Who stole your clicks?

_{Day 222/100}

Hey—It's Tim. 

When links were votes and everyone pretended to be a librarian

Once upon a dial-up, search was quaint.

Larry and Sergey cooked up PageRank, a gloriously nerdy idea: links are citations, citations are votes, and the most-cited pages should rise.

It was elegant, almost wholesome - academic publishing meets the open web. That 1998 paper read like a love letter to math and optimism: crawl, index, count links, return something useful.

And for a hot minute, it worked.

Like with other important things, people realised you could buy votes.

Then came the crack in the teacup:

if links are votes, you can buy votes.
Or swap them.
Or blast them from a thousand blogspot graveyards at 3 a.m.

Keyword stuffing, doorway pages, article spinners - the web’s version of counterfeit money. Google swung back with updates and quality crusades, but the arms race was on.

Even Google’s own guidance had to expand from “don’t be spammy” to naming-and-shaming tactics that shouldn’t exist in a sane universe.

(Expired domain abuse? Site-reputation abuse? Yes, those are real categories now.)

Welcome to the Monetized Maze

(The House Sells the Map)

Here’s where the premise of SEO goes from “optimize for users” to “negotiate with a slot machine.”

Search results turned into an ever-expanding carnival of SERP features - snippets, panels, ‘People also ask,’ and now generative answers - which increasingly answer the query up top and siphon off intent before your site has a chance to say hello.

Independent studies have long argued that a huge share of searches end with zero clicks to the open web. Whether the exact percentage is 36% or 60-something, the direction is obvious: the results page itself is the destination, and your website is the consolation prize.

That’s not a marketplace; that’s a moat.

Receipts, Red Flags, and the Algorithm’s Mood Board

Let’s get uncomfortably specific:

• The original scaffolding: PageRank institutionalized the idea that link structure encodes authority. That’s the DNA of SEO’s obsession with links - still with us, just buried under layers of “it’s complicated.”
  

• Quality raters & E-E-A-T: Google’s own handbook tells human raters to evaluate Experience, Expertise, Authoritativeness, and Trust. It’s not a direct ranking factor, but it telegraphs what the machine is trying to approximate - and what SEOs must theatrically perform.
  

• Spam’s new wardrobe (2024): Google formally cracks down on scaled content abuse, expired domain abuse, and site-reputation abuse - because industrialized sludge and “borrowed authority” became a business model. Translation: the game got exploited so hard (again) they had to invent new rulebooks (again).
  

• AI Overviews (May 2024): Google rolls out AI-generated summaries to hundreds of millions, promising links “prominently included.” In practice, the machine often satisfies the user’s curiosity right there, then sprinkles a few citations like parsley. Traffic anxiety spikes. When those summaries misfired spectacularly, Google publicly dialed scopes and guardrails. The direction of travel, though? Still toward answers above the fold.
  

• Zero-click gravity: Independent analyses (yes, imperfect) keep converging on a bleak thesis: more queries resolved in-SERP, fewer clicks to creators. Even Similarweb’s own marketing notes the SERP is increasingly the endpoint. That’s not a bug of SEO - it’s a feature. It’s the logical endpoint of putting a trillion-dollar ad company in charge of information triage.
  

• The brewing backlash (2025): As AI Overviews ingest the web to re-answer the web, infrastructure giants start shipping licensing and control levers for AI crawlers. Translation: the supply side is arming up. The open-web détente is cracking.

Back to the Fairy Tale We Outgrew

We began with links as earnest little votes.

Then votes became currency, currency bred scams, and the sheriff turned the town square into a kiosk that answers your question before a local can speak.

The “entire premise of SEO” was simple: help people find the best page. The modern premise is uglier: convince the gatekeeper you deserve a cameo in its answer.

If you’re playing this game, play it eyes open - optimize for users, yes, but build direct demand, own your channels, and design for a world where the answer box eats first.

Make them cite you

Because that whole sordid timeline ends in an answer-first web, Penfriend’s mission is simple: build content the machines are forced to quote and humans are eager to trust.

VIBE proves the human layer, Search Score ensures machine coverage, and structured claims/schema get you named in AI Overviews, Perplexity, and ChatGPT.

Zero-click is distribution - so we optimize for citations, mentions, and branded recall, and help you route that attention into demand you own (email, community, product).

While the kiosk talks, Penfriend gets you quoted, tracked, and paid - anti-slop by design, built for LLM visibility, and ruthlessly aligned with the only metric that matters now: being the source everyone else has to cite.

✌️ Tim "PageRank Romantic" Hanson
CMO @Penfriend.ai

Same brain, different platforms: X, Threads, LinkedIn.

P.S. This why naming your processes and being public about them is some important.

Coin a name for your core idea and stamp it in the H1, URL, and first paragraph with one dated fact. LLMs quote names; kiosks paraphrase nouns.

Make them cite you, not “some page.”

Penfriend.ai
Made by content marketers. Used by better ones.