Link authority is no longer a single-scoreboard game. For years, building backlinks meant one thing: earn enough trust signals to rank higher in Google. That model still holds, but it’s no longer the full picture. Today, users bounce between traditional search results and AI-generated answers, and these two systems don’t evaluate backlinks the same way. A strategy built for one won’t automatically perform in the other.

This matters because your audience is already split. Some people still click through search results and read full pages. Others ask ChatGPT or interact with AI Overviews and never visit a website at all. Both groups represent real visibility, real brand impressions, and real business value. If your link building strategy only accounts for one side, you’re leaving the other exposed to competitors who’ve already adapted.

In this article, I’ll walk through what recent data tells us about whether backlinks matter for AI the same way they matter for search, why those differences exist, and how I approach building a link strategy that covers both systems without doubling the budget.

What the Data Shows About Backlinks in Search vs AI?

Recent studies have started to quantify how backlinks correlate with AI citations. SE Ranking analyzed 129,000 unique domains across over 216,000 pages and found that referring domains were the single strongest predictor of ChatGPT citation likelihood. Their AI Mode research confirmed the pattern: domains with fewer than 300 referring domains averaged about 2.5 citations, while those with over 24,000 reached 6.8.

Ahrefs ran a separate study across 3,311 head terms in ChatGPT, Perplexity, and Google’s top 100 SERPs, focused more on citation overlap than backlink correlation. Together, these findings shaped how I now approach link building for both systems.

Backlink volume and referring domains across platforms

The data showed that top URLs in SERPs and AI Mode carry far more backlinks than those cited by ChatGPT. More than 75% of the 3 million URLs across all three sources sit in the 200 to 400 backlink range for positions one through ten. Position one shows a much wider spread, but the difference between positions two through ten is small.

Referring domain patterns mirror this. A minimum count of unique referring domains is needed to appear in the top ten at all, but adding more doesn’t cleanly translate into higher positions. What this suggests is a threshold effect – you need enough backlinks and referring domains to be in the conversation, but once you’re past that point, other factors start to matter more.

ChatGPT, by contrast, cited URLs that needed far fewer backlinks and referring domains. This was one of the clearest signals that LLMs weigh authority differently than search engines do.

What Domain Rating tells us (and what it doesn’t)

When I looked at domain rating across these same URLs, the pattern was even more spread out. More than 75% of top-ten URLs across all three sources had a DR of 40 or above, with medians sitting somewhere around 60 to 80. But the variation was wide, some lower-DR pages ranked well, and some high-DR pages didn’t rank at all.

What stood out most was the floor. Pages with almost no backlinks or very low authority almost never got picked – not in search, not in AI Mode, and not in ChatGPT. DR is useful as a directional proxy when comparing yourself against competitors in the same niche, but it shouldn’t be treated as a target to chase or a score that guarantees placement.

Summary

The data confirms that a minimum level of link authority is needed to compete in both search and AI, but the relationship between backlinks and position isn’t linear. SERPs and AI Mode reward authority more heavily, while ChatGPT appears to weight content structure and context over raw link metrics.

Why AI and Search Reward Authority Differently?

Understanding the data is one thing, but the real value comes from understanding why these systems behave differently. The answer lies in how each one retrieves and evaluates information.

How Google still uses link signals

Google’s algorithm continues to treat external signals (backlinks, referring domain relationships, and the credibility they carry) as a sorting mechanism. Higher positions in both SERPs and AI Mode still correlate with more backlinks and stronger referring domain profiles. From my experience, this hasn’t changed much in recent years. What has changed is how well Google detects artificial patterns, which means the quality and context of those links matters more than it used to.

When it comes to AI Mode, it’s worth remembering that it’s still a Google product. It pulls from Google’s index and evaluates pages using many of the same ranking signals. That’s why the backlink data for AI Mode looks so similar to traditional SERPs, they share the same foundation.

How LLMs approach authority

ChatGPT and similar LLMs operate on a completely different model. They don’t crawl the web in real time the way Google does, and they don’t evaluate backlinks as part of their retrieval process. Instead, they pull from content that is easy to quote and explain: tutorials, guides, comparisons, and structured data.

What I’ve noticed is that most AI mentions of a brand come from other websites, not from the brand’s own pages. This means unlinked brand mentions, contextual references, and editorial inclusions on third-party sites all carry weight in how LLMs associate your brand with a topic. The link itself matters less than the mention, the surrounding context, and how quotable the content is.

AI citations favor freshness, chunked formatting, and clear structure. If your content reads like a wall of text with no headers or defined sections, it’s less likely to get pulled into an AI-generated answer, regardless of how many backlinks point to it.

Summary

Google still uses link authority as a ranking signal, while LLMs focus more on content structure, context, and brand mentions across the web. A strategy that only builds traditional backlinks may perform well in search but miss the signals that AI systems look for when generating answers.

Building a Backlink Strategy That Covers Both

I run this in stages because it prevents overbuilding and removes guesswork. The idea is to build a credible baseline first, and then split into two parallel paths: one aimed at search rankings and the other aimed at AI citations. Both paths share the same foundation, which keeps the strategy efficient and the budget manageable.

Stage 1: Build the baseline for credibility

The baseline exists to make sure your pages aren’t invisible. At this stage, I prioritize unique referring domains over raw backlink volume. One link per domain is enough. The goal is to create a natural-looking backlink portfolio with clean anchors and a link velocity that mimics organic growth, not a spike from zero to a hundred in a week.

The types of links that work best at this stage, according to any solid SEO guide, include:

• Guest posts and blog placements on topically relevant sites
• Link insertions into existing content where your brand fits naturally
• PR placements that carry dofollow or nofollow links
• Unlinked brand mentions on authority sites in your niche

Content on the linking pages should be editorial quality and varied in length. Not every placement needs to be a 2,000-word feature. Mixing formats and lengths makes the portfolio look natural and gives both Google and AI something meaningful to associate your brand with.

Stage 2A: Building for search rankings

Once the baseline is in place, this stage focuses on closing the referring domain gap between you and the top-performing competitors for your target keywords. I use competitive link analysis to identify domains that link to competitors but not to my client’s site, and then I work to fill those gaps.

What matters most here is topical relevance, page quality, and link placement within the content. A link placed high on the page, surrounded by relevant copy, carries more weight than one buried in a footer or sidebar. I also pay attention to the linking site’s own health — stable traffic, a clean anchor profile, and a positive ratio of incoming to outgoing links.

What I skip: sites with no organic traffic, irrelevant categories, and domains that wouldn’t pass a credibility check. These waste budget and add nothing to the portfolio.

Stage 2B: Building for AI citation

This is where the strategy diverges from traditional link building. For AI visibility, I target the types of pages that LLMs already cite – tutorials, guides, resource lists, and comparison content. The link itself doesn’t need to be dofollow. What matters is that your brand is mentioned in the right context, near the right keywords, on pages that are structured for easy extraction.

Effective AI-focused link placements share a few common traits:

• Content is chunked into clear, skimmable sections
• The page includes data points, statistics, or original findings
• Author schema and timestamps are present in the markup
• The brand mention sits within a helpful, quotable paragraph

I treat these external pages the same way I’d treat my own content. If the page ranks well on its own, the mention gains more reach. If it’s buried with no traffic, the citation opportunity shrinks. The content needs to earn its own visibility for the mention to count.

Summary

The staged approach (baseline first, then split into search-focused and AI-focused paths) lets you cover both systems without doubling your effort. Each stage builds on the one before it, and the same foundation supports both directions.

Which Tactics Don’t Work Across Both Systems?

Not every link building tactic ages well, and some popular shortcuts perform well in one system while producing zero results in the other. I’ve seen this firsthand while working on campaigns that relied on aggressive or manipulative methods.

Why keyword boosting doesn’t reach AI?

Keyword boosting is the practice of sending hundreds or thousands of low-quality signals to your site within a short window, and it can inflate organic traffic numbers temporarily. On the surface, the results look strong. Rankings climb, traffic spikes, and the third-party dashboards tell a convincing story.

But when I looked at clients who had used this approach, their AI citation counts were flat. Zero visibility. In one case, a site gained almost 100,000 backlinks in six months through a keyword boosting service. The referring domains were minimal, the anchor profiles were full of spam, and Google Search Console showed roughly 500,000 impressions but only 51 clicks on the target keyword. The traffic looked impressive in tools, but it carried no real business value (and AI never cited the site once).

The reason is straightforward. LLMs don’t evaluate link graphs the way Google does. They look at what’s being said about your brand, how clearly it’s said, and whether the source is structured in a way that’s easy to extract. Manufactured signals don’t produce the kind of content that gets quoted.

Why parasite SEO falls short?

Parasite SEO is another tactic that I’ve had to clean up after more than once. The approach involves placing content on high-authority third-party domains to borrow their ranking power, and it can work in SERPs, sometimes impressively so. I’ve seen parasite pages go through two separate life cycles, ranking well each time for competitive keywords with strong search volumes.

The problem is that these pages almost never survive without continuous support. The moment the tier 2 backlinks propping them up stop, the rankings collapse. And throughout both life cycles, the pages I reviewed produced zero AI citations. They lacked the fresh, relevant, authority-backed content that AI tools look for when deciding what to cite. Whatever traffic they generated was short-lived, and the brand behind it gained no lasting presence in AI-generated answers.

If you’re in a competitive niche and someone pitches parasite SEO as a growth strategy, think carefully about what you actually get. A temporary SERP position is not the same as building the kind of brand visibility that compounds across both search and AI over time.

How to tell if your strategy is built for only one system

If you’re unsure whether your current link building approach is covering both systems, there are a few warning signs worth checking:

• Backlink volume is growing but referring domain diversity stays flat
• Rankings exist but AI citation counts remain at zero
• Traffic numbers look strong in third-party tools but Google Search Console tells a different story
• Your brand is rarely mentioned on external sites outside of paid or self-placed links

Any one of these patterns suggests that your strategy may be producing signals for search without building the kind of presence that AI systems recognize. The fix isn’t necessarily to scrap what you’re doing — organic SEO specialists usually layer in the contextual, brand-focused placements that LLMs respond to alongside your existing efforts.

Summary

Shortcut tactics like keyword boosting and parasite SEO may produce short-term search results but consistently fail to earn AI citations. When backlink growth doesn’t come with referring domain diversity, genuine brand mentions, and content that AI can actually extract from, the strategy is working for one system at best, and building nothing lasting for the other.

Link Authority Is Not a Single Lever Anymore

Link authority still matters, but it no longer works as a single lever. Search engines and AI systems evaluate backlinks and brand signals through different lenses, and a strategy built for only one side will leave gaps in the other. The staged approach — build a credible baseline, then split your efforts between search-focused link building and AI-focused brand placement — gives you coverage across both systems without overextending your resources.

The brands that will perform best in this environment are the ones that stop treating link building as a volume exercise and start treating it as a positioning tool. Where your brand appears, in what context, and how clearly it’s referenced now matters as much as the link itself.

---

Frequently Asked Questions (FAQ)