Are We Building Links for Rankings Now or AI Visibility?

Joined
Aug 6, 2026
Messages
5
Reaction score
1
Something I've been pondering recently.

For years, the link-building discussion has basically revolved around the same things:

DA / DR, topical relevance, ref domains, anchor text, guest posts, niche edits, PBNs, etc.

But search is changing pretty fast.

Google no longer just gives us a list of 10 blue links. We’re seeing more and more cases with AI Overviews and AI Mode where the sources are selected after the query is answered.

Which makes me wonder if we’re looking at the value of a backlink thru the wrong lens.

For instance, consider two websites:

Site A
  • DR 70
  • A lot of backlinks
  • Very little organic traffic.
  • Publishes hundreds of guest articles
  • Mostly rankings for its own branded terms
Site B

  • DR 35 35
  • Much smaller link profile
  • Topical relevance is high
  • Authentic organic traffic
  • Useful and original content you can publish
  • Referenced by other websites and communities
I think 5 years ago Site A would be the immediate choice for a lot of people.

I'm not so sure that's still the obvious choice.

I’m wondering if things like Google’s artificial intelligence systems are getting better at finding useful sources rather than returning pages based on traditional ranking signals.
  • topicality
  • real traffic
  • source data
  • author/entity credibility
  • quotes from other relevant sites
  • being talked about in real life conversations
  • depth of report
  • editorial background
could be more significant than the strength of the domain alone.

And that poses another question for link builders:

What Are Backlinks Really Used for in 2026?

Is it still mostly:
Transfer the authority and help this page rank.

Or should we consider it more broadly as:
Help to position this website as a credible source to be found, linked and associated to a topic within search and standard artificial intelligence systems.

I'm particularly interested in people who have actually tested this, as opposed to general SEO theory.

If you have been link building thru 2025/2026, have you changed your site selection?

Are you looking at traffic and topical relevance more than DR/DA now?

And have you observed any difference between links that help improve traditional rankings versus links that appear to boost visibility/citations in AI-driven search?

I'd be interested to hear what people are actually seeing in their own projects.
 
Relevance and real traffic matter more than just DR. We’ve been getting good results from relevant sites, editorial links, and brand mentions. We’ve also noticed better visibility in AI search when the brand is mentioned across trusted sites.
 
Relevance and real traffic matter more than just DR. We’ve been getting good results from relevant sites, editorial links, and brand mentions. We’ve also noticed better visibility in AI search when the brand is mentioned across trusted sites.
Yep, that's right where I'm headed as well. DR can show you how strong a domain looks on paper, but traffic, relevance, and real mentions can tell you so much more about whether the site actually has influence.

What’s interesting is whether these signals are becoming useful not just for traditional rankings, but for how artificial intelligence systems find and validate sources. That’s where I think link building gets a lot more interesting in 2026.
 
Yep, that's right where I'm headed as well. DR can show you how strong a domain looks on paper, but traffic, relevance, and real mentions can tell you so much more about whether the site actually has influence.

What’s interesting is whether these signals are becoming useful not just for traditional rankings, but for how artificial intelligence systems find and validate sources. That’s where I think link building gets a lot more interesting in 2026.
Yeah, I agree. We’ve also started looking beyond DR and checking things like relevance, traffic and where the brand is being mentioned. Some of the smaller but highly relevant sites have given us better results than bigger sites with high DR.
 
both, and thats not a cop out. the pages getting pulled into ai overviews are almost always pages that were already indexed and already had links pointing at them. the citation layer sits on top of the ranking layer, it didnt replace it.

what actually changed is which page you want linked. before, everything got pointed at the commercial page and you hoped. now the page that gets cited is usually the one answering a specific question cleanly, so thats where links need to go.

DR on its own was always a proxy anyway. its just more obvious now, because a strong domain with nothing indexed gets cited by nobody.
 
I'm running mostly GSA SER with Ranker X on top across my own sites, and the thing I keep noticing is that the pages getting pulled into AI Overviews are the ones that were already indexed and already had links pointing at them. The AI systems are not crawling some parallel web. They read what is indexed, and what gets indexed and stays indexed is what has links behind it.

On the site A vs site B example, B is the better buy now, agreed. But the reason isn't that links stopped counting. DR was always a proxy and a poor one. A DR 70 with no traffic and 300 guest posts is a page nobody reads, so nothing ever cites it.

What actually changed for me is where the automation sits. The site and the good properties get relevance and real placements. Everything automated goes behind those, tier 2 and 3 pushing the tier 1 links so they index and stay indexed. Since AI answers get pulled from indexed sources after the fact, indexing is close to the whole game now, and that part is cheap and easy to automate.

Haven't seen a single link that moves citations but not rankings. Same links, two jobs.
 
both, and thats not a cop out. the pages getting pulled into ai overviews are almost always pages that were already indexed and already had links pointing at them. the citation layer sits on top of the ranking layer, it didnt replace it.

what actually changed is which page you want linked. before, everything got pointed at the commercial page and you hoped. now the page that gets cited is usually the one answering a specific question cleanly, so thats where links need to go.

DR on its own was always a proxy anyway. its just more obvious now, because a strong domain with nothing indexed gets cited by nobody.
That's a really good distinction. I think “what page gets the link” is becoming just as important as “what domain gets the link.”

The commercial page may have the authority, but it’s the useful, question-driven page that generally gets the visibility, gets cited, and maybe becomes the source that artificial intelligence systems draw from.

So perhaps the shift isn’t from links to AI citations, but rather from building links for domain strength to building links that strengthen the right pages and entities.

DR still matters but it seems like context is way more important than the number.
 
I'm running mostly GSA SER with Ranker X on top across my own sites, and the thing I keep noticing is that the pages getting pulled into AI Overviews are the ones that were already indexed and already had links pointing at them. The AI systems are not crawling some parallel web. They read what is indexed, and what gets indexed and stays indexed is what has links behind it.

On the site A vs site B example, B is the better buy now, agreed. But the reason isn't that links stopped counting. DR was always a proxy and a poor one. A DR 70 with no traffic and 300 guest posts is a page nobody reads, so nothing ever cites it.

What actually changed for me is where the automation sits. The site and the good properties get relevance and real placements. Everything automated goes behind those, tier 2 and 3 pushing the tier 1 links so they index and stay indexed. Since AI answers get pulled from indexed sources after the fact, indexing is close to the whole game now, and that part is cheap and easy to automate.

Haven't seen a single link that moves citations but not rankings. Same links, two jobs.
The "same links, two jobs" part is probably the most interesting here.
I agree, I don't think you'd separate out AI visibility from traditional SEO as if they're completely separate systems. If a page isn’t indexed, discoverable and already showing some level of authority, there’s not much for an AI system to work with to start with.

What I’m still curious about is what happens after that baseline is established.

Both pages are linkable and indexable, but one is always referenced, the other is not. And that’s where I think relevance, original information, entity signals and the actual answerability of the page to the query may start becoming the bigger differentiators.

So I agree with you that indexing is fundamental. And I'm just wondering if indexed + links gets us in the game, while useful + trusted + specifically relevant determines whether we actually get picked.
 
Back
Top