What a delivery looks like
Below is an internal test on a public blog post (not a paid client engagement). We show it so you can judge the process — not as a performance claim. No client traffic numbers here.
1. Sample decay queue
Illustrative ranking. Toggle treat vs control.
| URL | Signal | Est. recovery | Group |
|---|---|---|---|
| /guides/local-seo-checklist | Title still says 2023; “Google My Business” ×4 | High | TREAT |
| /blog/speed-tools | Mentions retired Mobile-Friendly Test | High | TREAT |
| /blog/schema-101 | Stable; matched topic | — | CONTROL |
| /resources/ppc-glossary | “AdWords” in present tense | Med | TREAT |
| /blog/naps-vs-citations | Matched decline band | — | CONTROL |
| /guides/sge-seo | SGE name; pre–AI Overviews | Med | TREAT |
Illustrative rows for UI demo — not a real client's queue.
2. Before → after (excerpt)
From an internal refresh of a public post that still said “SGE”. Sources linked in the after version.
Title: Understanding Google SGE and Its Impact on Your SEO Strategy
Talks about SGE as the current product name. No link to Google's May 2024 AI Overviews announcement. No independent click-rate source.
Title: AI Overviews (formerly SGE): What Changes for Your SEO
Renames the product, cites Google's announcement, adds Pew Research click rates with a live link.
3. Automated claim checklist (real run)
Output of check_claims.py on that internal draft. Result: PASS — 2 SOURCED, 0 FLAG.
| # | Status | What |
|---|---|---|
| 1 | SOURCED | AI Overviews rollout May 2024 — link to blog.google |
| 2 | SOURCED | Pew 8% vs 15% click rates — link to pewresearch.org |
We also run a deliberate negative test (invented stats, “guaranteed #1”, broken link, outdated terms): the checker returns FAIL with 7 flags. That's the bar every delivery has to clear.