Section 05 · Why commercial visibility was always fragile

Entity & cluster findings

None of this caused the June event — the seasonality audit was clear about that. But these structural findings are DOM-, schema-, and GSC-verified, window-independent, and they explain why the phone was always ringing off a thin base.

§ 5.1 — Treasure Coast

A region-brand in a town-level market, pinned to the wrong coast

"Treasure coast" commercial demand, 5.5 months of GSC queries

≈ 2 / ~100

Roughly two commercial queries and ~100 impressions. The market is 100% town-level — the site's core pages and H1s anchor a region almost nobody searches commercially.

Schema business address

Palm Bay

1020 Sally St SE, Palm Bay — which is the Space Coast, not even the Treasure Coast. The site behaves exactly like a business pinned to one Palm Bay point with no city entity.

Rank vs distance from the pin

Top 10 ↔ p49–79

Top-10 only in Palm Bay / Sebastian / Vero / Melbourne (near the pin); position 49–79 in Port St. Lucie / Stuart / Fort Pierce — the towns the franchise is named for.

The mechanism is displacement, not deranking. Positions held (e.g. "snake removal near me" at pos ~3) while impressions went to ~0; average commercial position drifted 11.2→21.6 but from a base of ~1.7 clicks/day. That pattern fits SERP-feature / local-pack displacement — consistent with the 2026 AI local packs (1–2 businesses vs 3, ~32% business carry-over per the June 2026 industry data) plus local-ads crowding.

§ 5.2 — The Florida cluster

Three sites, one street address

The sibling audit found an aggravator the first round missed entirely:

fort-pierce.aaacwildliferemoval.com duplicates TC's entire 42-town /service-area/ map URL-for-URL (84 duplicate paths), declares areaServed="Treasure Coast", and carries the same street address in schema — and already outranks TC in Fort Pierce itself (p11–15 vs p49–79).
palm-beaches carries the same Palm Bay address too. Three FL sites, one NAP.
The branded query "aaac wildlife removal" splits three ways in Florida.
Google's doorway policy (updated 2026-05-15) describes multi-subdomain city targeting almost verbatim. A cited franchise-consolidation case (AlphaGraphics) gained +105% organic.

Round 5: is Fort Pierce the cause of TC's 2026 losses? No.

Age of the duplication

4 years

FP's 165 duplicate town pages were published Jan 2022 (TC's: Oct 2021) and are unchanged (3 pages touched, Oct 2024). It coexisted with TC's good 2024 commercial performance.

FP's Mar 2026 impression surge

11k → 74k

Driven by old 2021–22 blog posts ("animals that eat rats": 283 clicks / 144k impressions) — part of the network-wide blog redistribution, not town pages capturing TC traffic. FP's Jun–Jul clicks grew 136→225.

FP service-area pages earn

~2 clicks each

Near zero. No evidence FP took TC's 2026 losses; TC's Apr–Jun commercial collapse has no FP-side change to pair with.

Verdict: fragility fix, not traffic-loss cause. Dedup still worth doing — it costs nothing (FP's town pages earn nothing), removes doorway-policy exposure, stops the 3-way brand split, and consolidating cluster signals is likely a precondition for TC town pages to win 1–2-slot SERPs. Expectation: necessary but not sufficient; pair with pin/GBP re-anchoring and page differentiation.

§ 5.3 — Dallas

A thin clone sitting on the network's biggest prize

The proposed "Fort Worth whiteout" (FW crowding Dallas out) was contradicted by the data; what remained is more interesting.

Dallas head-term positions

p14–32

On "raccoon / snake / squirrel removal dallas". "wildlife removal dallas" fell from pos ~6–9 to 16–22 in Nov 2025 and never recovered. Fort Worth sits p5–6 on its own head terms with a 68% commercial click share.

Unconverted commercial demand pool

50.9k

Commercial-geo impressions in 5.5 months — the network's largest, bigger than Houston's — against a 5% commercial click share.

The clone evidence

4 pages

All four Dallas /services/ pages are still slugged *-fort-worth — stamped from the FW template. Schema areaServed covers only "Dallas; Irving; Coppell" against an 18-suburb metro.

Why "whiteout" was rejected: zero service-area town overlap between the Dallas and FW sites, and zero cross-bleed — Dallas gets 0 "fort worth" impressions; FW captures ~2 Dallas clicks at p20+; collin-county has no location pages at all. The better explanation is brutal metro competition plus a thin cloned site. The prize is real: 50.9k impressions of commercial demand going unconverted.

Same pattern elsewhere: txgulfcoast mirrors TC (region anchor, Baytown pin, 9 commercial clicks in 5.5 months) — it is TC's problem repeated, not a Houston shadow (zero town overlap; Houston takes trivial gulf-coast impressions).

§ 5.4 — What the literature adds

Context from the research agents

Local-pack rank anchors to the GBP pin and the searcher's location; the GBP "service area" field does nothing for ranking.
Keywords in the business name remain a top-3 pack factor.
AI local packs (1–2 slots, ~32% business carry-over vs the classic 3-pack) expanded May–Jun 2026 — the leading no-ranking-change explanation for lost calls.
Google's doorway policy refresh (2026-05-15) describes the FL cluster's pattern; the AlphaGraphics franchise consolidation case went +105% organic.
Region-name as a cause of the June drop was rejected empirically (region-named sites +2.8% vs city-named −13.9%; central-ohio +105%) — the region problem is about the commercial floor, not the June event.

Cross-site content note: 23 exact-duplicate question titles exist across TC and Dallas (bodies distinct, checked) — a cross-site cannibalization signal. Dallas also has a taxonomy mess (raccoon/raccoons, "cayote", city names inside the animal taxonomy) and ≥1 post with a mismatched title tag.