Discussion: Link Profiles and Niche

Prentzz

Prentzz

BuSo Pro
Joined
Mar 19, 2015
Messages
289
Likes
377
Degree
1
I've heard this idea over and over that certain link profiles can fly in one industry and not in another. For example, people might argue that because almost all of the prominent successful sites in X industry have hundreds of directory links, blog comments, web 2.0's, blatant PBNs etc, that the algorithm is more lenient towards this when compared to an industry Y where the sites typically have more 'natural' link profiles. They would argue that if you did the same in industry Y you would get slapped. On the flipside, if you run everything more natural in industry X you would barely move the needle.

The argument is essentially that the Google algorithm is granular to niches, allowing for different anchor text ratios and levels of 'spam' before triggering any 'flags'. Not sure how else to explain this idea..

Thoughts?

I'm sure this is going to be anecdotal because I haven't found any research or even reliable case studies on this idea.
 
Maybe in the past, different backlink profiles would fly in different industries, but these days (and this is only opinion) post-Penguin I'd say that Google is handling spam a lot better. Most trashy spam is ignored, most profiles (even if they're dofollow) are ignored. Stuff like Web 2.0's are handled with page and domain level page rank metrics (they don't have any).

But as far as anchor text ratios goes, maybe it's still granular. If you think about it, all rankings are comparisons against other pages qualified for ranking for any specific term, so you could say every single keyword has it's own granular weightings that work or don't.

I don't think Google is following the logic that one page in a niche is better than another because it has more PBN links, no matter what that niche is. They might get fooled by it, but I don't think they allow it.
 
Future State said:
Maybe in the past, different backlink profiles would fly in different industries, but these days (and this is only opinion) post-Penguin I'd say that Google is handling spam a lot better. Most trashy spam is ignored, most profiles (even if they're dofollow) are ignored. Stuff like Web 2.0's are handled with page and domain level page rank metrics (they don't have any).

But as far as anchor text ratios goes, maybe it's still granular. If you think about it, all rankings are comparisons against other pages qualified for ranking for any specific term, so you could say every single keyword has it's own granular weightings that work or don't.

I don't think Google is following the logic that one page in a niche is better than another because it has more PBN links, no matter what that niche is. They might get fooled by it, but I don't think they allow it.
Click to expand...

Thanks for the reply. I agree, Google is generally handling spam better, but in the dirty corners of the internet where I tend to rank I still see generally dirty link profiles on sites that rank excellently. These can't be explained away by their quality links, because these alone would not rank the site.

The only argument against that is potentially you could do this in any niche and get away with it, but people just don't because of propaganda white hat etc etc.

Site category grouping is fairly simple, you only need to look at how Majestic does their TF and CF algorithms to see that it's relatively easy to do accurately. Google MUST be doing this to some extent, otherwise link relevancy would not exist. You need to categorize pages and sites overall for relevancy to be measured.

So, they already have the groupings and they would be able to find averages for each of these and use that to tweak the algorithm.

Every industry is different. The automotive industry for example has hundreds of extremely popular forums, so it would be weird to me for Google to treat forum links in the automotive industry the same as they would in an industry where forums are practically non-existant.

Maybe they don't literally tweak the algorithm for forum links, but potentially it could allow for more UGC or no-follow links in one niche than another.

Thoughts on this?
 
Prentzz said:
Every industry is different. The automotive industry for example has hundreds of extremely popular forums, so it would be weird to me for Google to treat forum links in the automotive industry the same as they would in an industry where forums are practically non-existant.
Click to expand...

They did say they were going to start treating nofollow as a hint, probably for exactly the reasons you're stating. Not all user generated content is trash. If there's other signals that show it's high quality or there's strong moderation on a forum, why not count them. Maybe they'll count as half of what a normal dofollow might as a maximum and downgrade them based on content quality or domain authority. Who knows.
 
Ryuzaki said:
They did say they were going to start treating nofollow as a hint, probably for exactly the reasons you're stating. Not all user generated content is trash. If there's other signals that show it's high quality or there's strong moderation on a forum, why not count them. Maybe they'll count as half of what a normal dofollow might as a maximum and downgrade them based on content quality or domain authority. Who knows.
Click to expand...
Thanks for commenting. I agree that plenty of UGC is quality enough that it should be counted, which is why for my main business I'm very happy to post on forums and leave blog comments. The top competitor in my industry has plenty of legit forum links, coupon site links, director/business listing links and sponsored posts that are either nofollow or not, almost always with it mentioned that they are sponsored.

In my opinion, all of these are good links that I would want, so long as we also get some higher quality links in contextual content. Wouldn't want ONLY those lower quality links, but I'd be happy for it to represent 70% of my link profile in the first year or two, and my experience reviewing legit ecommerce sites as they grow is that this sort of ratio is natural, if they aren't actively building contextual links.

Thoughts on this?

Also @Ryuzaki do you have any opinion on whether Google in any significant manner alters their link algorithms depending on niche and the link profiles of sites in that niche?
 
Prentzz said:
Also @Ryuzaki do you have any opinion on whether Google in any significant manner alters their link algorithms depending on niche and the link profiles of sites in that niche?
Click to expand...

Ask me tomorrow and I might say something different, but at this moment I'm thinking that they don't need to do that.

If they grab all of the pages that give off the on-page signals that make a page relevant to a query, then look at the backlink profiles of the pages and domains those pages belong to, they're going to naturally have that granularity we're talking about. An automotive query should fetch pages with a lot of forum links, for instance.

I don't think they need to change anything in the way they weight the types of links, because the communities within the niches do it for them.

If they treat the links the same as they would in any other niche, you might find that automotive pages have a lot less contextual links and more forum links. So their average score might be a Page Rank 2 while pages in another niche might average to PR4.

But what's going to happen is that the scores will be normalized, where each page's score will be divided by the score of the highest ranking page. So the top ranking page of a PR4 divided by 4 will be have the best score of 1.00. The 2nd place page of PR3 would have a 0.75 score, and so forth. So it all evens out if they normalize it.

As far as changing thresholds for when you trigger the tripwire, maybe with anchor texts they would, since every community will have different cultures or know more about SEO, etc. But as far as letting certain amounts of spam slide, I think they're able to ignore it now. I don't think they let certain amounts of spam count either, that would muddy up the calculations and doesn't signal quality in any way.

On the flip side, say they do take it to the granular level manually rather than letting it occur naturally. This is where the wisdom of "match and exceed" comes in. Just don't exceed too far and become an outlier, I'd think.
 
You're basically touching on the entire premise of why I invented SERPWoo.

To evaluate what's working per keyword, per niche, per industry... instead of the super regurgitated generic vanilla SEO that worked for 1 website or a generic understanding of all SERPs.

We basically hand you a SERP history for your chosen keyword, like watching the stock market go up and down over time.

Combine those keywords into a project to make a monitoring a niche easy for you. Comb through those results now and come up with what's "working" for your niche or industry.

Dip those results into our new ZORA Audit tool and Monitoring program ( to find real-time links of your competitors if you wanted ), and you end up with stuff like:



I have enough data at my fingertips to see how different algo updates impact different niches. How different content ( length, intent, age ) impacts different niches. How HTML FORMs for one niche helps, but in another, it doesn't do anything. Same with images and breadcrumbs in one niche pushes up rankings, but in another, it won't.

Links are like that too many times.

However, you got to see what the benchmark is and then test your way up to that benchmark.

As an example, I have a series where I rank a brand new domain from scratch to the top 5 of Google for a B2B SaaS term dominated by large companies and I did it based on advice that goes against what most SEOs will tell you, even on this forum sometimes. I did it:

  1. With a new domain, that does not contain my keyword
  2. A slow and very poor theme
  3. No SEO game plan
  4. Less than 7 total pages of content
  5. Of the above 7 pages of content, the total word count is 1000 or less sitewide.
  6. Only used my keywords on the homepage
  7. Did this all in less than 7 months total from start to finish
  8. Almost no backlinks
That series can be gotten to from here.

You just gotta find out your benchmarks, your thresholds, and then start experimenting from there per keyword, per niche, per industry.

Links play a part in that like anything else. Different kinds of content help in different niches, different links help in different niches too.

.
 
Last edited:
You must log in or register to reply here.
Back