Thread by @darth_na

We can always try asking
(again).

Hey, anyone at @googlesearchc, or maybe @JohnMu or @searchliaison ...

Does Google actually try to verify experience or expertise,
or is it based on linguistics/phrasing?

#SEO #EEAT

https://twitter.com/lilyraynyc/status/1656296085386657792

My money is on it being linguistics/phrasing.

G simply cannot know if Bob bought the giant pink 18" floppy dragon pillow (yeah, some of you went there!), or is simply making it up.

(Note: they may be able to guess if Bob's full of shit though, due to 34K reviews in a week ;))

For example ... you're a person, not a machine.
You have good judgement, right?

So you know I made this up...

https://twitter.com/darth_na/status/1613170327843999744

Now ... I know some people claim G can't do a "people database", and thus can't really do "authors",
but personally - I don't see an issue.
Technically, it should be smaller than the Link Graph (fewer people than sites, fewer edges etc.).

But not all qualifications are public.

So G would have only a sample.

They could use association and inference.
Similar to the concept of Authority sites and Distance for links - they look at mentions/citations/links from such sites to "people" (thus why mutual interlinking between works/profiles is important).

But that doesn't cover things like "reviews".
Not unless G are scoring people with some form of "trust" metric.

That's a fair bit of resources,
vs simply parsing content looking for "I" and "we" etc.

But I'd be happy for someone at @googlesearchc to say I'm wrong ;)

So which is it @JohnMu / @searchliaison ...

Shallow/(relatively) Simplistic parsing of text,
or a more complicated record/system tracking and scoring people?

And honestly - people really should know.

Either, people are getting confused with EEAT,
and thinking they must hire expensive people etc. to gain an edge,
or you've got a collection of data on people that the public should know about,
or it's just smoke and simple parsing.