Completions

For multi-location SEO + content + technical SEO leadership

Four thousand location pages published last week. The next Google spam update could deindex all of them.

Google formalized the scaled-content-abuse policy in March 2024. The Helpful Content System has been dampening rankings since 2022. The doorway-page penalty has existed since 2015. Surfer SEO, Scalenut, NeuronWriter, ContentatScale, Clearscope ship content generation. The pre-publish content-distinctness gate that scores every generated page against the operator corpus + LLM-as-judge substantive-content evaluation + Google scaled-content-abuse heuristics before publish is operator-side architecture.

By Jay Christopher11 min read

What this gets you

  • Pre-publish content-distinctness gate— every generated page evaluates against corpus distinctness + LLM-as-judge substantive- content score + Google scaled-content-abuse heuristics before publish. Pages exceeding threshold publish; pages failing route to revision queue.
  • Failure-mode-matched revision workflow — low corpus distinctness routes to template- pattern revision. Thin substantive content routes to content-enrichment workflow. Scaled-content- abuse pattern routes to template-redesign workflow. Per-template revision propagates the fix to every page generated from the template.
  • 3-stage multi-location content-quality pipeline integration — pre-publish gate (this skill) + publish substantive content (neighborhood-FAQ-authoring + per-location-content-generation) + post-publish cannibalization detection (local-cannibalization- defense). Three stages cover the full content- quality lifecycle.
  • Per-template + per-location coverage measurement — per-template gate pass rate over time surfaces which templates produce the patterns Google dampens. Per-location pass rate surfaces per-location revision needs. Trend lines drive template-revision prioritization.
  • Google policy alignment with regulator-grade audit trail — gate decisions align with Google scaled- content-abuse policy + Helpful Content System + doorway-page guideline. Audit trail captures every gate decision per page per template per cycle for retrospective + recovery-from-manual-action evidence.

The scaled-content-abuse policy was formalized in March 2024

A multi-location operator built a programmatic-SEO pipeline in 2022 to capture long-tail per-location traffic. The pipeline generates per-location landing pages from a template that combines location name + service category + a paragraph of boilerplate text + a few per-location placeholders. The operator published 4,000 pages across 200 locations and 20 service categories. The pages indexed. Organic traffic to per-location pages grew steadily through 2023.

Google updated the Helpful Content System in 2023 + rolled out the formalized scaled-content-abuse policy in March 2024. The operator did not change anything in the pipeline. The pages kept generating. The operator content team monitored standard SEO metrics + did not see immediate impact. Six months later organic traffic to per-location pages dropped 35 percent over a single 60-day window. The corporate SEO director investigated.

The investigation revealed Google had algorithmically dampened a large fraction of the operator per-location pages under the scaled-content-abuse + Helpful Content System policies. The pages were near- duplicate. The template scaffolding dominated. The per-location content was sparse. Google dampened + the operator team did not know which pages were affected because Google does not surface affected- page lists. The team faced a multi-month manual audit + revision workflow to recover the affected pages.

The pre-publish content-distinctness gate prevents this pattern at the pipeline level. Every page evaluates before publish. Corpus distinctness scores the page against every other page in the same category. LLM-as-judge evaluates substantive content. Google scaled-content-abuse heuristics pattern-match for the boilerplate-heavy + thin-content + doorway- shape signatures. Pages exceeding distinctness threshold publish. Pages failing route to revision queue with failure-mode-matched workflow. The team catches the pattern at gate before it accumulates in the catalog rather than discovering it through organic-traffic drop after Google enforces.

The gate also handles the recovery scenario for operators with existing affected pages. The audit trail surfaces which pages historically failed gate criteria + which template patterns produce the failure mode. Per-template revision propagates to every page generated from the template. The recovery runs at template-revision pace rather than per-page manual revision pace.

What is in market — and what each category leaves to you

The content-generation primitive is mature. The pre- publish content-distinctness gate integrated with the page-generation pipeline at multi-location scale is operator-side architecture.

Content generation + on-page SEO — Surfer SEO, Scalenut, NeuronWriter, ContentatScale, Clearscope, Frase

Excellent at AI-assisted content generation + keyword research + content briefs + on-page SEO recommendations + content quality scoring. The pre-publish content-distinctness gate that evaluates every page against the broader operator corpus to detect near-duplicate scaled-content patterns + LLM-as-judge substantive-content evaluation + Google scaled-content-abuse pattern matching + the 3-stage pipeline integration are operator-side architecture above the content- authoring primitive.

AI content detection — Originality.ai, GPTZero, Writer, Sapling, Copyleaks, ZeroGPT, Turnitin

Strong at AI-written content detection through perplexity scoring + LLM-fingerprint detection. The content-distinctness gate uses AI-detection primitives as one of multiple inputs (Google does not penalize AI-written content per se; it penalizes thin + duplicate + doorway content regardless of authorship). The broader distinctness evaluation goes beyond AI-detection into substantive-content + corpus-distinctness layers.

Programmatic SEO platforms — Webflow programmatic + Framer GPT + custom Next.js/Astro pipelines

Strong at the page-generation primitive with template + data + publish workflow. The pre-publish gate sits between page-generation and page-publish in the pipeline. The gate is an operator-side build that the programmatic-SEO platform calls as part of the publish workflow.

Google Search Central documentation + Google Search Console manual-action surface

Google publishes scaled-content-abuse policy + Helpful Content System guidance + doorway-page policy. Google Search Console surfaces manual actions after enforcement. The pre-publish gate prevents the pattern before enforcement; the manual-action surface is the after-the-fact signal the gate exists to avoid.

The 4,000-page programmatic SEO catalog the director discovers Google dampened during quarterly traffic review

The status quo at most operators running programmatic SEO at scale. The pipeline ships + the pages index + Google updates + the dampening happens silently + the director discovers the 35-percent traffic drop during quarterly review + the multi-month recovery workflow begins. The pre-publish gate prevents the pattern from accumulating in the catalog upstream of the enforcement.

The pipeline, end to end

  1. Position in the 3-stage multi-location content- quality pipeline. Pre-publish gate (this skill loop 27 content- distinctness-gate) + Publish substantive content (loop 9 neighborhood-FAQ-authoring + per-location- content-generation) + Post-publish cannibalization detection (loop 18 local-cannibalization-defense). Per-location-page-generator agent owns the pre- publish and publish stages; downstream detection sits in adjacent agents.
  2. Corpus distinctness scoring. Every generated page evaluates against every other operator page in the same category via token-overlap + embedding-similarity scoring. Pages exceeding token-overlap threshold against the corpus flag as near-duplicate. The threshold tunes per category + per operator content history.
  3. LLM-as-judge substantive-content evaluation. Each page evaluates against an LLM-as-judge prompt that scores whether the page provides unique value beyond template scaffolding. The judge evaluates per- page specific information density + per-location context depth + per-location case examples + neighbor- hood-specific references.
  4. Google scaled-content-abuse heuristic matching. Pattern matching for the signatures Google scaled- content-abuse policy targets — boilerplate- heavy text + thin content + doorway-shape with city- swap template + repetitive section structures + lack of unique value beyond template. The heuristics tune per Google policy updates.
  5. Gate decision per page. Pages exceeding all three thresholds (corpus distinctness + substantive content + Google heuristics) publish. Pages failing one or more route to revision queue with failure-mode signal surfaced. The decision logs to the audit trail.
  6. Failure-mode-matched revision workflow. Low corpus distinctness routes to template-pattern revision (reduce template + increase per-location- specific content). Thin substantive content routes to content-enrichment workflow (add per-location case examples + neighborhood context + local-event references). Scaled-content-abuse pattern routes to template-redesign workflow (the underlying template approach is producing pages Google will dampen).
  7. Per-template revision propagation. Per-template revision propagates the fix to every page generated from the template. The recovery runs at template-revision pace (one template revision fixes 500 pages) rather than per-page manual revision pace (500 manual revisions). Template revision version-controls with effective date for rollback.
  8. Per-template + per-location coverage measurement. Per-template gate pass rate over time surfaces which templates produce the patterns Google dampens. Per- location pass rate surfaces per-location revision needs. Trend lines drive template-revision prioritization. The signal feeds template-revision cadence per cycle.
  9. Existing-catalog audit for recovery scenarios. Operators with existing programmatic-SEO catalogs run the gate retroactively against the published catalog. Pages that fail current criteria flag for revision. Audit trail surfaces which template patterns produce the failure modes. Template revision propagates the recovery.
  10. Google policy update tracking. Google scaled-content-abuse policy + Helpful Content System guidance + doorway-page policy update over time. The heuristic library tracks Google policy announcements + ingests updates into the gate evaluation. Pages already published may flag as affected under updated criteria.
  11. Manual-action recovery workflow. Operators with existing manual actions run gate retroactively + identify affected pages + revise via failure-mode-matched workflow + submit reconsideration request to Google with the audit trail as evidence of corrective action.
  12. Audit trail per gate decision. Every gate decision stores page reference + template reference + distinctness score + substantive-content score + heuristic-match flags + decision + actor + resolution. Audit trail queryable per template per page per time period for retrospective + reconsideration evidence + policy-update impact analysis.
  13. ROI measurement. Gate-pass rate per template trend over time. Organic traffic per template trend over time. Manual-action count pre vs post deployment. Recovery cycle time from Google enforcement to remediation. Per-template revision velocity. Signal feeds heuristic tuning + template-revision prioritization per cycle.

Frequently asked

What is programmatic SEO and what is the risk?

Programmatic SEO is the practice of generating large numbers of pages from templates plus data to capture long-tail keyword traffic. Multi-location operators use programmatic SEO to publish per-location landing pages, per-service-area pages, per-product-category pages at scale. The technique works when each generated page carries substantive distinct content beyond the template scaffolding. The risk is the Google scaled-content-abuse policy (formalized in March 2024) plus the broader Helpful Content System plus the doorway-page penalty that has existed since 2015. Pages that score as near-duplicate scaled content, doorway pages, or thin content lose ranking through algorithmic dampening or land manual actions that deindex the affected URLs.

What is a content-distinctness gate and how does it work?

A content-distinctness gate runs as a pre-publish check in the page-generation pipeline. Every generated page evaluates against three measures before publish — corpus distinctness (token-overlap score against every other operator page in the same category), substantive-content score (LLM-as-judge evaluation of whether the page provides unique value beyond template scaffolding), and Google scaled-content-abuse heuristics (pattern-matching for boilerplate-heavy + thin-content + doorway-shape pages). Pages exceeding distinctness threshold publish. Pages failing route to revision queue. The gate prevents the scaled-content-abuse pattern from accumulating in the operator catalog rather than discovering it after Google enforces.

How is this different from Surfer SEO, Scalenut, NeuronWriter, ContentatScale, Clearscope, or Frase?

Those platforms ship the content-generation primitive with keyword research + content briefs + AI-assisted writing + on-page SEO recommendations + content quality scoring. They are excellent at the content-authoring layer. The pre-publish content-distinctness gate that evaluates every page against the broader operator corpus to detect near-duplicate scaled-content patterns, the LLM-as-judge substantive-content evaluation, the Google scaled-content-abuse pattern matching, the integration with the 3-stage multi-location content-quality pipeline (pre-publish gate + publish substantive content + post-publish cannibalization detection), and the per-location-page-generator agent integration are operator-side architecture above the content-authoring primitive.

What is the 3-stage multi-location content-quality pipeline?

The pipeline organizes three skills across three temporal stages. Pre-publish gate (this skill — content-distinctness gate runs before page publish). Publish substantive content (neighborhood-FAQ-authoring + per-location-content-generation produce substantive distinct content). Post-publish cannibalization detection (local-cannibalization-defense + content-conflict-resolution detect cross-location keyword cannibalization after publish). The three stages cover the full content-quality lifecycle from generation through publication through post-publish monitoring. Per-location-page-generator agent owns the pre-publish and publish stages; downstream detection sits in adjacent agents.

How do you distinguish legitimate per-location pages from doorway pages?

Per-location pages are legitimate when each carries substantive distinct content reflecting actual per-location differences — per-location services + per-location pricing + per-location team + per-location hours + per-location testimonials + per-location case examples + neighborhood-specific context + local-event references. Doorway pages are illegitimate when each page is boilerplate plus a swapped city name plus no substantive distinction. The distinctness gate evaluates against the substantive-content score (LLM-as-judge), the corpus distinctness threshold (token-overlap), and the Google scaled-content-abuse heuristics. Pages that meet all three criteria publish; pages that fail route to revision queue. The gate enforces the legitimacy criterion at the pipeline level rather than relying on per-page editorial review at multi-location scale.

What happens when the gate fails a page and how does revision work?

Failed pages route to a revision queue with the failure-mode signal surfaced (corpus distinctness too low + substantive content too thin + scaled-content-abuse pattern detected). The revision workflow proposes specific revisions matched to the failure mode. Low corpus distinctness routes to template-pattern revision (the boilerplate is too dominant; reduce template + increase per-location-specific content). Thin substantive content routes to content-enrichment workflow (add per-location case examples + neighborhood context + local-event references). Scaled-content-abuse pattern routes to template-redesign workflow (the underlying template approach is producing pages Google will dampen). Per-template revision propagates the fix to every page generated from the template.

Hire the agent that gates every programmatic SEO publish

The per-location-page-generator agent owns the pre- publish content-distinctness gate — corpus distinctness + LLM-as-judge substantive-content evaluation + Google scaled-content-abuse heuristic matching — sitting on top of whichever content- generation primitive (Surfer SEO, Scalenut, NeuronWriter, ContentatScale, Clearscope, Frase) or programmatic-SEO platform (Webflow programmatic, Framer GPT, custom Next.js/Astro pipelines) or AI content detection (Originality.ai, GPTZero, Writer, Sapling, Copyleaks) you license downstream. 3-stage multi-location content- quality pipeline + failure-mode-matched revision workflow + per-template revision propagation + per- template coverage measurement + existing-catalog audit for recovery + Google policy update tracking + manual- action recovery workflow + audit trail.

Hire the page-generator agent

We scope on the call and send a private checkout link after.

Related reading: Multi-location SEO architecture · URL hierarchy authoring · Canonical-tag management