Content Quality
E-E-A-T, thin content, and the city-page templating problem — verified September 25, 2026
This score is driven almost entirely by one systemic issue: the ~50 city landing pages are a single template with the city name swapped in 3–4 places, functioning as classic doorway pages. Everything else on the site — E-E-A-T signals, service pages, the "blog" — is mediocre-to-thin but not as severe. Methodology: every page below was fetched live via curl and diffed with Python's difflib.SequenceMatcher after normalizing out city names, per Google's September 2025 Quality Rater Guidelines framework.
The #1 Finding: City Pages Are One Template
Fetched and compared /tree-service-provider-katy-tx/, -sugarland-tx/, -bellaire-tx/, -the-woodlands-tx/, -tomball-tx/, and -lake-charles-la/ (now removed, see below) — six pages spread across the full service-area map.
| Comparison | Similarity (Sept 25) |
|---|---|
| Katy vs. Bellaire | 97.8% |
| Katy vs. Tomball | 97.8% |
| Bellaire vs. Tomball | 100.0% (byte-identical once normalized) |
| Katy vs. Sugarland | 99.9% |
| Katy vs. The Woodlands | 97.7% |
Standard city pages land at 530–549 words each — suspiciously tight clustering consistent with one template being cloned with only the city token replaced. What actually varies: the H1 hero heading, one "We Also Serve [City], TX!" line, and the title/meta description. Everything else — the "why choose us" section, service list, bio blurb, and footer — is 100% boilerplate. There is no city-specific content anywhere: no local neighborhoods or landmarks, no local tree species or climate notes, no local case studies, no locally-tagged photos.
Original fix recommendation: add genuine per-city content — local neighborhood names, common local tree/soil issues, a locally-relevant testimonial or project photo, real drive-time framing. Even 150–250 words of real local specificity per page would materially change the doorway-page risk profile.
Status (Oct 4, 2026, updated three times same day): round 1 added a genuinely new, unique ~80-100 word paragraph to the top of all 43 real service-area city pages (99.8% → 91.0% similarity). Round 2: varied the "Why Choose Us" 3-blurb section across 5 templates (same real facts — 24 years since 2002, ISA Certified Arborist Daniel Arias TX-4490A, equipment/safety — different wording) and added a second new paragraph using real Houston-region facts (clay soil, humid subtropical climate, hurricane season) (91.0% → 76.0%). Round 3 (client kept pushing to close the gap further): varied the 4 remaining service-category blurb cards (Commercial / Residential / Storm Damage / Tree Diagnosis — previously 100% byte-identical across all 43 pages) across 5 templates each, same real claims reworded. Klein vs. Katy re-diffed: 99.8% → 59.5% similarity, measured the same way all three times. Every sentence across all three rounds uses only real, verifiable facts — no invented local landmarks, job specifics, or statistics. Still real progress, not full completion: the remaining overlap is mostly shared site chrome (nav/footer) and the bio/credential sidebar box, which would need real local specifics (landmarks, project photos, testimonials) from the client to differentiate further — that's the genuine ceiling without new client input. Also: Sulphur LA and Lake Charles LA (the 2 out-of-territory pages) were removed entirely (not just noindexed) at the client's request, and a third out-of-territory page — Georgetown, TX, ~170 miles away near Austin, missed in the original audit — was found and noindexed.
Diffed two of the flagged pairs in full. /tree-service-provider-the-woodlands-tx/ vs. -tx-2/: byte-identical <title>, meta description differing by exactly one phrase ("Houston based," dropped on the -2 variant), both self-canonicalized and indexable, 97.9% identical body text. The Sugar Land pair follows the same pattern at 98.4% identical. All 6 pairs now 301-redirect to a single surviving URL (Session 2, 2026-09-30), re-verified live Oct 4, 2026. See Technical SEO for the full 6-pair table.
E-E-A-T Signal Assessment
| Factor | Score | Notes |
|---|---|---|
| Experience | 30/100 | Real named testimonials with dates ARE embedded on the homepage (Eric D., Kevin Kaplan, Rita Smith, and others, Feb 2020–Jul 2022) — a genuine positive. But no first-hand case studies, and the gallery has 260 images with 236 sharing just two generic alt strings ("Houston Tree Service" / "Tree Trimming Service") — no per-photo specificity. |
| Expertise | 45/100 | Daniel Arias is named owner with a real, checkable "ISA Certified Arborist Credential ID TX-4490A" on nearly every page — but it's never hyperlinked to a verification tool, and there is no author bio page, team page, or photo of Daniel Arias anywhere. |
| Authoritativeness | 20/100 | No external citations, press mentions, or third-party trust badges rendered on the pages. One gallery alt tag oddly references an unrelated business ("Chimney Sweep San Antonio"), suggesting reused stock/template gallery assets. |
| Trustworthiness | 40/100 | NAP is consistent in the footer/schema of the pages sampled here (16135 Barklea Rd, Cypress TX; 713-966-9955) — but see Local & Maps for a serious NAP exception found elsewhere on the site. "Years of experience" claims are stale and internally inconsistent (see below), and the newest visible customer review is from July 2022 despite 37 live GBP reviews as of August 2026. |
High Findings
Grepped the phrase across pages, fetched the same day (2026-09-25):
| Page | Claim (Sept 25) |
|---|---|
| Homepage | |
| City pages (Katy, The Woodlands, etc.) | |
| Homepage (separate paragraph) |
2002 to 2026 is actually 24 years — none of the three numbers on the site were correct, and they disagreed with each other. This is exactly the kind of low-effort inconsistency the September 2025 Quality Rater Guidelines flag as a trust/quality red flag.
Status (Oct 4, 2026): the client confirmed "since 2002" (24 years) is the real anchor. Found and fixed every instance of the 3 conflicting numbers — 119 occurrences across 67 posts, in two passes (a city-page-specific phrasing variant needed a second scoped regex after the first pass). Re-verified live sitewide: zero stale instances remain.
Nearly every sampled page's dateModified clusters within a single ~1-hour window on 2025-05-07 (16:51:31–17:47:36 UTC) — a bulk find/replace or plugin update, not 47 independent instances of someone rewriting unique content that day. The -2 duplicate pages and the service pages checked were last modified a full year earlier (2024-08-08 / 2024-05-04) and never touched in that batch — explaining why the duplicate pairs keep drifting further out of sync.
| Page | Word count | Risk |
|---|---|---|
| /tree-removal-services/ | 383 | High — below the 800-word service-page floor |
| /tree-services-cypress-tx/ | 247 | Very high — a list of 7 service names with one-line descriptions |
| /tree-maintenance-tips/ | 603 | 11 distinct topics crammed into 603 words — each gets roughly one paragraph or less |
/tree-maintenance-tips/ is a single static WordPress Page, not a post-type content hub — no archive, no categories, no author byline, no visible post date. The entire site is Pages, not Posts; there is no evidence of an actual content-marketing program.
Medium Finding
/tree-solutions-gallery/ has 260 <img> tags; alt="Houston Tree Service" appears 120 times and alt="Tree Trimming Service" appears 116 times — duplicate boilerplate alt text on 91% of the gallery's images, with no per-photo captions or job/city specificity.
Status (Oct 4, 2026): most filenames are meaningless sequential numbers (Thumb-109, Thumb-111...) with zero real distinguishing info — writing per-photo specificity for those would mean inventing details, which wasn't done. Instead, found 24 images that each appear specifically in exactly one dedicated service page's content (Tree Removal, Tree Trimming, Stump Grinding, Storm Damage Cleanup, Commercial, Deep Root Fertilization, Root Barriers, Tree Diagnosis) — a real, verifiable signal, not a guess. Updated those 24 to real service-specific alt text, verified live. The remaining ~161 images have no real distinguishing data available without new input from the client (same blocker as the case-study content item) — recommend batch-renaming source files by job/city before future uploads, as already noted below.
Missing-Content Opportunities (prioritized)
- Real local content per city page: neighborhoods, species/soil notes, local testimonials and photos
- A real About/Team page with Daniel Arias's bio, photo, and the ISA credential linked to a verification source
- A real, dated, authored blog (WordPress Posts, not Pages) — split the 11 crammed tips-page topics into single-topic posts
- Depth on the thin service pages: process detail, safety certifications, FAQs — not padding, real answers
- Unique, descriptive alt text and captions on gallery photos, ideally tied to specific jobs/cities/species
- A refreshed on-site reviews display reflecting the real, current 5.0/37 GBP rating (see Local & Maps)