Case Study 2 — The Roundup Farm: Right Format, No Authority, No Experience

A complementary angle. Case Study 1 examined a real, documented Google change to the featured-snippet format. This one examines the limit the whole chapter builds toward (§10.7): a site can get the format exactly right and still collapse, because format is Gate 1 and authority plus genuine experience is Gate 2.

This is a labeled composite — not one named company, but a pattern that has played out publicly and repeatedly, most visibly around Google's Helpful Content Update (launched August 2022) and its reviews system updates. The updates and Google's published guidance are real and Tier-1 (Chapter 6 covers them in depth). The company and every specific number below are illustrative; the mechanism is what is real.

Background: a business plan made of formats

A small team spots what looks like an easy opportunity. Comparison and "best of" content — the commercial-investigation formats from §10.6 — sit right before the purchase, so they attract high-value searchers and, when monetized with affiliate links, pay well per visit. The team reasons: these are the formats that win commercial queries; let's produce them at scale.

So they build "the Roundup Farm." Over a year they publish thousands of pages, every one in a format that matches its query's intent perfectly:

  • "Best [product] for [use case]" roundups — hundreds of them.
  • "[Product A] vs [Product B]" comparison pages with tidy tables.
  • "Top 10 [category] of 2026" listicles, refreshed with the year in the title.

The formats are correct. The pages have comparison tables (eligible for table snippets, §10.3), FAQ sections mirroring the People Also Ask box (§10.4), clean headings, and a keyword in every title. On a format checklist, the Roundup Farm scores nearly perfect. And for a while, it works: the pages are new, the niche had thin competition, and many rank on page one. Traffic and affiliate revenue climb. The team concludes the strategy is validated and pours in more of the same.

There is just one thing every page lacks: anyone who has actually used the products.

The issue: Gate 1 passed, Gate 2 hollow

Walk the Roundup Farm through the two-gate model from §10.7.

  • Gate 1 — intent/format match: passed, cleanly. Every page is the right shape for its query. That is exactly why the pages were eligible to rank at all, and why they did rank while competition was thin.
  • Gate 2 — the ranking contest (quality, experience, authority): hollow. The comparisons are assembled from other people's specs and reviews, not from first-hand testing. There is no author who has held the products, no original photos, no measurements, no "we tried this and here's what the spec sheet doesn't tell you." The site has few earned links because nothing on it is worth citing — a rewritten spec table is not a linkable asset (§10.5). Its authority is borrowed, and thin.

For a window, Gate 1 alone was enough, because Gate 2 is only decisive when there is real competition and when Google's quality systems catch up. Both caught up.

FIGURE C2.1 — "Right format, hollow center: the arc"              [constructed teaching example]
  PHASE            WHAT HAPPENS                                   WHY
  Build            1,000s of "best X" / "X vs Y" pages, all in    Formats match intent (Gate 1);
                   the correct format                              thin niche, little competition
  Early ranking    Many pages hit page one; traffic + affiliate   Gate 1 is enough when no strong,
                   revenue climb                                   experienced competitor is present
  Reassessment     A Helpful Content / reviews update rolls out;  Google's systems reward first-hand
                   rankings and traffic fall broadly, site-wide    experience + helpfulness; "search-first,"
                                                                   experience-free content is demoted
  Aftermath        Traffic doesn't "bounce back" next week; the   The signal is about the SITE's overall
                   whole site is discounted, not one page          helpfulness, not one bad page
  (All figures/timeline illustrative; the SHAPE — a broad, durable demotion of experience-free content
   at scale — is the real, repeatedly-documented pattern.)

What it shows

The Roundup Farm is §10.7 dramatized, plus a warning the chapter only gestured at:

  1. Format is necessary, not sufficient — and its sufficiency is temporary. The right format got the pages eligible and even briefly winning, in a thin niche. It could not make them durable, because it never addressed why Google (or a reader) should trust this page over a genuinely experienced one. When competition and quality systems arrive, format alone folds.
  2. Google's target is precisely "right format, no substance." The Helpful Content system was built to find content made primarily to rank rather than to genuinely help — and mass-produced roundups with no first-hand experience are close to its archetype. The reviews system goes further, explicitly rewarding reviews that show evidence of actually using the product. A comparison written by someone who never touched the products is exactly what these systems demote.
  3. Experience is the moat a service business has and a farm doesn't. This is the hopeful mirror of the story, and Rivertown's whole advantage. A technician who installs both tankless and tank water heaters every week can write the "tankless vs tank" comparison the Roundup Farm never could — the honest one, with the hidden costs and the "in an old house, choose this" judgment. That is the "Experience" of E-E-A-T (Chapter 5), and it is the difference between a comparison that lasts and one that evaporates in an update.
  4. "It worked for months" is a trap. Early success in a thin niche validated the wrong lesson. The team scaled because it was ranking, which meant a larger site to demote when the reassessment came. Format metrics ("we cover every query, every page is optimized") looked great right up until they didn't.

Outcome

In the version that ends in learning, the team stops treating format as the strategy and starts treating it as the packaging. They cut the thousands of hollow pages down to a much smaller set they can genuinely stand behind — categories where someone on the team has real, demonstrable experience — add first-hand testing, original photos and data, and named authors, and slowly rebuild trust and links. Recovery from a site-wide quality demotion is slow and uncertain (Chapter 6 is honest about this: there is nothing to "fix" in a day, and reassessment takes time). In the version that doesn't end in learning, the team blames the format, tries a different format at the same hollow scale, and repeats the collapse.

The lesson

A great format cannot overcome weak authority, and format at scale with no experience is exactly what Google's quality systems are built to catch. The Roundup Farm did the thing this chapter teaches — match the format to the intent — and only that thing, and it wasn't enough, because it was never the whole job. The durable move is the one §10.7 and the Strategy File prescribe: aim correct formats at queries you can genuinely win, back them with real experience and earned authority, and be willing to build fewer pages that deserve to rank rather than a thousand that merely fit.

Connect it to Rivertown. The Roundup Farm is the anti-Rivertown. Rivertown's format plan (Figure 10.3) deliberately skips "best furnace brands" — the roundup query it could format but not authoritatively win — and aims its comparisons at "tankless vs tank" and "repair or replace," where its technicians' first-hand experience is a moat no farm can copy. Same formats; opposite outcome, because one has a hollow center and the other has a real one.

Discussion questions

  1. The Roundup Farm passed Gate 1 (format) and briefly ranked. Explain precisely why that early success was evidence of nothing about the strategy's durability — and what a more honest early metric would have been.
  2. Google's reviews system rewards "evidence of first-hand experience." List five concrete things a genuine reviewer includes that a spec-rewriting farm cannot fake. How many of them could a service business like Rivertown supply for a "tankless vs tank" page?
  3. This case is a labeled composite, not a named company. Why does the book prefer a labeled composite here over naming a specific site and quoting its traffic drop? (Tie your answer to the citation-honesty rules.)
  4. A stakeholder says, "So formats don't matter — only authority does." Correct them: state the exact role format does play, using the two-gate model, without overselling or dismissing it.
  5. The team's fatal move was to scale because it was ranking. Describe a governance rule (a decision the team could have written down in advance) that would have prevented mass-producing hollow pages — and tie it to the "publish less, but better" idea developed in Chapter 12.