Add EEAT credentials reference document#32
Open
stefan-burke wants to merge 11 commits into
Open
Conversation
Reference document of verifiable trust signals (Experience, Expertise, Authoritativeness, Trustworthiness) for use in marketing copy. Sourced from non-AI portions of the existing site: FAQs, testimonials, sidebar copy, footer, privacy notice, equipment listings, package specs, and older location pages. AI-generated marketing paragraphs (those carrying ChatGPT export artifacts) were excluded. https://claude.ai/code/session_01WhGXZskvCs5L8GjSveZKFw
Add Eleventy frontmatter (permalink: false, eleventyExcludeFromCollections: true) so the working reference document doesn't generate a public page or appear in collections when the site is built. https://claude.ai/code/session_01WhGXZskvCs5L8GjSveZKFw
Restructure the credentials doc around what is actually verifiable, what sits in line with industry baselines (MUTA membership, BS EN 13782, £5m+ public liability, fire-retardancy certs), and what is missing. Key corrections: - Royal wedding line clarified: Complete Marquees supplied a marquee for a private party connected to a royal wedding, not the royal wedding itself. Sidebar copy reworded recommendations added. - "Hundreds of events" reframed as the company's own self-stated figure, not externally verified. - "24 years of continuous operation" softened to "24 years on the trading record" with continuity self-stated. - "GDPR-compliant" softened to "publishes a GDPR-style privacy notice". - "Company number 1090110" hedged as site-stated, with a Companies House cross-check recommended. - Capri ownership reframed as commodity, not differentiating. - Family-run + 20 years repositioned as mid-pack for serious operators, not exceptional. - Pricing comparison added: at £995 for an 80-guest package vs the £2,000-£3,600 wedding-tier benchmark, Complete Marquees sits in the mid-market value tier. - Added explicit gap list: no advertised MUTA membership, no published PL insurance figure, no BS EN 13782 / fire-retardancy / SafeContractor documentation, no published terms of business, no site-visit policy. - Added a three-tier market comparison table. - Added eleven concrete recommendations to harden credentials. https://claude.ai/code/session_01WhGXZskvCs5L8GjSveZKFw
VOICE.md is a tone-of-voice guide adapted for the wedding-and-corporate
clientele: dry, plain-spoken Hampshire, calm rather than matey, with
explicit guards against wedding-magazine vocabulary ("magical", "stunning",
"bespoke") and the country-house / Boden-catalogue voice. Frontmatter
keeps it out of the public site build.
Also corrects the EEAT credentials doc to reflect that Complete Marquees
is run as a husband-and-wife operation by David and Joanne (the testimonial
names Joanne; David is not yet on any public-facing page). The owner /
team page recommendation is updated accordingly.
https://claude.ai/code/session_01WhGXZskvCs5L8GjSveZKFw
These are working-reference documents at the repo root, alongside README.md. Without this exclusion they get merged into combined/src/ and Eleventy attempts to process them, which broke the Build and Deploy CI step. The Eleventy frontmatter on each file (permalink: false, eleventyExcludeFromCollections: true) is kept as a belt-and-braces fallback in case the merge config changes. https://claude.ai/code/session_01WhGXZskvCs5L8GjSveZKFw
Two fixes to unblock the Build and Deploy workflow.
1. The chobble-template's eleventyComputed validation requires every
tagged content item to either provide a "name" frontmatter field or
set "eleventyExcludeFromCollections: true". pages/thank-you.md and
pages/not-found.md provided neither, breaking the build with:
Error: Item is missing required "name" field in
./src/pages/thank-you.md
at blocks (./src/_data/eleventyComputed.js:226)
Both are utility pages (form thank-you and 404), not collection
members, so eleventyExcludeFromCollections: true is the right marker.
2. While here, add CLAUDE.md to sourceExcludes alongside README.md /
VOICE.md / EEAT-CREDENTIALS.md. CLAUDE.md is internal AI guidance and
was being merged into the build, rendered, and published as
/CLAUDE/index.html on the live site (Pagefind even indexed it). It
should not be public.
Verified locally with `bun scripts/build.js` - build completes cleanly,
_site/CLAUDE/ no longer exists, _site/EEAT* and _site/VOICE* never did.
https://claude.ai/code/session_01WhGXZskvCs5L8GjSveZKFw
Belt-and-braces fallback to match VOICE.md and EEAT-CREDENTIALS.md. The sourceExcludes entry in scripts/consts.js already prevents these files reaching the build, but if that exclude is ever forgotten or overridden the frontmatter ensures Eleventy doesn't generate a public page. https://claude.ai/code/session_01WhGXZskvCs5L8GjSveZKFw
Working draft of a five-phase plan to bring the location pages up to the same standard as the EEAT and VOICE docs. Includes: - Triage of all 27 pages (only 3 are AI-rewritten; aldershot and andover heavily, hampshire lightly; the rest are original-voice but generic). - A target per-page shape with shared includes for the standard EEAT block, site rules and packages CTA. - A phase-1 quick-wins pass (strip ChatGPT export attributes, emoji headings, em-dashes; reword the royal-wedding sidebar line per EEAT-CREDENTIALS.md section 3) that visibly improves the site without rewriting copy. - A phase-2 boilerplate consolidation step, a phase-3 pilot page, a phase-4 batched roll-out and a phase-5 SEO consolidation review. - Open questions for the owner that need answering before phase 3: which towns we've genuinely delivered in, named venues per town, MUTA / PL insurance status, and whether to keep all 27 pages. - A definition-of-done checklist for each page. Excluded from the site build alongside the other internal docs. https://claude.ai/code/session_01WhGXZskvCs5L8GjSveZKFw
Owner confirmations folded in: - £10m public liability cover (held, not previously displayed) - All 27 location pages have genuinely been delivered into - No MUTA membership - No specific named-venue references available for location pages - Royal-wedding line is to be dropped from public copy Sidebar CTA (_includes/sidebar-cta.html, every page) updated from "From small gatherings to a royal wedding..." to "Family-run from Havant since 2002, with £10m public liability cover and our own install team." EEAT-CREDENTIALS.md: - Owner-confirmed-facts section replaces the royal-wedding correction - Gap analysis: PL flips from "not stated" to "£10m, above £5m baseline" - MUTA stays as a gap (now confirmed not held) - Tier comparison: Complete Marquees row now notes £10m PL specifically - Recommended actions: "publish the £10m figure" replaces "publish PL" - Marketing-copy notes lead with £10m PL as the highest-leverage line LOCATION-PAGES-PLAN.md: - Section 8 now captures the owner answers - Phase 5 (consolidation) is dropped - New section 11: thin-page risk analysis sorts the 27 pages into three tiers (writable now / writable but thin / at risk) and lists per-town asks for the Tier B and Tier C pages so the owner can supply one specific per town to thicken the lead paragraphs. - Flags an unrelated credibility leak: "Packages from only £330" in the second sidebar CTA, vs the cheapest published package at £390+VAT. https://claude.ai/code/session_01WhGXZskvCs5L8GjSveZKFw
A plain-language list of 33 questions to pass to the owners, ordered roughly by impact and written in the house voice (per VOICE.md) so it reads as a normal note rather than a survey form. Grouped into eight sections: 1. Insurance and paperwork (PL certificate, fire-retardancy, other accreditations) - the biggest single trust unlock 2. The two of them - photos and short bios for an owner / team page 3. How they actually work - site visits, lead times, deposit / cancellation / weather terms, payment methods 4. The £330 sidebar mystery (sidebar shows from £330, packages page shows from £390 plus VAT) 5. Towns we'd love a specific for - the 13 Tier B/C towns from LOCATION-PAGES-PLAN.md section 11, presented as a fill-in table 6. Recent jobs and recent customers - new photos and fresh testimonials to replace the four undated ones currently on the site 7. Towns to add or drop - in particular flagging that Havant itself doesn't have a page 8. Voice and last bits - things we shouldn't say, things they do say worth picking up, and the Gmail vs domain-email decision Essentials are marked so they can prioritise. Excluded from the site build alongside the other internal docs. https://claude.ai/code/session_01WhGXZskvCs5L8GjSveZKFw
The owner questionnaire has been sent. Removing the working copy from the repo and dropping its sourceExcludes entry. https://claude.ai/code/session_01WhGXZskvCs5L8GjSveZKFw
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Adds
EEAT-CREDENTIALS.md— a reference document compiling every verifiable trust signal (Experience, Expertise, Authoritativeness, Trustworthiness) for Complete Marquees, drawn from authentic, non-AI portions of the existing site.Intended as a working reference for marketing copy, sales calls, and proposal writing — most of these signals already exist on the site but are buried under AI-generated paragraphs that dilute them.
Sources used
Authentic site content: FAQs page, testimonials, sidebar CTAs, footer, privacy notice, equipment listings (with prices), package specs, older un-rewritten location pages (Salisbury, Southampton).
Sources excluded
AI-generated marketing paragraphs were ignored — identifiable by ChatGPT export artifacts left in the HTML (
data-start,data-section-id,data-message-model-slug="gpt-5-3"), checkmark emojis, em-dashes, and the typical "elevate / unforgettable / seamless" vocabulary.Highlights surfaced
Test plan
https://claude.ai/code/session_01WhGXZskvCs5L8GjSveZKFw
Generated by Claude Code