Every page here is also a document. Use your browser's print or save-as-PDF and you get the same PDF that goes out by email, which is why these are laid out like sheets of paper.
EZO Content Pipeline · Pillar 3 walkthrough · SEO session · 2026-07-28

Your rules, machine-enforced - and your sheet, machine-fed

Left: the QA check running the team's own SOP + SEO rulebook on 4 real in-flight pages. Right: the proposed Gumshoe routing that ends hand-updating the LLM Optimization Plan sheet. You own the rules and the mapping - correct both in this session.

● live / built ◕ in motion this week ○ proposed / queued ■ blocked - named owner score = percent of the 9 SEO checks passed (0-100) PASS = score 80+ and no CRITICAL finding n/9 = checks passed of 9 run per page HIGH / MED / LOW = finding severity as set in the ruleset
● A · The check on 4 real in-flight pages (run: pipeline/runs/draft-qa)

9 deterministic SEO checks per page - H1, meta title, meta description, slug, separator format - encoded 1:1 from the Sprint SOPs and the SEO rulebook. Exact character counts, no hand-measuring. Pages labeled by role cluster, not by person.

SaaS Management Platform - feature pagewriter draftPASS
100 · 9/9
✓ zero findings - meta title, description, H1 and slug all inside SOP ranges
Hardware Asset Discovery - feature pagewriter + SEO + PMM clusterPASS
89 · 8/9
HIGHh1-length H1 is 44 characters; SOP allows 20-40
Problem Management - feature pagewriter + SEO + PMM clusterPASS
89 · 8/9
LOWslug-format slug has 2 words; rulebook prefers 3-6
IT Asset Management (ITAM) - solution pageowner unresolved in trackerFAIL
67 · 6/9
MEDmeta-title-length 88 characters; rulebook allows 50-60
MEDmeta-description-length 162 characters; rulebook allows 145-160
LOWslug-format slug has 2 words; rulebook prefers 3-6
● After the check: FixPass repairs, then re-checks (eval harness, planted violations)
100%
planted problems the rule checker caught
86%
of fixable violations auto-fixed, then independently verified
0%
unjustified edits - every change carries a reason
◕ B · Gumshoe → your sheet, automated (proposed - you correct the mapping)

API key received; the API is not wired yet, so step one is mapping what it exposes against what you paste by hand today. Nothing writes to your sheet until the mapping below has your sign-off.

◕ Gumshoe
key in hand; endpoints confirmed against your real reports first
→
○ Weekly pull
Mondays + on demand; raw pull kept for replay
→
○ Schema guard
tabs + headers pinned; drift = stop, write nothing, alert
→
○ LLM Optimization Plan Q3 2026
your tabs; machine rows stamped with pull date
Gumshoe data (as in your sheet today)Destination tab · columnHand rows today
Pages LLMs pick as sources + how oftenFresh analysis - which pages are picked as sources · url, page type, Citation frequency25,269○
Blog topic citation weightBlogs - unique topics to build · LLM citations (importance)23○
Reddit threads cited in answersSocial - Reddit · Citation frequency311○
YouTube citationsSocial - YouTube videos · LLM citations, # cited videos19○
Competitor citation pages behind each gapMoney-page gap tabs · # competitor pages, # domains16-868○
Visibility, LLM rank, mentions (weekly)Full Funnel Data · "1. AEO & LLM" → feeds the weekly AEO report (report itself already runs)1 col/week◕
● Stays yours, never machine-written: the planning tabs - ITAM / ITSM / EAM planned activities (Activities, Brief, Action, Contributor(s), Dependency, Publish date, ITSM's Gumshoe prompt column) and Plan for Q3. The machine feeds evidence; you decide the plan.
◕ Weekly AEO report
same Monday pull generates it; "not measured" never becomes a silent 0
→
■ Asana table (later)
same rows routed into Asana once write-back is approved (owner: CMO)
Every mark on this page is defined in the legend - nothing unexplained. "Hand rows today" = rows currently maintained by hand in that tab of the LLM Optimization Plan Q3 2026 sheet. Left-half scores and findings are real output from the check run on the team's in-flight drafts (pipeline/runs/draft-qa/summary.json); FixPass tiles are from the eval harness on planted violations. Right half is a proposal for this session: correct the column mapping, choose in-place vs staging tabs, confirm the Monday cadence.