure that pages that have a high EAT value have certain types of links, certain types of mentions and other types of patterns.

I would also like to know which ones specifically.

---
By the way BERT is no mystery, you can even download it here: [github.com...]

There's no such thing as a "intelligent" AI yet.

[aristotle]

If I may inject a bit of humor, I once saw a story about fake "board of directors" pages. They're not hard to create -- you just make up some fictitious bios of corporate executives, financial officers, attorneys, Ph.D.s, etc, with facial photos copied from other sites.

On a more serious note, I like goodroi's suggestion about "TEALrank". I think it or something very similar is probably an important part of this.

[Wilburforce]

So how do you think Google is (or could) implement EAT mindset into their ranking algo?

I'm not sure that E-A-T is describing anything new: it is just using different terms.

PR used to be a pretty good indicator for all three elements, so you might ask the same question a different way: what better indicators are there?

While Google has put a lot of work into negating link manipulation (so a site with an apparently fantastic backlink-profile may no longer rank as it once did), I think it is still all about links. Good backlinks gained naturally are probably the best E-A-T indicator available, and therefore probably carry more ranking weight than anything else.

So in direct answer to the question, I don't think Google is implementing a new mindset, just refining what it has always done. Renaming it doesn't make it a different animal.

I'm not sure how much difference it makes if you have to scroll past ten ads to get to the first organic result: all the E-A-T in the world won't outrank a dollar.

[samwest]

I personally think EAT means "eat their lunch". When after 20 years of expertise, authority and trust and holding within the top three, I am now being out ranked by two Pinterest pages one of which is a single unrelated pin, and two absurdly ad filled, non authoritative aggregation blogs (on every related query no less) ...the entire idea and trust in E-A-T goes right out the window.

If there is AI involved in their top secret sauce (quite sure there is)...it appears to be broken...or intentionally biased to shape traffic. Big surprise.

[explorador]

Google has been strongly encouraging webmasters to focus on E-A-T (expertise, authoritativeness, trustworthiness) ... How?

My guess and just a guess, is one of the news report fact-style redacting modes on articles, going to the point, less words, stories divided into paragraphs that each describes fact based small stories. This is a style used in order to allow readers to jump and still navigate texts quickly in "scanning mode".

Why? again, just a guess, because Google enjoys using our content to answer questions directly on results. Such styles as explained above make it easier to extract and show, as Google has been doing for a while. Then people won't even have to click and jump to your site to get "useful" snippets of text, useful for Google and readers, but not turning exactly into visits. If you think about it... this is already happening.

[aristotle]

the entire idea and trust in E-A-T goes right out the window.

Well some niches might not have any high-authority sites competing in them. So if none of the sites have much authority, then the E.A.T. part of the algorithm won't have much effect on the rankings for that niche. Thus if one of the sites drops in the rankings, it's probably due to other factors rather than to authority.

[iamlost]

Fun retrospective discussion links:

Trust and Authority - they are not the same thing [webmasterworld.com], September 2008.

How To Establish EAT (Expertise, Authority, Reputation) [webmasterworld.com], August 2014.

Bing: Trust and Authority Usually Sees Higher Rankings [webmasterworld.com], October 2014.

Google's Knowledge Graph:Knowledge-Based Trust: Estimating the Trustworthiness of Web Sources [webmasterworld.com], March 2015.

How Google can determine trust [webmasterworld.com], May 2016.

Google Adds Trust Project Labelling [webmasterworld.com], November 2017.

E-A-T-ing well now even more critical [webmasterworld.com], May 2019.

While my broader thoughts on the matter of how EAT might be implicitly determined by an algorithm are mentioned in some of the above, given the explicit question and the new day with my feet up looking for an excuse not to mow the grass...

Note: an important disclaimer is that each of the following has inherent weakness, is game-able; the critical flaw in implicit approximation of explicit goals.

Expertise: expert skill or knowledge in a particular field.
* credentials aka awards/certification/licensing in field.
* career aka number of years working in field.
* performance aka record of publication/speaking on field.
* peer respect aka number of citations, mentions by others in field.

Authority: power to influence others because of recognised expertise/knowledge in field.
* expertise, see above
* legal/rational institutions in field, i.e. professional associations, research universities, government oversight/regulating organisations.
* link graph nodes of exceptional authority aggregation in field/niche/vertical.

Trust: firm belief in the reliability, truth, ability, or strength of an entity.
* credibility in field aka authority (see above).
* reliability aka testimonials, mentions.
* safety/security aka level of confidentiality (eg PII), ethical/legal/social accolades versus complaints.
* focus aka is interest in field total or tangential, professional or hobby, etc.

The foregoing is largely field/niche/vertical oriented; I suspect there is a probable broader website accessibility/usability perspective input as well although such always seem to be minor ‘tiebreak’ level influences (probably because the greater their weight the greater the impact on ‘too big to exclude’ sites).

Dangnabit the grass didn’t mow itself.

[glakes]

I personally think EAT means "eat their lunch". When after 20 years of expertise, authority and trust and holding within the top three, I am now being out ranked by two Pinterest pages

I couldn't agree more. Look at all the scraped pages and images that are ranking. If Google can't/won't figure out the original source of the material and rank them, which should be quite easy, then why would they program EAT in the algo?

I don't believe AI exists for ranking the best content. I believe AI exists to drive ad revenue, which consists of not only displaying ads that a specific user will likely click but also dummying down organics to achieve paid clicks. When you're the dominant 800LB gorilla in any market, you don't have to produce the best of anything - you just have to make people believe it's the best there is.

I don't see Google investing much of any money into EAT because that would cut into profits and potentially conflict with how Google's AI maximizes revenue. You can't even get an English speaking Ad rep from Google, for a 30 minute conversation each month, while spending $xx,xxx.xx+ monthly. That's how cheap Google is.

IMO, EAT is another distraction that keeps webmasters chases their tails. In my ecommerce industry, the vast majority of consumers could care less about expertise, authority and experience. Most want cheap which is why they buy Chinese products, with known problems, from businesses run by people that can't even speak English. Most people, both consumers of products and information, will consume anything that is spoon fed to them. Look at how many people cast their votes for politicians in elections. Having done no research, they vote for the name on the ballot that is most familiar to them. That's how they search too - using Google because that's the most familiar search engine in their limited reality.

[explorador]

Consider 3 important trends under the light of social behavior and -what's happening in the world right now-:

1. The world doesn't behave like this forum WebmasterWorld, so "this is happening", result? wrong, false, post your source, hey your source is not enough for me so I say this is false, your source is biased, what you said is biased, wrong, and it only promotes your own agenda and point of view, etc, you get the idea. Here in WebmasterWorld you will see discussions in good faith and IMHO the moderators do a great job too, not to mention most members are adults with some years in the business, but the world is in chaos right now.

2. Lot's of content on the web is being used or someone is at the moment trying to use it to build some arguments for a discussion that would go out of proportions. I've been finding for some time terrible extractions of content in order to build accusations, false arguments, etc. It's not just me:

Philosopher Axel Kaiser gave a nice speech describing how most common expressions in the now-world are nearly impossible without entering in some hate speech-someone-waiting-to-get-offended. You have to consider this under the light of how many authorities in diff fields are now having issues even at lectures, because the new audiences EXPECT them to say something surrounded by sources, books, links, studies, etc (even if they won't take such info seriously), anyone posting something or giving a lecture will be quickly under attack "and how did you get to that conclusion?", and having 10, 20 years of experience will not serve as an answer.

1. Content is king? yes, but what's good content? what's expertise, authoritativeness, trustworthiness? there are lots of publications on the web about products, services, reviews, methods, processes, etc. And they used to rank pretty well, but then came the no-content-providers who would just steal that content in order to post 10 reasons why whatever is not working as you expected. It doesn't end there, right now there is a spike on content created in order to criticize YOUR content. The 10 reasons why... articles and videos are now being replaced by comparisons between your content and other content, or "why your content is wrong". Someone could easily create articles on webmaster forums comparison including WebmasterWorld and it would be very easy, ending with what it seems someone who really knows and will even include links to the sites and "sources".

Remember, today everyone thinks is a philosopher, a news reporter, a social warrior, an idea creator and protector, and everyone is talking about complex topics in order to tell people how and why they are wrong. I've been following closely some content creators that IMHO create pretty solid quality stuff and they rank X on their platforms (web, FB or youtube) but suddenly they are being outranked by people who don't create content, all they do is criticize their content and tell people "why they are wrong", surprisingly they get traffic, solid traffic and views.

[jediviper]

So no-one from here has noticed the possible connection of implementing EAT signals to the May update?

I have a couple of sites that I am monitoring in my niche (also from our portfolio), that it makes almost obvious the fact that the sites with more official info (governmental links, company's info at the footer etc) started to rank higher after the 5th of May and the opposite happened for other sites that had no such info.

[explorador]

Don't want to sound negative (or just be negative), but that sounds like the MFA's issue a while back (G creating a problem). Unless Google can read like a human, I foresee lots of websites using official sources-links for their content, even if there is absolutely no connection with the content, just like the famous internet and technology quotes by Benjamin Franklin (memes).