Every couple of years someone declares PageRank dead. Then a big site loses half its traffic after a manual link audit, or gains 40 positions after one high-authority editorial pickup, and the conversation quietly drops. PageRank isn’t dead. The way most practitioners measure it is just so far removed from what Google actually computes that it might as well be measuring something else.
The Gap Between Toolbar PageRank and What Google Actually Runs
Google killed the public Toolbar PageRank score in 2016. What died was the exposed proxy — not the underlying algorithm. The internal system still runs. Gary Illyes confirmed at multiple conferences that PageRank (or a descendant of it) remains one of the most significant signals in Google’s ranking stack. The problem is that without a live public score, practitioners defaulted to third-party proxies: Ahrefs DR, Moz DA, Semrush Authority Score. These correlate with rankings. They are not PageRank. They model link graphs differently, treat link loss differently, and none of them have access to Google’s actual crawl index.
Optimizing for DR instead of actual link equity flow is how you end up with a high-DR domain that passes almost nothing to its target pages. Common problem. Fixable once you understand the mechanism.
What Actually Blocks Link Equity From Flowing
PageRank distributes equity through followed links, divided across all outbound links on a page. In practice, four things kill the flow before it reaches your page:
1. Nofollow at the Link Level
rel="nofollow", rel="ugc", and rel="sponsored" all tell Google not to pass PageRank. Google treats these as hints now, not directives — so some equity may bleed through on editorial nofollows from very high-trust domains. Don’t count on it. Build for followed links.
2. Link Dilution on High-Outlink Pages
A link from a page with 200 outbound links passes roughly 1/200th of that page’s equity, minus the damping factor (approximately 0.85 in the original paper; Google’s version is likely in that range). A link from a focused editorial page with 8 outbound links passes substantially more. This is why a link from a Wikipedia disambiguation page and a link from a tightly written feature article on the same domain are not equivalent, even when third-party tools assign them identical DR scores.
3. Crawl Depth and Indexation Gaps
If the linking page isn’t indexed, or is indexed but rarely recrawled, the equity it theoretically carries may not factor into Google’s live graph when your page is being evaluated. We see this with links from large sites where the linking page sits five or six levels deep and gets crawled infrequently. The link exists. The equity effectively doesn’t — not yet, possibly not ever at meaningful amplitude.
4. Redirect Chains
301 redirects pass PageRank, but each hop leaks some. A single clean 301 is fine. Three hops — an old link that hits a migrated URL, which was restructured, which 301s to your current URL — and you’ve lost real equity at each step. Auditing redirect chains on your highest-value inbound links is one of the least glamorous things you can do. It’s also one of the highest-ROI.
Internal Link Equity Is Undervalued. Seriously.
Most link building discussion fixates on external acquisition. Fair — external links are harder to get. But internal links pass PageRank too, and you control 100% of that flow. If you’ve acquired strong external links landing on your homepage or a high-traffic category page, you’re sitting on distributable equity that most sites leave stranded.
The pattern we see repeatedly: a site earns a handful of strong editorial links to a top-level URL. That URL accumulates real PageRank. But the pages that actually need to rank — product pages, specific guides, conversion-critical content — are linked from the homepage via one small nav link alongside 40 other items. The nav dilution kills most of what was earned externally.
Better approach: identify which external-facing pages have the strongest inbound link profiles, then build a deliberate internal link pathway from those pages to the URLs that need ranking support. One contextual in-content link from a high-equity page beats ten footer links. Not a hack — just how the math works.
Entity Associations Change How Equity Is Weighted
This is where PageRank intersects with something that wasn’t in the original 1998 paper: topic and entity relevance. Google’s systems now evaluate not just how much PageRank a link passes but how relevant the linking entity is to the receiving entity. A link from a high-authority domain in an unrelated vertical passes less contextual weight than a link from a mid-authority domain strongly associated with your topic cluster.
Pure authority-chasing — going after the highest DR links regardless of topic — has diminishing returns for this reason. A DR 80 link from an unrelated niche may move less than a DR 60 link from a publication Google has heavily associated with your subject area. I can’t give you a precise formula without access to Google’s internals, and I won’t pretend otherwise. But the directional signal from ranking outcomes is consistent enough to change how you prioritize outreach.
How to Actually Measure Link Equity Flow (Without Toolbar PageRank)
You can’t get exact PageRank scores. You can proxy the signal reasonably well by combining a few data sources:
- Google Search Console crawl data — pages crawled more frequently tend to have more real equity pointed at them. Cross-reference with your internal link graph.
- Index coverage of linking pages — before crediting a link, verify the linking URL is indexed. Use the
site:operator or crawl against a fresh Googlebot user-agent. - Outbound link count on linking pages — raw number matters for dilution math. Pull it during prospecting, not after you’ve already landed the link.
- Redirect chain depth on inbound links — Screaming Frog or a custom crawl against your inbound link list will surface chains you didn’t know existed.
None of these individually gives you PageRank. Together they let you reason about equity flow well enough to make better decisions than DR alone allows.
One Genuine Caveat
PageRank is one signal. It interacts with content quality, entity disambiguation, user signals, freshness. A site with weak content but strong link equity can rank — but that position is fragile; a helpful content or core update can strip those rankings in a single cycle. The sites we see hold rankings through algorithm updates tend to have both real link equity and content that actually answers what the query implies. Equity alone buys you less than it did in 2015.
Practical priority: fix your equity flow problems first, because they’re often structural and recoverable quickly. Then make sure the pages receiving that equity deserve to rank on content merit. Fixing the plumbing on a page that doesn’t deserve to rank is just faster delivery of a bad result.
Pull the top 20 inbound links to your domain right now and check how many linking pages have more than 100 outbound links. If the answer is most of them, your effective equity intake is significantly lower than your DR suggests — and that’s the gap worth closing before you commission another link.