Darth Autocrat (Lyndon NA)

.
:: Google's Link Guide Changes ::

So ... G made some updates/additions to their docs,
and I'm seeing a few mentions of it ...
... and ...
I'm not seeing any real reason for it?

I don't think anything is "new" per-se?

https://t.co/C2xSyzMwZ4

>>>

#SEO #Links

>>>

Not squishing links together is common sense

Link text is known

Descriptive/informational link text is common sense

G using context (surrounding text) is known

G attempting to extract URLs for discovery is known

Not using JS reliant links is known + common sense

>>>

So it's really just a collation of the things that Google/Googlers have said for .... a decade+?

(much of this stuff was covered when I was a TC, so over 10 years ago!)

.

The bit that concerns me though ... is the Outbound Link bit...

>>>

>>>

"...
Linking to other sites isn't something to be scared of; in fact, using external links can help establish trustworthiness (for example, citing your sources).
..."

I can see:
1) People declaring OBLs as ranking factors
2) People jumping on (E)EAT (again!)

>>>

>>>

For those of you that didn't know;
links are a factor for the origin page, 
as well as the destination 
(the text in an <a> carries more weight than the same text in a <p>)

But that's not the same as saying that outbound links to good sites is a ranking factor.

>>>

>>>

So, if at all possible,
can someone from @googlesearchc please clarify that a bit more?

Are you saying/suggesting OBLs improve ranking 
(if so, how), 
and if there is influence from EEAT
(if so, is it actually part of the algo, not just QR metrics)
.

Thank you.

(Oh, and if you are able to help ... can you also make any required distinctions between Discovery/Crawl queue and Ranking - such as identifying, parsing and extracting URLs from JS based links ... do those pass PageRank etc., or are they solely for discovery still?)

Darth Autocrat: 
Internet business consultant 
(SEM (SEO/PPC), CM & SMM, with UX, CRO & ORM)

I Fix things for other SEOs, (or what they broke).
He/Him

Chris Moran

Barry Adams 📰

.
:: AI Content questions ::

There's some solid thinking from @chrismoranuk in the linked to tweets,
including a number of thoughtful and important questions.

But I have some of my own...

>>>

#SEO #AI #GenerativeAI
Via @badams https://t.co/F2HhJf0kKD

Lots of businesses right now seem to be asking the question: "With generative AI poised to completely shake up the internet and more, what do we need to build to avoid being left behind?" And I think it's a really useful starting point to really consider where we are now

Head of Editorial Innovation @guardian. Ophan, digital publishing, news, culture change, data and analytics

>>>

How did something (several things!) with so many (known) issues get to go public?

How comes these (potential) flaws were not declared upfront/immediately?

Why were certain types of "prompts" permitted,
(ones known to trigger "creative" responses)

>>>

>>>

Why were not attempts made to alert the user to verify/validate generated content (if not automate those checks)?

Why are there questions regarding the content used by (some of) these tools?

>>>

>>>

How did certain sources get "approval" for inclusion,
despite ownership/IP etc. being an established and largely aware-of concept?

What efforts were made regarding filtering out "bad sources" (inc. those of poor taste/nature, inaccuracies etc.)?

>>>

>>>

Considering the nature of the output, and the structure of the data, why were sources (intentionally?) not tagged/incorporated?

etc. etc. etc.

We're not talking about a bunch of naïve, overly excited little script-teenies!
We're talking established professionals!

>>>

>>>

We're not talking about some ad-hoc, multi-faceted, cellular/insulated setup with randoms chipping in here or there!
We're talking structured, organised, well funded setups,
with experienced people, in positions!

>>>

>>>

So ... I'm left with the big, broad questions!

WHY!?

Why is it, 
with this many brains were involved,
this many experienced people,
in various projects,
permitted (encouraged?) to produce things with so many obvious/noted flaws,
and to go public with them?

>>>

>>>

I have a horrid suspicion that the vast majority of specific/detailed questions ... can be easily answered by that Big "Why?",
and I don't like a single one of the potential answers,
esp. when you look at the companies and personalities involved!

Kristina Azarenko

.
:: Website migration issues ::

Good list from @azarchick! 
Off the top of my head... 30?

1. Redirects no Rewrites
2. No redirects/Rewrites (dupes)
3. 302's not 301's (slower)
4. JS/Meta not server/script
5. Redirect chains

>>>
#SEO #Migrations https://t.co/r6DUsFlKeC

8 common issues with website migrations

A 🧵

SEO Hero | Building MarketingSyrup SEO Academy. 
Creator of THAT! technical SEO Course. Join 300+ ppl 👉  https://t.co/JJsvakM6jL
Tweet about Tech #SEO + Mindset

>>>

6. Incorrect order (older redirects in effect, not new)
7. Incorrect destination of redirects (HP etc.)
8. Untested redirects (fails/wrong dest.)
9. Destination is/are noindex
10. Destination is/are disallowed
11. Origin set to noindex (uses CLE + Meta/JS)

>>>

>>>

12. Origin set to disallow
13. Redirects taken down to soon (old DN drops)
14. Content is changed as well
15. Design is changed as well
16. Content + Design are changed as well
17. Pages are "cut"
18. Pages are "merged"
19. Nav/ILs get "optimised" at same time

>>>

>>>

20. Canonicals are left with old values
21. Canonicals are changed (see (14. "merged")
22. Resources (CSS/JS/Images) point to old DN
23. Internal links (ILs) point to old DN
24. Tools (GA/GSC) aren't setup/migrated etc.
25. No tracking of indexed pages/rankings prior

>>>

>>>

26. Reversal within weeks
27. Multiple migrations within a short time
28. Platform changes before redirects settle
29. Mass content changes before redirects settle
30. Destination/New site has "errors"

... so ... what did I miss?

1/🧵
🚨:: Does Google use Bounce Back Rate? ::🚨

*sigh*
It's not even 2 years since I last covered this!

🧵&gt;&gt;&gt;

#SEO #BounceRate #BounceBackRate #BounceBackToSERP https://t.co/ReMgT3uW6N

1/🧵
🚨:: Does Google use Bounce Back Rate? ::🚨

*sigh*
It's not even 2 years since I last covered this!

🧵>>>

#SEO #BounceRate #BounceBackRate #BounceBackToSERP https://t.co/ReMgT3uW6N

2/?

Here's the thread I did at the end of 2021...

https://t.co/RzmRDif91i

But if you don't want to read through that,
I'll do a quick summary here...

>>>

3/?

Bounce Rate (BR) is an "Analytics" metric.
Dwell Time is a made up metric.
Click Through Rate (CTR) is a SERP metric.

G have stated none are used for Rankings.

But they "do" use user-behaviour in the SERP,
for various things (such as gauging algo performance etc.).

>>>

4/?

So we instantly hit Google's "semantic slide",
where they can answer truthfully/honestly,
and yet mislead at the same time.

Simply having a different name permits it.

Through in that the metrics people keep on asking are simplistic, shallow, limited etc.

>>>

5/?

There are numerous patents and papers for
"implicit feedback"
(user search/SERP behaviours)
which cover a range of potential uses,
ranging from spotting trending queries,
including algo performance gauging,
through to potential satisfaction proxies.

>>>

6/?

People will likely do better to get the concept of
"search sessions".
This may be a single query, or include multiple variations/refinements,
clicking on "suggestions" etc.
(So long as the primary topic/nature remains the same, it's the same session)

>>>

7/?

How often do you do a search, 
click a result,
decide the page isn't for you,
hit the back button,
click another result
(or click a suggestion etc.)
?

There you go.

>>>

8/?

It's NOT:

* Bounce Rate 
(only n% have GA setup, and BR includes closes etc.)

* Click Through Rate
(On it's own, weak/noisy - consider "last click" instead)

* Dwell Time (long/short click)
(Short time + a new search session is not the same as short/long + 2nd click)

>>>

9/?

* Pogo Sticking
(The influence would be dependent on whether the search session remains, or is amended etc.)

Each every single metric and shallow reference is flawed, as they can be noisy.

Instead, you need to look at multiple features.

>>>

10/?

If you've seen the latest tweet by @indexsy  https://t.co/p4HCYbMrC7

You'll note the phrasing "used to have".
This could be a simple back reference,
or suggesting that it's not used "now".

Personally, I'm willing to bet G still user SERP Behaviours,
but not simply.

>>>

11/?

So - do yourself (and everyone else) a massive favour;
don't over-simplify the concept of SERP Behaviours for Ranking.

Terms like "Bounce Rate" have no relevance.
Things like "CTR" are too simple/noisy.

>>>

12/?

G would have to combine multiple factors to make a strong enough signal to punch through all the noise.

AND ... they'd have to implement protections against click-farms/bots etc.

>>>

13/?

This is an age-old concept, and plenty have tried to abuse it ... so G are likely looking at 
Quantity over Duration (consistency/trend)
as well as multiple individual factors (returns, position, CTR etc.).

If you want more details:

https://t.co/RzmRDif91i

@darth_na's Latest Twitter Threads

Read threads by Darth Autocrat (Lyndon NA) and get notified about future ones.