Word counts are reasonable across product, collection, and blog content. The dominant content-quality drag is two staging-clone pages indexed for 12–18 months at near-100 % similarity to their canonical originals; a long tail of 14 thin brand-collection pages (107–115 words each); and product descriptions that read as HTML-escaped spec dumps rather than purchase-aiding prose. E-E-A-T signals are present on the local-business side (address, hours, owner mention) but absent on the editorial side (no blog author bylines, no "press / featured in", no dealer credentials).

Word Counts by Page Type

Page typeAvgMinMaxSample size
Homepage1,4531
Product (PDP)56254758824
Collection (PLP)1,0081072,42447
Page (CMS)58113976411
Blog post9075531,22413
Policy1,6221213,1224

Thin Content (<200 visible words)

14 collection pages fall below the 200-word threshold. These are typically brand-only landing pages where a 150–250 word brand introduction would unlock [brand] for sale, [brand] Houston, and [brand] pre-owned rankings.

URLWord count
/collections/anonimo114
/collections/bremont110
/collections/bucherer107
/collections/bulova110
/collections/carl-f-bucherer115
/collections/david-yurman110
/collections/harry-winston112
/collections/laine111
/collections/mb-and-f108
/collections/other113
Note: hub vs content distinction
Collection pages with low word counts but many images and internal links function as navigation hubs. They shouldn't be flagged as "thin" by simple word counts alone. However, even hub pages benefit from a 150–250 word brand introduction (history, in-stock highlights, why-buy-from-us). The brands above are exactly the high-margin enthusiast brands where copy investment pays back fastest.

Duplicate Content — Critical

/pages/about-us vs /pages/new-about-us — 99.4 % similarity
Same hero image, same body text, same internal links. Only the page <title> and URL differ. The "new" version's title is literally "New About-us" — clearly an unfinished draft from an abandoned redesign. Last modified 2024-05-14. Both pages are in the indexable sitemap.
/pages/contact-us vs /pages/contact-us-new — 74.2 % similarity
Both indexed. Both in the sitemap. Last-modified 2024-11-21 for the "new" version.
/pages/get-best-price-rolex-luxury-watches
No <H1>, last modified 2024-02-17, reads like an SEO-only landing page that has gone stale.
/pages/search-results-page
Internal placeholder for search results, indexed via the sitemap. Should not be discoverable.

E-E-A-T Signals (Experience, Expertise, Authoritativeness, Trustworthiness)

What works
  • Local presence is unambiguous: 73 / 100 pages mention "Houston", 35 / 100 mention "Bellaire", 11 / 100 mention the street address
  • Phone and email appear in the site header on every page
  • Human ownership referenced on About + Sell-a-Watch pages
  • Reviews page exists (/pages/watchvault_reviews, 3,480 words)
  • All four core policies present (Privacy, Refund, Terms, Shipping)
  • Blog has 12 BlogPosting articles, 553–1,224 words, on luxury-watch topics with reasonable depth
What's missing
  • No author bylines on any blog post. BlogPosting JSON-LD has no author property. AI Overviews discount unattributed editorial.
  • No dealer credentials: no GIA, IWJG, BBB badges, no years-in-business statement
  • No customer testimonials with names + dates on the homepage
  • No "press / featured in" section (Hodinkee, Robb Report, GQ etc.)
  • Authentication / authenticity story buried. "We authenticate every watch" should be explicit on PDPs and About
  • No AggregateRating or Review JSON-LD anywhere — reviews exist as text but not as schema

Product Description Quality

All 24 product meta descriptions are 320-character spec-sheet dumps
Example: "Bezel Color: BlackMovement: Mechanical (Automatic)Water Resistance: 3000 m (300 ATM)With Papers: NoReference Number: M17731Brand: BreitlingCase Color: BlackDepartment: MenType: WristwatchDealer Code: …" — no commas, no benefit terms, no CTA, no price, no purchase-aiding context. Google truncates at 155–160 chars; the visible portion conveys nothing useful to a buyer.
In-page Product description (rendered HTML inside JSON-LD)
The description field of the Product JSON-LD on a sampled Rolex GMT-Master II PDP is 19,705 chars of HTML-escaped spec markup. Google strips the HTML, but the dense unstructured prose provides almost no additional retrieval signal beyond the structured fields already present.

AI Citation Readiness (Cross-Reference)

  • All AI bots return 200 (verified)
  • Site captured in Common Crawl CC-MAIN-2025-43, multiple snapshots
  • /llms.txt exists but is the auto-generated Shopify default — minimal customization, missing brand portfolio, address, hours, FAQ summary
  • /llms-full.txt is only 41 lines / 1.4 KB — basically duplicates llms.txt
  • Most pages have flat H2 → paragraph structure with no sub-sections. Passage-level retrieval (used by Google Passage Ranking and AI Overview citation extraction) needs question-shaped H2/H3 headings followed by 50–150 word direct answers.

See the AI Search Readiness section for full GEO analysis.

Action Items

C1Unpublish duplicate staging clonesXS · 15 min

Why: /pages/new-about-us (99.4 % identical to /pages/about-us) and /pages/contact-us-new (74.2 % overlap with /pages/contact-us) are both indexed in sitemap_pages_1.xml. Sitewide canonicalization signals are split.

How:

  1. Shopify admin: set both pages to "Visible: Hidden" (returns 404 + drops from sitemap)
  2. Add 301 redirects: /pages/new-about-us → /pages/about-us and /pages/contact-us-new → /pages/contact-us via Shopify URL Redirects
  3. Submit sitemap_pages_1.xml re-crawl in Search Console

Owner: Site admin · Acceptance: curl -I https://watchvault.us/pages/new-about-us returns 301 to canonical; both URLs absent from rebuilt sitemap.

M2Write 200-word brand introductions for thin collectionsM · ~5 hrs

14 collection pages under 200 words (Anonimo, Bremont, MB&F, Laine, Bucherer, Bulova, Carl F. Bucherer, David Yurman, Harry Winston, etc.). Each: 1–2 sentence brand history, watches typically in stock, what to look for in a pre-owned [brand], CTA to call/visit.

Acceptance: Every collection page ≥ 200 visible words above the product grid.

M3Add author bylines + Person schema to all blog postsM

AI Overviews and E-E-A-T heavily weight author identity. Currently BlogPosting JSON-LD has no author.

  1. Create author profile page(s) (e.g., /pages/author/mike-watchvault)
  2. Render <a rel="author" href="…"> byline in templates/article.liquid
  3. Update BlogPosting JSON-LD to include author: {"@type":"Person","name":"…","url":"…"}

Acceptance: Every blog post has visible byline + Person schema; valid in Rich Results Test.

M-PASSAGERe-template editorial pages with passage-level citability structureM

Most editorial pages have flat H2 → long paragraph structure. Convert About, Sell-a-Watch, FAQs, and blog posts to question-shaped H2/H3 + 50–150 word direct answers immediately below. AI Overview retrieval extracts passages under headings; this is the single fastest intervention for AI-citation rate on long-form content.

L7Build dedicated "About Watch Vault" trust pageM

Direct-answer page targeting [is watch vault legit], [watch vault houston reviews], [watch vault BBB] queries. Include FAQPage schema and dealer-credentials list.

Cross-references: Product meta-description rewrite is filed under On-Page (H6). AggregateRating / Review schema additions are under Schema (M4, L6).