Chapter 34 — Key Takeaways

A one-page reference. If you remember nothing else from this chapter, remember that a publisher's traffic comes through doors it does not own — and that the only real insurance is to own a door of your own.

The one idea

Publisher SEO is the whole book at maximum stakes, because a pure content publisher has the least to fall back on when Google changes its mind. Build compounding evergreen depth, earn genuine author trust, keep the archive clean, and own an audience Google doesn't mediate — so that when the algorithm moves (and it will), you wobble instead of fall.

The three publisher doors

Surface Driven by Can you target it? Treat it as
Search / Top Stories a query (newsworthy) + trust + freshness Yes — intent, quality, speed Your controllable base
Google News automatic inclusion + news policies No submission/fee; manage presentation only A door that's already open
Google Discover inferred interest — no query No — eligibility, not entry; no keyword Upside, never a baseline

Rules of thumb

  • You don't submit or pay for Google News. Inclusion is automatic for genuine news that follows the policies; the Publisher Center only manages presentation. Anyone selling "News inclusion" is selling an open door.
  • There is no keyword in Discover. You become eligible (indexed + policies + a great image + real trust) and hope to be matched. Large images (~1,200px, max-image-preview:large) help; nothing guarantees.
  • Compelling but never clickbait. Discover rewards curiosity and penalizes the curiosity gap.
  • Run a portfolio, not a monoculture. Breaking spikes and dies (half-life: hours–days); evergreen compounds (half-life: years, if updated). Analysis lives in between and can age into evergreen.
  • A bloated archive is a site-level liability. Prune/merge dead thin pages and update the evergreen ones (Chapter 12) — publishing less can raise the whole site's traffic.
  • Schema and Web Stories are not ranking tricks. NewsArticle schema aids understanding/eligibility (keep datePublished/dateModified truthful); Web Stories are a supported format, not a cheat. AMP lost its Top Stories privilege in 2021.
  • Paywall without cloaking. Use flexible sampling (metering or lead-in) plus paywall structured data (isAccessibleForFree: false) — declare the gate, don't hide it.
  • Syndication trades control for reach. A cross-domain canonical is a hint a stronger domain can override; the reliable protection is the republisher's noindex. Decide whether you're buying reach or rankings before you sign.

The mental reframe

  • "Google giveth and taketh" is structural, not personal. Google optimizes for searchers, not for your revenue, and changes its systems hundreds of times a year. That's the job, not hostility — which is exactly why diversification is obvious, not paranoid.
  • "Nothing to fix" is a diagnosis. After a core-update reassessment there's often no bug — the verdict is on your content's quality at the site level. The remedy is slow (better content, a cleaner archive, real trust), so start it before the drop.
  • A windfall is not a foundation. Bank every Discover spike into something you own — an email subscriber, a returning reader — and forecast the business on channels you can influence.

What you can do today

Open Search Console and check for a Discover report and its shape (steady vs. spiky). Sort ten of your pages into breaking / analysis / evergreen and name your best evergreen asset and one decayed page. Then answer the only question that ultimately matters: if Google halved your organic traffic next month, what channel would keep the business alive? If the answer is "nothing," that — not any single tactic — is your most important project.

Where this goes next

Chapter 35 takes on YMYL SEO (health, finance, legal), where WellPath is the home case and Google raises its trust bar the highest — the full treatment of the credentialed-author, review-process, never-fake-authority ideas this chapter only previewed.