Could it really be this simple? Outlink and rank.

[Sgt_Kickaxe]

Every site is a case study if you have access to enough of it's data and every diligent webmaster pays attention to the data at hand so...

The number of sites I've watched which follow this pattern is at 14... out of 20 or so.

- Launched 2+ years ago
- No updates in 1+ years
- Lots of pages ranking for multiple keywords in the top 50 but few in the top 5.

From there one change is made - the site adds a link to a "trusted" news article about the general topic. The more prominent the locations(on both ends) the better.. Nothing more is done but after the next spike in Google crawl activity, and after the typical major update that follows it, holy cr*p!

Suddenly the site is "trusted" judging by the number of keywords that jump into the top 5 across much of the site. My sample size is too small but something definitely makes a big difference. Oddly it doesn't work when you link to authority similar sites, it MUST be a media/news site. The biggest jumps I've seen were from sites linking to media from their index page and in all cases nothing happened until a spite in crawl activity followed by an update happened.

Could it be this easy? It takes a long time to test.... thoughts?.

[martinibuster]

Spammers do it and it doesn't help.

Brett suggested outlinking to authoritative sites in his seminal post about ranking in Google.

Many SEOs twisted that advice to subsequently link to "authority" sites, without regard to how it fits with the page.

Outlinks to "authority" sites became rote, a near religious act of SEO faith.

It's not just outbound links though. There's a lot more going on

I wrote an article about it on SEJ called Link Distance Ranking Algorithms.

[Kendo]

There are news sites and there are news sites that are mostly spam sites. I occasionally use a news wire service and the results are great. Searching our keywords after the last press release the results were actually more impressive than before. For that keyword string our site usually holds 2 of the first 4 positions, shifting positions each week. After the last press release our site and the news sites who featured our article filled 8 of the 10 spots on the first page of results.

But it is not a cheap process.

[ColinM]

No I do not think it is that simple.

If it were that simple the SERPs would be a complete mess.

There are billions of blackhat spam pages that link to a variety of articles, none of which you will see in the SERPs.

The biggest factor these days is simply relevancy IMO.

[Wilburforce]

We don't know the population-size of sites to which those criteria apply, but I'm guessing

1. 20 is not enough to be representative, and
2. Your selection may be biased - how were the 20 sites selected? - in some way.

Importantly, also, correlation does not imply causation.

On the question itself, no, I don't think it is that simple, but I can't employ scientific rigour to disprove every unlikely thing - I'm busy enough already.

[aristotle]

It might depend on how often the outlink is clicked. From what I've seen, google's algo puts a lot more value on links that are frequently clicked, as compared to links that are rarely clicked.

Also, as others have mentioned, the content of the two pages must be closely related.

Most likely spammers usually don't know how to do it in a way that might work.

[Sgt_Kickaxe]

Outlinks to "authority" sites became rote, a near religious act of SEO faith.
It's not just outbound links though. There's a lot more going on
I wrote an article about it on SEJ called Link Distance Ranking Algorithms.

I've read your articles and respect your views but I think you missed a little nuance in what I said. I specificaly said it did not seem to work when linking to "authority sites", I said it seems to work if you link to an article about the same general topic on a trusted media site. Only recently too, I've been a webmaster as long as most here.

This didn't work in the past which is why it caught my attention now. I have a theory but would appreciate other views.

Theory - Media might be used to filter results towards desired views and opinions by strengthening the signals approved media sites provide. We already know that citing a media site with the wrong views can tank your traffic but is the opposite fast becoming true? A byproduct of anti-misinformation algos perhaps?

What surprised me is that these are NOT political sites but they benefited greatly by associating with specific trusted MEDIA sites.... but not from similar "authority" sites(as you pointed out)... It has to be media. Also, just linking to the media site in general changed nothing, it had to be to an article about the subject of the website and, in almost all cases, a specific aspect of the subject.

It's like the link to the MEDIA site became an approval sticker if you will. It makes sense, too, if you wish Google to trust your content they might first want to see if they trust your cohort.

Small sample, hard to test, no insider info etc... but it's an observation that raises the question: Media site > authority site?

note: these were not my sites and they are not political. This was just an observation and question that needs more evaluation, imo. I'm very aware that a bazillion things can and do influence results but nothing much had changed with these sites other than citations. OUTGOING citations.

[martinibuster]

I think you missed a little nuance in what I said.

Nah, it's the same. What tends to matter are the inlinks.

If they don't have any burned inlinks and they do that... well... then...

[FranticFish]

Not trying to shoot down the idea, but I think (as others have said) the inbound linking history of all 20 sites over time is at least as important as what happened to their content over time.

You don't mention whether or not you have crunched the link profiles of the 20 sites and compared them all against each other. What patterns (if any) did you find there?

Were there things that were common to the 6/20 - or the 14/20? Or were there things that were common to the 20/20?

Maybe they do combine to produce the effect you noted, but you have to try to disprove your theory too :)

[robzilla]

Theory - Media might be used to filter results towards desired views and opinions by strengthening the signals approved media sites provide. We already know that citing a media site with the wrong views can tank your traffic but is the opposite fast becoming true? A byproduct of anti-misinformation algos perhaps?

"[Google Search and Google News] do not make subjective determinations about the truthfulness of webpages, but rather focus on measurable signals that correlate with how users and other websites value the expertise, trustworthiness, or authoritativeness of a webpage on the topics it covers."

[NickMNS]

@Robzilla Tha's what Google say, but everyone knows they don't always do what they say.

The theory makes perfect sense, Google can easily develop a complex algorithm for measuring outbound links to specific website which report favorably on issues it agrees with. The algo would only need to keep track of favorable and not favorable websites, then create a scoring system to count those links and prominence of those links adding a few points for top of the page favorable links then deducting points for below the fold unfavorable links and so on. It also adds more point for do follow links. There is likely also bonus points for multiple links to the same website. This would be easy for Google and it's army coders, barely an inconvenience.

Alternatively, Google could just take the topic of the page, which it does anyways for "normal" search and simply demote sites that it doesn't agree with. But why keeps thing simple when instead one can make in complicated. More complex makes for a better theory.

[Andem]

For what it's worth, I seem to have a similar experience to Sgt_Kickaxe when I've researched other sites. It's not linking to news/media sites specifically but sites that I would consider to have more "authority" than the site linking out. I've never been able to definitely pin this down and have done no testing on any of my properties.