TL;DR - WTF is PageRank Anyway?
PageRank is Google's original method for measuring authority by treating links as votes, weighted by the importance of whoever's doing the voting. I've been building sites in vape and cannabis, one of the most link-hostile industries on the internet, for fifteen years now, and PageRank has quietly decided the outcome of almost every ranking battle I've fought in that time.
Contrary to popular belief, PageRank never died. It evolved into something called PageRank-NearestSeeds, which cares less about how many links you have and more about how close those links sit to a small set of trusted "seed" sites.
A domain's homepage carries its own PageRank score, and that score gets inherited by every other page on the site, which is the real reason new vape and cannabis websites sit flat for months no matter how good the content is.
Internal links distribute that same authority around your own site, and most vape ecommerce catalogues waste huge amounts of it through sloppy architecture.
Most people think PageRank is dead. It isn't.
I hear this constantly, usually from someone quoting a five-year-old blog post that quoted an even older Google statement.
Here's the thing nobody in this argument seems to say out loud: Google doesn't rip out its foundational systems and start again. It layers new signals on top of old ones. NavBoost, user engagement data, entity recognition, all of it got built on top of the link graph, not instead of it.
This actually got confirmed properly in 2024, when Google's internal Content Warehouse documentation leaked. It showed that the concept is still very much alive inside Google's ranking systems, and that a site's homepage carries a PageRank score that gets associated with every single document published on that domain.
Shaun Anderson, whose research I lean on heavily for this kind of thing, has a name for that mechanism: Homepage PageRank, sometimes called Homepage Trust. That's Anderson's own interpretive label for a pattern he identified across the leaked documentation, not a literal field name lifted from Google's code.
The mechanism is real and well-evidenced, per the leak. The name for it is his analysis, and I credit it as such.
Why does Google run things this way? Because pure content scoring has no defence against volume. You can publish a thousand pages overnight. You cannot fake years of earned, editorial trust overnight, not without leaving a trail that's easy to spot.
The plain-English version, for anyone who's never touched a ranking factor in their life
Think of PageRank as compound interest for trust. A new site starts with almost nothing in the account. Every quality link is a deposit. Over time those deposits earn their own interest, because a page that's trusted enough to be linked to by other trusted pages becomes more valuable itself, and passes that value onward.
Miss the early deposits and you're not just behind, you're behind on interest you'll never get back.
Entity SEO, which I'll come back to later, works the same way as a business reputation. You don't build a reputation by shouting about yourself. You build it by being consistently, verifiably the same thing, in the same niche, over a long period, until other credible parties start vouching for you without being asked.
The sandbox isn't a punishment. It's arithmetic.
- What most people believe: new sites get sandboxed, a deliberate probation period Google runs to test whether you're serious.
- Why they're wrong: there's no evidence of an intentional holding pen anywhere in the leaked documentation. What looks like a punishment is a straightforward consequence of a scoring model.
- What's actually happening: a brand-new domain starts with a homepage that carries effectively zero PageRank, and every page you publish on it inherits that same low baseline until the homepage itself earns enough trust through quality links.
Without some kind of baseline trust signal, Google has no reliable way to tell a legitimate new business apart from a spam domain that went live an hour ago. Homepage PageRank acts as a stand-in score for new pages until they've had time to accrue their own signals.
If you're launching a new vape or cannabis site, your first move isn't more content. It's homepage authority. This is the exact question we start with in a brand audit, because there's no point optimising twenty product pages if the domain underneath them has no trust to lend.
Not all links are equal, and restricted markets make this brutally obvious
- What most people believe: link building is a volume game. More links, more PageRank, better rankings. Simple.
- Why they're wrong: that reading was crude even for the original algorithm, and the modern version makes it actively wrong.
- What's actually happening: the production version running inside Google today is PageRank-NearestSeeds, abbreviated pagerank_ns in the leaked documentation. It scores a page not by how many links point at it, but by how close it sits, within the link graph, to a small set of trusted seed sites, likely major universities, government domains, and top-tier news publishers.
Proximity to genuinely trusted anchors in the web graph is far harder, because you'd need those trusted sites, or sites close to them, to vouch for you specifically.
Every extra hop a link takes to reach a trusted seed quietly taxes its value. In most industries that discount is annoying. In vape and cannabis, where mainstream press coverage and .edu or .gov citations are genuinely hard to earn, that discount compounds against you far harder than it does against, say, a SaaS company that can get a TechCrunch mention with one good funding announcement.
- Real-world example: a single link from a site that's itself trusted and closely connected to a major news outlet has done more for clients I've worked with than dozens of links from disconnected, low-authority directories, even when the directory links looked fine on paper metrics.
Stop auditing your backlink profile by volume or domain rating alone. Audit it by distance. This is exactly why link acquisition sits inside our fractional SEO lead engagements rather than being treated as a bolt-on task, because closing that distance in a restricted category takes deliberate strategy, not a shopping list.
Internal links are road signs, not garden hoses
- What most people believe: PageRank is something you earn from other websites. Internal links are just navigation, a courtesy for users.
- Why they're wrong: the leaked documentation confirms Google maps how authority moves inside your own site through an attribute called onsiteProminence, which measures how important a page is within its own domain by simulating traffic flow from the homepage and other high-traffic pages.
Anderson uses a heat and laser metaphor for this that I think about on every single site audit I run. Picture your strongest pages, homepage, best-performing guide, as heat sources. Internal links aren't hoses that spray that heat evenly. They're lasers. Point several of them at one page and you concentrate real energy on it. Spread the same links across a two hundred item footer menu and every page gets lukewarm at best.
- Your internal architecture is a direct signal of editorial judgement.
- A clean structure that funnels authority towards genuinely important pages reads as a maintained, trustworthy resource.
- A flat site where everything links to everything reads as neglect.
I still see vape ecommerce catalogues where thirty near-identical flavour pages, "Blue Raspberry 20mg," "Blue Raspberry 10mg," "Blue Raspberry Ice," all link to the same generic footer and nothing else.
- No links up to a buying guide.
- No links across to genuinely related products.
- It's the equivalent of building a power station and running one thin wire to an entire county.
Work out which pages actually need to rank, then deliberately build internal pathways towards them from your most prominent existing content. That discipline is the backbone of the topic clusters we build inside our content marketing services work, because a topic cluster without deliberate internal linking is just a folder of unrelated pages.
The company you keep still matters
Modern PageRank is both an endorsement of your brand but it also works the other way too which means you need to be careful who you link out to as well. The leaked spamrank attribute measures how likely your page is to link out to known low-quality or spammy domains, and consistently vouching for bad neighbourhoods on the web starts to cost you your own credibility.
I bring this up specifically for vape and cannabis publishers because "further reading" and affiliate sections are where this quietly goes wrong.
A lazy resource list, built once and never revisited, can end up pointing at grey-market retailers or defunct affiliate partners with nobody noticing until rankings slip for reasons that look mysterious from the outside. Audit outbound links on the same schedule you audit inbound ones.
Where this bites hardest in a real vape ecommerce catalogue
Everything above is fine in principle. It becomes useful the moment you can point at a real product catalogue and say, that, right there, is the mechanism causing your problem.
The flavour cannibalisation problem
Split thirty flavour variants of the same device across thirty thin, near-duplicate URLs, and you're not just diluting internal PageRank through the laser problem above.
You're also confusing Google's semantic understanding of the site, because a compressed vector representation of your site's content, built from everything you publish, gets blurrier the more near-duplicate pages you add rather than sharper.
I've flattened this exact pattern more than once, replacing a sprawling flavour catalogue with a single strong category page and letting only the genuinely high-demand variants keep dedicated URLs. The category page consistently outranks the twenty thin pages it replaced.
The terminology problem restricted markets don't talk about enough
Vape and cannabis terminology is riddled with words that mean something else entirely outside the industry. Juice. Mod. Cloud. Flower. Wax. Shatter.
Google has to resolve which meaning applies through entity recognition, effectively deciding whether your "juice" page is about e-liquid or orange juice, in the same way it has to work out whether "mod" means a vape device or a car part.
I think of this as a specific, sharper version of Anderson’s Disconnected Entity Problem whereby your own industry language, used without enough surrounding context, becomes a liability rather than an asset, because the machine reading it has no reliable way to disambiguate what you mean.
Reinforce every ambiguous term with clear definitions, structured product data, and consistent internal linking between related entities, or Google will resolve the ambiguity using noisier signals from elsewhere on the web, ones you don't control.
The Authority-Relevance Gap
This is the framework I keep coming back to, because it explains almost every stuck client site I've ever inherited.
The Authority-Relevance Gap is the distance between how much trust a page has earned, its PageRank, its domain authority, its distance from trusted seed sites, and how clearly Google's systems understand what that page is actually about, its entity clarity, its topical consistency with the rest of the domain.
Most restricted-market sites fail on one side of that gap or the other.
- Either they've built up real authority over years but publish content so scattered or ambiguously worded that Google struggles to match it confidently to search queries.
- Or they're brand new and painfully clear about their topic, but Google simply doesn't trust them yet, no matter how well-structured the content is.
Closing that gap on both sides at once, rather than treating link building and content as two separate workstreams run by two separate teams who never talk to each other, is the single most reliable lever I've found in fifteen years of restricted-market SEO. It's the entire premise behind running fractional SEO lead and content strategy together rather than apart, and it's the kind of pattern that only shows up once you've seen enough case studies to notice it repeating.
Practical takeaways
- Homepage authority sets the baseline for every page on your domain. Fix that before optimising individual pages.
- Link distance to trusted seed sites matters more than raw link volume, and it matters more in vape and cannabis than almost any other industry.
- Internal links are a finite resource. Point them deliberately at the pages that actually need to rank.
- Audit your outbound links as carefully as your inbound ones. Guilt by association is a real signal.
- Industry-specific terminology needs extra context to avoid being misread by Google's entity systems.
- Authority and relevance have to be built together. One without the other leaves you stuck.
FAQ
Is PageRank still used by Google today? Yes. Google's own leaked documentation confirms the concept remains active, now running as an evolved model called PageRank-NearestSeeds rather than the original public algorithm.
Why do new vape and cannabis websites struggle to rank even with good content? Because a new domain's homepage starts with almost no PageRank, and every page on that domain inherits that low baseline until the homepage itself earns enough trust through quality links.
Does internal linking really affect rankings? Yes. Google measures how authority flows within a site through an attribute called onsiteProminence, meaning a page with no internal links pointing to it will struggle regardless of how good the content is.
Can outbound links hurt my rankings? They can. A signal called spamrank treats outbound links to low-quality domains as a mark against your own site's trustworthiness, so outbound link hygiene matters as much as inbound link acquisition.
What's the single most efficient way to speed up how fast new content ranks? Invest in a small number of genuinely earned links to your homepage rather than spreading effort across individual pages. That lifts the baseline authority for everything you publish afterwards, present and future.
If you want to know exactly where your own domain's authority currently stands before deciding what to fix next, my brand consultation audit gives you a systematic overview of where your business at right now.
It’s great for anyone that wants or needs a second pair of eyes on their current strategy. Most of my brand consults, a good 75% of them, anyway, are from brands that are looking for reassurance about their current SEO strategy.
Because if you’re paying an agency or in-house SEO to do work for you and you suspect something isn’t quite right, the smart move is to get a second opinion, and that’s what my brand consultation audit is all about.