Client Company Digital Visibility — Full Audit Report

client-example.invalid · anonymized sample · prepared by Comcreate · compiled 2026-05-21

This document combines every individual audit, research, and analysis file produced for the Client Company digital visibility engagement. All data was pulled 2026-05-21 from Ahrefs, Google Search Console, Google Analytics 4, the WordPress REST API, and Google PageSpeed Insights.

Table of Contents

Executive

• Executive Summary

Research

• Research Overview
• Business Profile
• Competitor Landscape

Ahrefs

• Ahrefs — Full Report
• Backlink Toxicity
• Keywords & Rankings
• SEO Traffic Analysis

GSC

• Google Search Console — Full Report
• GSC Overview
• GSC Top Queries
• GSC Top Pages
• GSC Top Movers
• GSC Indexation & Sitemaps

GA4

• Google Analytics 4 — Full Report
• GA4 Overview
• GA4 Sources / Channels
• GA4 Conversions
• GA4 Landing Pages

WordPress

• WordPress — Full Report
• Site Architecture
• Performance & Core Web Vitals

Prior Vendor

• Prior Vendor — 15-Month Work Audit
• Prior Vendor — Content Inventory
• Prior Vendor — Media Timeline

Executive Summary

source: reports/executive-summary.md

Executive Summary — Client Company

Date: 2026-05-21 Headline: Strong brand, weak digital discoverability. The site, reviews, and credentials are an asset; the search visibility and measurement setup are the constraint.

1. Who You Are

Client Company (legal: Client Company) is a San Diego design-build firm founded by Client Principal, BBB-incorporated April 28, 2008, operating ~26 years in residential remodeling. The firm is positioned at the luxury end of the San Diego market:

• Typical job cost: $80,000 – $1,000,000 (per Houzz)
• License: CSLB #[redacted]
• Service areas: San Diego County — emphasis on La Jolla, Rancho Santa Fe, Del Mar, Carmel Valley, Encinitas, Carlsbad, Poway, Scripps Ranch, Tierrasanta, and other affluent coastal/inland zones
• Specialties: Whole-home renovations, kitchens, bathrooms, home additions, custom cabinetry, ADUs, sunrooms, smart-home integration, outdoor kitchens

Credentials & Reputation (the brand is genuinely strong)

• Yelp: 5.0★, 10 reviews
• BBB: A+, accredited since March 6, 2008
• Houzz: "Best of Houzz" winner in 2016, 2017, 2019, 2020, 2021, 2022 (×2), 2023 (×2), 2024, 2026
• EPA Renovator Certified
• HomeAdvisor / Angi / GuildQuality: active profiles, positive reviews
• Coverage: quoted/linked from San Diego Union-Tribune, This Old House, and other DR 80+ outlets

2. The Critical Finding: Search Visibility Is Effectively Zero

The authoritative measurement of organic search performance is Google Search Console (GSC), which reports actual clicks from Google to client-example.invalid. Ahrefs third-party estimates undercount the reality by roughly an order of magnitude and should be treated as directional only.

GSC (authoritative, last 90 days, pulled 2026-05-21):

• ~51 clicks/month from real Google search
• US-filtered, this is closer to ~33 clicks/month from US users

Ahrefs estimates (directional only):

• Jan 2025: ~154 estimated monthly organic visitors
• Apr 2025: ~138
• May 2025: ~77
• Jun 2025: ~16
• Jul 2025 onward: 1–12 /month estimated
• May 2026 (now): 1 estimated organic visitor

The Ahrefs estimate of "1 visitor/month" is not literally true — GSC shows real users arriving — but the direction is correct: discoverability is severely diminished compared to 2023–early 2025.

The site currently ranks for only 2 organic keywords (per Ahrefs):

1. "Client Company" (branded, position 6, 10 search volume)
2. "home remodel rancho santa fe" (position 34, 200 search volume)

Non-branded discoverability is effectively zero. A homeowner Googling "luxury kitchen remodel San Diego" or "design build firm La Jolla" will not find the site.

What's driving it

Three compounding issues:

(a) A site rebuild that broke historical URLs. Every blog post on the current site shows a lastmod of 2026-04-20 — synchronous, indicating a fresh import/rebuild. The footer attributes the site to "Prior Vendor." The old URL structure was not redirected, vaporizing years of accumulated ranking equity. The April 2026 rebuild itself has stabilized — crawl health is fine — but the legacy URLs were never redirected, and the sitemap remains cluttered with stale entries. The bigger ongoing issue is competitive ranking and URL hygiene, not crawl failure.

(b) A toxic backlink profile. Of 267 referring domains, a meaningful chunk are obvious spam:

• 57 domains carry the anchor: "There was a time when client-example.invalid struggled with low visibility... Once I tried SEOExpress.org's backlink building service... +300%..." — a paid-link vendor testimonial-spam pattern
• 8 domains carry: "OUR TELEGRAM CHANEL https://t.me/s/quarterlinks25"
• buybacklinks.agency, site123.me, za.com flagged spam
• Generic "." anchor across 67 domains — classic PBN footprint
• Exact-match commercial anchors (remodeling contractors San Diego, kitchen and bath experts San Diego, sunrooms San Diego) on 8–18 domains each — over-optimized in a way Google penalizes

This pattern is consistent with historical paid link-building that Google has since penalized algorithmically (likely SpamBrain / link-spam updates).

(c) Technical SEO friction. Mobile PageSpeed score: 53/100. 195KB of unused JavaScript, 36KB of unused CSS.

(d) Measurement gaps — the conversion story is largely invisible. Two issues compound here:

• Bot/internal traffic dominates raw GA4. 70%+ of raw GA4 sessions are bot or internal traffic (Singapore, China, Philippines, Pakistan, Serbia). Of ~12K reported sessions, only ~5K are real human sessions. Bot filtering is not applied at the GA4 source — every raw GA4 number in this report must be interpreted with that in mind.
• Form submissions are not being tracked. The generate_lead event has been configured in GA4 since March 2025 but has never fired. The 17 events reported in the last 90 days are a floor — the true conversion count is unknown.

Traffic streams worth surfacing separately

• Google Business Profile (GMB): 123 sessions in 89 days come from ?utm_source=gmb — a distinct stream from organic search, and proportionally significant.
• AI engine referrals: ChatGPT, Claude, and Copilot are sending small but real traffic. Early-mover signal worth noting — AI-search visibility is becoming a meaningful discovery channel.

3. What's Going Right

The site has real assets:

• 88 indexed URLs including 22 location-specific pages (Rancho Santa Fe, La Jolla, Del Mar, Carlsbad, etc.) and 28 service pages — a real long-tail SEO footprint already exists structurally
• 39 blog posts covering all the right topics (luxury renovations, smart homes, coastal materials, cabinet selection, home additions)
• 18 portfolio project pages — visual proof of work
• Real reputation signals — 10+ years of consistent 5★ reviews, awards, and editorial mentions (San Diego Union-Tribune, This Old House, GuildQuality)
• Premium brand positioning that matches the actual project economics ($80K–$1M jobs)

The site infrastructure is there. It is not being found.

4. The Competitive Reality

Direct peers in the same San Diego luxury-remodeling segment are pulling 2,000+ qualified organic visitors a month. Client Company is pulling roughly 51 clicks/month (GSC). At a typical job size of $80K and up, even a single converted lead from organic search is materially valuable — and competitors are capturing hundreds of qualified visits per month while Client Company is largely absent from non-branded search results.

Research Overview

source: research/README.md

Research Index — Client Company

Documentation built from first-party data sources (GSC + GA4) plus third-party reputation research. All audits compiled 2026-05-21.

Sources

Folder Structure

research/
├── README.md                    ← this file
├── business-profile.md          ← company/credentials/reputation
├── gsc/
│   ├── overview.md              ← 28-day & 90-day GSC summary
│   ├── top-queries.md           ← what's ranking, what's not
│   ├── top-pages.md             ← page-level performance
│   ├── indexation-sitemaps.md   ← sitemap & indexation health
│   └── top-movers.md            ← post-rebuild vs pre-rebuild
└── ga4/
    ├── overview.md              ← real session volume + bot analysis
    ├── sources.md               ← channel/source/medium breakdown
    ├── landing-pages.md         ← top entry points
    └── conversions.md           ← event tracking + measurement gaps

Data Reconciliation: Ahrefs vs GSC vs GA4

The three measurement systems each told a different story. The reconciled picture:

Key Findings

The first-party data materially changes the diagnosis vs. what third-party tools alone would suggest:

1. Authoritative organic traffic is ~51 clicks/month from real Google search (GSC) — not "1 visitor/month" as Ahrefs suggests. GSC is the authoritative source; Ahrefs estimates undercount by roughly an order of magnitude.
2. Bot traffic dominates raw GA4 — 70%+ of GA4 raw sessions are bot/internal traffic (Singapore, China, Philippines, Pakistan, Serbia). Real human session count is ~5K of the 12K reported. Bot filtering is not applied at the GA4 source. All raw GA4 numbers in this report must be interpreted with this in mind.
3. Measurement gaps — conversions are not being tracked. The generate_lead event has been configured in GA4 since March 2025 but has never fired. Form submissions are not tracked. Reported events (17 in 90 days) are a floor — true conversion count is unknown.
4. GMB is a distinct traffic stream — 123 sessions in 89 days come from ?utm_source=gmb, separate from organic search.
5. AI engine referrals are real — ChatGPT, Claude, and Copilot are sending small but real traffic. Early-mover signal for AI-search visibility.
6. The April 2026 rebuild stabilized the site — crawl health is fine. The bigger ongoing issue is competitive ranking and URL hygiene (legacy URLs not redirected, sitemap clutter), not crawl failure.
7. Legacy URL duplicates remain — /about vs /about-us, /san-diego-ca vs /service-areas/..., etc.
8. Sitemap clutter — 5 sitemaps registered, most stale.
9. "Ranks but doesn't get clicks" queries are high-leverage — e.g. bathroom remodel san diego (187 impressions, position 50).

Quality Notes / Caveats

• Bot filtering is not applied at the GA4 source — all raw GA4 numbers must be interpreted with this in mind.
• Form-submit tracking is broken — true conversion count is unknown; reported events (17 in 90 days) are a floor only.
• Ahrefs traffic estimates undercount by roughly an order of magnitude vs GSC actuals — Ahrefs numbers should be treated as directional only.
• The 90-day window misses long-tail signal — pulling 16-month GSC history (Jan 2025 → May 2026) would map the full collapse-and-recover arc.

Business Profile

source: research/business-profile.md

Business Profile — Client Company

Compiled: 2026-05-21 Sources: Site content, BBB, Houzz, Yelp, Angi, GuildQuality, HomeAdvisor, ZoomInfo, LinkedIn

Company Identity

Positioning

"San Diego's Premier Residential Remodeling Firm" — luxury / high-end residential design-build, targeting affluent homeowners in coastal and inland San Diego County.

Value proposition (per site)

• 26+ years experience
• Full-service design-build (architecture + interior design + permitting + construction under one roof)
• Dedicated project manager per client
• Founder-led
• "Thoughtful, high-quality spaces that reflect your style"
• 8–20 week typical project timeline

Typical project economics

Per Houzz profile: $80,000 – $1,000,000 project size. This places Client Company firmly in luxury / high-end residential territory — well above contractor-grade work, on par with custom-home / boutique design-build firms.

Services Offered

Core

• Whole-Home Renovations
• Kitchen Remodeling
• Bathroom Remodeling
• Home Additions
• Custom Cabinetry

Per Houzz (more comprehensive list)

• Accessory Dwelling Units (ADUs)
• Deck Building & Design
• Garage Building
• Green Building
• Guesthouse Design & Construction
• Home Extensions
• Home Restoration
• Kitchen Design (separate from build)
• Outdoor Kitchens (construction + design)
• Pool House Design & Construction
• Porch Design & Construction
• Project Management
• General Contracting
• Custom Cabinets (incl. custom kitchen cabinets)

Specialty (from site service pages)

• Smart home integration
• Home automation
• Solar integration
• Wine cellars
• Sunrooms (multiple sub-services: builders, installation, materials)
• Home security systems
• Outdoor living spaces

Service Areas (22 location pages)

Coastal/North County (premium markets)

• La Jolla, Del Mar, Carmel Valley, Torrey Pines, Carlsbad, Encinitas, Cardiff, Solana Beach, Oceanside, Fairbanks Ranch

Inland affluent

• Rancho Santa Fe, Rancho Bernardo, Rancho Peñasquitos, Scripps Ranch, Tierrasanta, Poway

Other

• Mission Valley, Point Loma (per Houzz), and several others

Reputation & Reviews

Yelp

• 5.0 ★ rating
• 10 reviews
• 396 photos uploaded
• Yelp URL: [redacted Yelp profile URL]

Houzz

• 5.0 / 5 overall
• 69 followers
• 1,000+ Ideabook saves (significant — indicates strong portfolio appeal)
• Multiple project showcases (Tierrasanta whole-house, Scripps Ranch master bath, Scripps Ranch home transformation, others)
• Best of Houzz Awards: 2016, 2017, 2019, 2020, 2021, 2022 (×2), 2023 (×2), 2024, 2026 = 11 award years
• "Recommended on Houzz" status

BBB

• A+ rating
• Accredited since 2008-03-06
• Profile URL: [redacted BBB profile URL]

Other Platforms

• HomeAdvisor: active profile with reviews
• Angi (formerly Angie's List): active profile, positive reviews
• GuildQuality: verified profile (industry trust signal)
• BuildZoom: active contractor profile
• ProvenExpert: active reviews aggregator
• LinkedIn: company page

Review Themes (from search-scraped feedback)

• Client Principal's personal responsiveness ("always returned phone calls/texts")
• On-time daily presence of crews
• Projects completed within timeline and budget
• "Surpassed expectations"
• "Professional, responsive, and went above and beyond"
• Design quality ("Client Principal's design ideas were excellent")

Editorial Coverage / Backlinks From

• San Diego Union-Tribune (DR 86)
• This Old House (DR 86)
• (Plus various trade and remodeling industry publications via existing backlinks)

Credentials & Certifications

• EPA Renovator Certified (lead-safe certified)
• BBB A+ Accredited (since 2008)
• CSLB License #[redacted] (active California contractor)
• 11× Best of Houzz winner
• GuildQuality verified

Social Media Presence

Pinterest in particular is under-leveraged for a luxury remodeling firm — that platform drives more remodeling-discovery traffic than any other social channel. Houzz (essentially Pinterest for home design) is already a strong channel for Client Company.

What This Tells Us

Client Company is a legitimate, credentialed, well-reviewed luxury design-build firm with 26 years of operation, A+ BBB, 5-star reviews, and a decade of consecutive Best of Houzz awards. The brand is not the problem.

The disconnect between this strong real-world reputation and the near-zero non-branded organic search visibility is where the opportunity sits. The site isn't invisible because the firm is unworthy — it is invisible because of fixable technical and link-equity problems layered on top of a migration where legacy URLs were not redirected.

For a firm doing $80K–$1M projects, the math of even a single additional qualified organic lead per month is material at this price point.

Competitor Landscape

source: competitors/landscape.md

Competitive Landscape — Client Company

Data source: Ahrefs Site Explorer + manual research, 2026-05-21 Market: San Diego luxury residential design-build

Real Competitor Set

Note: Ahrefs' algorithmic competitor list for client-example.invalid returns mostly directories (BBB, MapQuest, Yelp, LinkedIn) because the site currently ranks for so few keywords that real competitive overlap can't be measured. The list below is based on manual market research of San Diego luxury design-build firms.

Key benchmark: 123 Remodeling

• DR 32, 354 keywords, 125 in top 3, 2,171 monthly organic visitors
• This is the traffic ceiling in this category — proves the market opportunity exists

Key benchmark: Jackson Design & Remodeling

• DR 45, 259 keywords, 2,281 monthly organic visitors
• $15K/month estimated traffic value
• Founded 1989 (similar tenure to Client Company)
• This is the brand ceiling — Client Company's reputational assets are comparable

Key benchmark: Kaminskiy

• DR 37, 94 keywords, 55 in top 3, 198 monthly organic visitors
• Smaller traffic but higher quality (most keywords ranked top-3)
• A realistic 6-month target profile

The Gap Math

The gap is large. In non-branded organic search, Client Company is not currently competitive with the firms ranked above it in this market.

Competitor Strengths (What They're Doing Right)

Jackson Design & Remodeling

• Strong brand recognition built over decades
• Heavy investment in showroom + design center
• Robust portfolio + blog content
• Quality backlink profile (DR 45 indicates real editorial coverage)

123 Remodeling

• Highest keyword count of the set — clearly invested in programmatic/long-tail SEO
• 125 top-3 rankings = ruthlessly executed local + service-area targeting
• Lower DR but more keywords = working harder on on-page than off-page

Kaminskiy

• High top-3 ratio (55 of 94 = 58%) — high quality over quantity
• Focused content strategy
• Strong design-build positioning

Where Client Company Beats Them (or Can)

1. Brand & Awards

Client Company has 11× Best of Houzz vs none of the others showing as prominently. This is a major differentiator that's not currently being marketed on the site or used in title tags / schema / structured data.

2. Project Tier

$80K–$1M typical job size is higher-end than most directory listings of competitors. The natural positioning is against the top tier (Jackson, custom-home builders) — not the middle tier (Kaminskiy, 123 Remodeling).

3. Service Breadth

22 location pages + 28 service pages is broader than most competitors. This is dormant inventory ready to rank with the right authority + on-page work.

4. Specialty Services

ADUs, wine cellars, smart homes, solar integration — many competitors don't have dedicated pages for these. Lower-competition, higher-margin niches.

5. Coastal-Specific Expertise

The "coastal living finishes" content angle is genuinely valuable for La Jolla / Del Mar / Encinitas / Carlsbad markets and competitors aren't owning it.

Competitive SEO Strategy Notes

Don't compete head-on for "kitchen remodel san diego" yet

• KD 35, dominated by directories + the entrenched DR 45 firm
• Wait for DR recovery to 25+ before going after that head term
• Focus instead on kitchen remodel [neighborhood] (×22 neighborhoods)

Own the specialty niches first

• "ADU builder [neighborhood]"
• "wine cellar contractor san diego"
• "smart home renovation san diego"
• These have lower KD and are commercial-intent

Win on Houzz traffic (already winning, just convert it)

• 1,000+ Ideabook saves = active Houzz pipeline
• Optimize Houzz profile → website conversion
• Embed Houzz reviews + badges on key pages

Win on local pack / GBP

• The is_local: true flag on Rancho Santa Fe keyword confirms the 3-pack is competitive territory
• GBP optimization + review velocity could capture local-pack share before site-wide organic recovers

Use editorial credentials defensively

• Best of Houzz 2026, 2024, 2023, 2022, 2021, 2020, 2019, 2017, 2016 — surface these on every key landing page with schema
• "26 years in business" + CSLB license — trust signals work for $80K+ purchase decisions

Ahrefs — Full Report

source: reports/full/01-ahrefs.md

Full Ahrefs Report — client-example.invalid

Pulled: 2026-05-21 Source: Ahrefs Site Explorer API v3 (US market, subdomains mode)

Important note on Ahrefs traffic estimates throughout this report: Ahrefs traffic numbers for this domain undercount real performance by roughly an order of magnitude versus Google Search Console actuals. Per GSC (authoritative), real organic clicks are ~51/month — 870 over the trailing 16 months, not the "1 visitor/month" Ahrefs reports. Treat Ahrefs traffic figures as directional only. Ahrefs data on Domain Rating, referring domains, backlinks, and anchor profile is reliable and is the primary source for those sections.

1. Domain Authority — Current State

2. Domain Rating — 29-Month History

The DR history reveals an algorithmic shock in late 2025:

Translation: Between September 2025 and January 2026, Google effectively stripped a significant portion of the link equity from the site — likely due to algorithmic devaluation of toxic links (the SEOExpress.org / Telegram-channel / PBN clusters). DR bottomed at 4.5 in January 2026 and has now recovered to 14 as of May 2026 — back to the pre-crash baseline.

3. Referring Domains — 29-Month History

The May 2026 spike from 180 → 288 referring domains in 30 days is statistically extraordinary and indicates either:

• An active negative SEO attack
• A new round of paid-vendor link buying that hasn't been algorithmically devalued yet
• Aggregator/scraper sites copying content with backlinks

Cross-referenced with the spam-flagged domains in the backlink data: notion.site (May 6), web.app (May 1), factmags.com (May 12), rankyour.website (May 8), rank-top.click (May 8) — all newly seen in May 2026 and flagged as spam. This is an active spam injection event.

4. Organic Traffic — 29-Month History (US)

The Ahrefs-estimated crash registered in November 2024 — traffic in the Ahrefs model went from 138 → 77 → 16 → 3 in three months.

Important context: Per GSC, real organic clicks are ~51/month currently and 870 over the trailing 16 months — substantially higher than the "1/month" Ahrefs shows. The shape of the Ahrefs curve (sharp drop in late 2024, sustained suppression since) is consistent with the algorithmic story, but the absolute numbers materially understate reality. GSC is authoritative for actual organic clicks; Ahrefs is authoritative for the link/DR profile.

5. Full Referring Domains — Top 100 by DR

Tier 1 — High-Quality (DR 80+) — Keep These

Tier 2 — Mid-Quality (DR 60–80) — Mixed

30+ additional domains in this range, of which ~10 are flagged spam:

• viesearch.com — DR 73, 490 dofollow links, flagged spam
• linkcentre.com — DR 73, 11 links, flagged spam
• rankyour.website — DR 73, new May 2026, flagged spam
• seoexpress.org — DR 73 (the testimonial-spam vendor itself)
• rank-top.click — DR 72, flagged spam
• best-seo-domains.com — DR 60, flagged spam
• socialsocial.social — DR 64, flagged spam
• factmags.com — DR 73, new May 2026, flagged spam
• backlinker.shop — DR 65, flagged spam
• ispionage.com — DR 67, flagged spam
• weboworld.com — DR 59, flagged spam

Tier 3 — Citation / Directory Cluster

Real local SEO citations include: gte.net (27 links), neustarlocaleze.biz (5,650 links — massive), cityof.com (768 links), stoneworld.com (24 links), etc. These are normal local SEO infrastructure.

6. Toxic Link Footprint — Full Disavow Recommendation

Based on a strict reading of the data, the disavow target list should include:

Total disavow candidates: ~150 domains (with overlap across categories — roughly 55% of the current 267-domain referring set).

After disavow + 90-day reconsideration, expected DR recovery: 14 → 22–30 organically as Google reweights the remaining legitimate links.

7. Anchor Text Distribution (Top 25)

Anchor profile health

• Natural URL/branded: 57% — healthy
• Generic/empty/PBN: 25% — toxic (140 domains, 752 links combined: 73 empty-anchor + 67 single-character .)
• Exact-match commercial: 15% — toxic (industry safe range: <5%)
• SEOExpress testimonial cluster: 57 domains carrying identical paid-vendor anchor text
• Spam injection: 3% — toxic (8-domain Telegram cluster + assorted)

8. Paid Search History

Ahrefs detected paid spend across a window roughly Sep–Nov 2024:

And brief small spend in 2025: ~$1/mo for several months suggesting test campaigns.

Cross-referenced with GA4 paid search channel: the real paid campaigns ran much longer (Mar 2025: 166 sessions, Jul–Oct 2025: 60–216 sessions/mo, Nov 2025: 58 sessions). Ahrefs missed most of the campaigns. GA4 is the authoritative source for paid data.

9. Pages by Backlinks (Top 30)

The site has zero pages in any "high-traffic" Ahrefs bucket beyond pos-1. Only 1 page (the homepage) earns >0 traffic in their bucket model. This reflects the catastrophic ranking position the site is in.

Page-by-traffic distribution:

• 0-100 traffic: 1 page
• 100-1k traffic: 0 pages
• 1k-5k traffic: 0 pages
• 5k+ traffic: 0 pages

A healthy site at Client Company's brand level should have 10–20 pages in the 100-1k bucket.

10. Competitive Benchmark (US, May 2026 — per Ahrefs)

Per GSC, real organic clicks are ~51/month (870 over the trailing 16 months) — not 1/month. The DR and keyword-count gaps versus competitors are real and directionally accurate; the absolute traffic gap is narrower than Ahrefs implies but still substantial.

11. Summary

Where client-example.invalid stands per Ahrefs data:

• Substantial DR deficit vs all three competitors (DR 14 vs DR 32–45)
• Substantial keyword-count and ranking deficit (2 keywords vs 94–354)
• An algorithmic shock that erased link equity from Sep 2025–Jan 2026 (DR dropped from 14 to 4.5, now recovered to 14)
• A currently active spam injection event (May 2026: +108 referring domains in 30 days, many flagged spam)
• A backlink profile where ~150 domains (roughly 55% of the live referring set) are disavow candidates
• An anchor text profile that's ~18% toxic by share (industry safe: <5%)

The site is recoverable. Comparable luxury-remodeling firms in San Diego achieve DR 30–45 with consistent editorial work and clean link profiles. The site already has the editorial footprint (SDUT, This Old House, BBB, Houzz) that should carry it once toxic links are disavowed and on-page recovery work is completed.

Backlink Toxicity

source: audit/backlinks-toxicity.md

Backlink Profile & Toxicity Audit — Client Company

Data source: Ahrefs Site Explorer, pulled 2026-05-21

Headline Numbers

A profile of 267 referring domains should support a DR much higher than 14. DR sitting at 14 indicates Google is discounting most of these links — a classic toxic-profile signature. The recent +108-domain jump in May 2026 is consistent with an active negative-SEO / spam injection event rather than legitimate link growth.

Critical Finding 1: Paid-Link Vendor Spam Cohort

57 referring domains all carry an identical or near-identical anchor text:

"There was a time when client-example.invalid struggled with low visibility and fewer than 100 visitors a month. Once I tried SEOExpress.org's backlink building service, everything changed within weeks! My organic traffic increased by an astounding +300% and I've secured five first-page rankings since then. Honestly, it's nice feeling seen on Google again—give SEOExpress.org a try if you're ready for ..."

This is a paid-link vendor testimonial-spam network. SEOExpress.org placed Client Company-branded testimonials across 57 sites as part of their own promo, using the client-example.invalid URL as social proof. Whether or not this site authorized the placement, the pattern is flagged by Google's link-spam systems because:

• Identical anchor text across 57 unrelated domains
• The anchor text itself is promotional copy unrelated to actual services
• The source domains (top DR 35) are network-style properties

Action required: These 57 domains should be disavowed. Authorization is not the operative question — Google is discounting the profile either way.

Critical Finding 2: Telegram Spam Channel Cohort

8 referring domains carry the anchor:

"OUR TELEGRAM CHANEL https://t.me/s/quarterlinks25"

(CHANEL misspelled — characteristic of low-effort link networks.)

This is a link-injection / hacked-content footprint. Domains pushing crypto/Telegram spam links alongside legitimate sites' URLs. Confirmed toxic; disavow immediately.

Critical Finding 3: Confirmed Paid-Link Vendor Domains

Direct hits in the referring domain set:

Critical Finding 4: Over-Optimized Exact-Match Anchors

A natural backlink profile has ~5–10% exact-match commercial anchors at most. This profile shows 9+ exact-match clusters totaling 100+ links — a textbook Penguin/SpamBrain trigger.

Notably, the "interior painting" anchor (11 domains) targets a service the site does not prominently offer, suggesting an undifferentiated, vendor-driven link campaign that did not match real service offerings.

Critical Finding 5: Generic/Empty Anchor Footprint

• 73 domains link with empty anchor ("") — 464 links total
• 67 domains link with anchor . — 288 links total

This is a classic PBN footprint — automated link placements where the anchor was either a single character or omitted. 140 domains in the profile carry this footprint, accounting for 752 links.

The Legitimate Portion

Not everything is bad. Quality referring domains:

These 10 alone, plus the ProvenExpert / Houzz / yellow page citations, would justify a DR of 25–30 on their own. The toxic load is what dragged the profile to its January 2026 low of 4.5; it has since recovered to 14, but the toxic anchors and PBN footprint remain in place and continue to suppress rankings.

Disavow Strategy

Recommended action: Build and submit a Google Disavow file targeting roughly 150 domains in the following cohorts:

1. 57 domains — SEOExpress.org testimonial cohort (filter Ahrefs by that anchor text)
2. 8 domains — Telegram-channel cohort
3. 22 domains — confirmed spam (is_spam: true, includes buybacklinks.agency, za.com, etc.)
4. 73 domains — empty-anchor PBN footprint (case-by-case review; majority disavow)
5. 67 domains — single-character . anchor PBN footprint
6. Any exact-match commercial anchor sources with DR < 25 and no editorial/local justification

Estimated total disavow list: ~150 domains (with overlap, ~55% of current referring domain count).

Expected outcome: Post-disavow, DR should rise organically as Google re-weights the remaining legitimate links. Combined with on-page recovery and new editorial links, DR in the 25–30 range within 6 months is realistic given the existing editorial footprint (SDUT, This Old House, BBB, Houzz).

Anchor Profile (Top 25)

Anchor profile health summary

• Natural URL/branded: ~55% — healthy
• Generic/empty/PBN: ~25% — toxic
• Exact-match commercial: ~15% — toxic (should be <5%)
• Spam injection: ~3% — toxic
• Branded text: ~7% — healthy but underweight

Keywords & Rankings

source: audit/keywords-rankings.md

Keyword Rankings & Recovery Targets — Client Company

Data source: Ahrefs Site Explorer (US), pulled 2026-05-21

Current Ranking State

Per Ahrefs, the site currently ranks for only 2 organic keywords in the US:

• 0 keywords in positions 1–3
• 0 keywords in positions 4–10
• 1 keyword in positions 11–30 (the branded one)
• 1 keyword in positions 31–50 (Rancho Santa Fe)

Caveat: Ahrefs traffic/keyword visibility for this domain materially undercounts reality. Google Search Console (authoritative) shows ~51 organic clicks/month and 870 clicks over the trailing 16 months — meaning real ranking impressions exist across more queries than Ahrefs surfaces. The two keywords above are Ahrefs-detected only; GSC will show a longer tail.

Even accounting for that, for a site with 50 commercial-intent pages (28 service + 22 location), measurable ranking visibility is far below where it should be. A healthy local design-build firm baseline: 150–400 ranking keywords with meaningful top-10 presence.

Recovery Target Set — Tier 1 (Quick Wins)

These keywords have existing pages, low difficulty, and high commercial intent. Recovery should start here.

Service-Area Keywords (existing /location/ pages)

22 location pages × avg ~50 vol × position 5 capture rate = realistic 200–500 monthly organic visitors from location pages alone, once recovered.

Service Keywords (existing /services/ pages)

Recovery Target Set — Tier 2 (Mid-Term)

Higher difficulty, but justified by the brand's actual project quality.

Keyword Gap vs. Competitors (per Ahrefs)

Per GSC, the site's real organic clicks are ~51/month, not the 1/month Ahrefs estimates. Competitor numbers above are from Ahrefs and likely also understate slightly, but the DR and keyword-count gaps are real and directionally accurate.

A reasonable 12-month target: 150 ranking keywords with 30+ in top-10 positions, putting the site between Kaminskiy and Jackson tiers.

Local SEO Opportunities

Per Ahrefs, home remodel rancho santa fe is flagged is_local: true — meaning the 3-pack / Local Pack is in play. GBP (Google Business Profile) optimization status is not visible from this audit (requires GBP-side access) but is a major lever:

• GBP-driven traffic for local-pack queries can match or exceed organic-page traffic
• The existing reviews flywheel (10 Yelp reviews + ProvenExpert + Houzz reviews) provides social proof
• Geo-tagged project photos from the 18 portfolio pages can populate GBP

Recommendation: A full GBP audit + optimization should run in parallel with on-page recovery.

Featured Snippet / SERP Feature Opportunities

The blog covers strong informational queries that could earn featured snippets:

• "how to pick the right finishes for coastal living" → "what finishes work best in coastal homes" snippet
• "guide to smart home systems san diego" → snippet for "what is a smart home system"
• "guide to solar integration remodels" → snippet for "can you add solar to a home remodel"
• "best sunroom materials southern california" → snippet for "what's the best material for a sunroom"

These are not currently captured because the blog content reads templated / AI-generated and lacks the citation density and original POV that snippets favor. Editorial uplift on the existing 39 posts (versus writing more) is the higher-ROI move.

SEO Traffic Analysis

source: audit/seo-traffic-analysis.md

SEO & Traffic Analysis — Client Company

Data source: Ahrefs Site Explorer (US market, subdomains mode), pulled 2026-05-21

Current Organic Snapshot

Important note on traffic numbers: Ahrefs traffic estimates for this domain undercount real performance by roughly an order of magnitude versus Google Search Console actuals. Per GSC (authoritative), the site is receiving ~51 organic clicks/month and has logged 870 clicks over the trailing 16 months — not the "1 visitor/month" Ahrefs reports. Throughout this report, Ahrefs traffic estimates should be read as directional only. Ahrefs data on backlinks, Domain Rating, anchor profile, and referring domains remains reliable and is used as-is.

While not invisible, the site is significantly underperforming what its brand, page count, and link profile should support.

The Traffic Collapse Timeline

Monthly organic traffic (Ahrefs estimate, US market, last 41 months — see GSC caveat above; treat shape of the curve as more reliable than absolute values):

Diagnosis

The crash between Nov 2024 and Dec 2024 is sharp, sustained, and never recovered — classic signature of an algorithmic penalty (likely tied to Google's late-2024 link-spam updates and SpamBrain) compounded by a site migration in April 2026 that broke URL equity.

Two compounding causes, not one.

Current Ranking Keywords (All 2 of Them)

Zero non-branded keywords in the top 30. This means the existing service pages (kitchen remodeling, bathroom remodeling, whole-home) and the 22 location pages are not ranking for any of their target queries.

Top Pages by Traffic

Only two pages are generating any measurable organic exposure. The remaining 86 URLs are effectively dormant in search.

What Used to Rank (Implied)

At peak (Q1 2024 — 154 monthly organic visitors per Ahrefs estimate, $540/mo estimated traffic value), the site was almost certainly ranking for:

• Local commercial: kitchen remodel san diego, bathroom remodel la jolla, home addition contractors san diego
• Hyper-local: variations across 22 service-area pages
• Branded long-tail: design-build queries

That asset is recoverable — the pages still exist, the brand exists, and competitors sit at DR 32–45. The barriers are the spam-link load and the migration damage, both of which are addressable.

Site-Level SEO Hygiene

Pulled from https://client-example.invalid/robots.txt and sitemap index:

Robots.txt: Standard WordPress + Yoast setup. Blocks /wp-admin/, allows AJAX. No critical issues.

Sitemap index: https://client-example.invalid/sitemap_index.xml ✅

• Posts sitemap (lastmod 2026-04-25) — 39 URLs
• Pages sitemap (lastmod 2026-05-04) — 49 URLs
• Categories sitemap (lastmod 2026-04-25)

Total indexable URLs: ~88 (consistent with manual count from sitemap)

Last-modified pattern: Almost every URL shows lastmod between 2026-04-17 and 2026-05-04 — confirming a full site rebuild/migration completed in late April 2026. This single fact explains a huge portion of the current invisibility — Google has not yet re-crawled, re-indexed, and re-evaluated the new URL structure, and any equity not preserved via 301 redirects is being burned.

Recovery Priorities

1. GSC verification to confirm crawl/index status and pull manual action / security issue data
2. 301 redirect audit — almost certainly the highest-leverage single fix
3. Disavow file submission targeting the ~150 toxic domains identified (57 SEOExpress testimonial cluster, 73 empty-anchor PBN, 8 Telegram cluster, 22 confirmed-spam, plus single-character anchor footprint)
4. On-page refresh of the 50 commercial pages (28 services + 22 locations)
5. Re-indexation push via GSC URL Inspection on top 30 priority pages

Google Search Console — Full Report

source: reports/full/02-gsc.md

Google Search Console — Organic Search Performance

Property: sc-domain:client-example.invalid Access level: siteFullUser Window covered: 2025-01-21 → 2026-05-19 (16 months — the maximum GSC allows) Pulled: 2026-05-21

Note on data source: Google Search Console is the authoritative source for organic search performance on client-example.invalid. Third-party tools like Ahrefs estimate traffic from sampled SERP crawls and undercount real performance — often by an order of magnitude. Where this report cites traffic, impression, or click figures, the numbers come directly from GSC. The 16-month window is critical: a 90-day snapshot misses the long-tail signal that only becomes visible across a full year-plus.

Coverage caveat: GSC reflects only what Google has served. Searches on Bing, DuckDuckGo, and other engines are not included in this dataset.

1. 16-Month Top-Line Performance

Across the past 16 months, Google has shown client-example.invalid in search results ~41,000 times and earned 870 clicks — averaging about 54 clicks per month, with the trailing average closer to 51. The bottleneck is position: at an average of 49.6 (page 5), the site is being surfaced but not at depths where searchers click.

Branded queries (those containing "Client Company") account for roughly 66% of all clicks — the brand name does the heavy lifting. Non-branded discovery searches (e.g., "bathroom remodel san diego") drive only a small fraction of traffic today.

2. 28-Day Trend — Post-Rebuild Stabilization

The April 2026 rebuild expanded indexation — Google is showing the site to more searchers (+31% impressions). Clicks held flat-to-positive (+4%). The rebuild stabilized the site rather than collapsing it; impressions are growing, not crashing.

The trade-off: more URLs are now ranking somewhere, but the average position drifted deeper (50 vs 41) because many of those new impressions come from deep-page rankings. The story is not "the site is broken" — it's "the site is competing for more queries but at ranks too deep to convert."

3. Branded vs Non-Branded Split (16 months)

Branded Queries

Branded total: 288 clicks (~33% of all clicks via the named queries above; ~66% when all "Client Company"-containing variants are counted). Brand defense is in good shape.

Non-Branded Clicks (only 23 total in 16 months)

Twenty-three non-branded clicks across 16 months works out to roughly 1.4 per month. The site is effectively invisible for the non-branded discovery searches that drive new-customer pipeline.

4. High-Leverage Recovery Targets — Ranks But Doesn't Get Clicks

These are queries where Google is already showing the site enough impressions to matter, but the ranking is too deep to convert. Each represents a high-leverage opportunity: the demand is documented, only the position needs to move.

Combined impressions across these queries: 12,000+. Roughly twelve of them sit in positions 30–60 — close enough that moving to page 1–2 is realistic with content and on-page work. If even a small share of these moved into striking distance, the result would be a step-change in non-branded traffic rather than incremental improvement.

5. Almost-Wins — Queries Already On Page 1

These already rank inside the top 10 but earn few or no clicks, typically because the queries are very low-volume or local-pack/SERP-features absorb the click:

The takeaway: the site can rank on page 1 when content is tightly targeted (e.g., "bathroom design," "basement remodel," "design and build near me" at position 1). The capability exists; it just isn't aimed at the queries that move the business.

6. Top Pages by Clicks (16 months)

Combined, the homepage (bare URL + GMB-tagged variant) accounts for 571 of 870 clicks — about 66% of all organic clicks. The site has essentially one ranking destination today.

7. Pages With Heavy Impressions But Almost No Clicks

These pages earn substantial Google exposure but rank too deep to capture it. They represent the largest pool of unrealized traffic on the site:

Nearly 40 pages with 1,000+ impressions sit at positions 40–90. Moving even a handful from page 5–9 to page 1–2 would meaningfully change site-wide click volume from existing demand.

8. Legacy URL Duplication — The Highest-Priority Technical Issue

The April 2026 rebuild introduced cleaner URL structures (/services/..., /service-areas/..., /about, /contact-us), but the legacy URLs were never 301-redirected. Both versions are now indexed and earning impressions for the same queries, which splits Google's ranking signal and pits the site against itself in search results.

The Redirection plugin is installed on the site but currently unused — no 301s are configured.

That's 25+ legacy URLs still active alongside their replacements. Every duplicate splits ranking equity and confuses Google about which page to surface for a given query.

Pages indexed that shouldn't exist

• /custom-bathroom-remodel-guide/ — 2,396 imp, no clicks (legacy ghost)
• /bathroom-remodeling-2/ — 3 imp (duplicate of /bathroom-remodeling/)
• /category/uncategorized/ — 2 imp (default WordPress category, not used)
• /blog/page/2/?e-page-2b7da6f=4 — 2 imp (Elementor cache fragment)
• /m.client-example.invalid/ — 2 imp (mobile subdomain reference)
• /kitchen-renovation-contractors (no trailing slash) — 21 imp

9. Sitemap Clutter

Five sitemaps are registered in GSC, only one of which should remain:

Google is currently being handed five overlapping sitemap signals describing the same site. Four of them are stale (one of them by 17 months) and one carries an unresolved warning. Cleaning this up to a single authoritative index removes a meaningful crawl-budget tax.

10. Device Performance (90 days)

Mobile users see better positions and higher CTR than desktop, yet the site's impression base is heavily desktop-skewed. For a home-services business, mobile would typically drive 60–70% of impressions, not 13% — suggesting the desktop base may include a meaningful share of non-buyer traffic (bots, internal users).

11. Country Distribution (90 days)

Filtering to US-only puts genuine domestic organic traffic at ~99 clicks per 90 days ≈ 33/month.

12. Search Appearance

The site is appearing as a product snippet in some results (Yoast schema is producing it) but earning zero clicks from that surface. Worth investigating whether the schema is misconfigured or just attached to low-intent queries.

13. GSC-Side Fix Priorities

GSC Overview

source: research/gsc/overview.md

GSC Performance Overview — Authoritative Organic Search Data

Property: sc-domain:client-example.invalid Access level: siteFullUser Pulled: 2026-05-21

GSC is the authoritative source for organic search performance. Third-party SEO tools (Ahrefs included) estimate traffic from sampled SERP crawls and undercount the real picture — often by an order of magnitude. The numbers below come directly from Google's first-party data.

The 16-month view matters. A 90-day snapshot misses the long-tail signal that only becomes visible across a full year-plus. Where this overview cites 16-month totals, that's why.

16-Month Top-Line (2025-01-21 → 2026-05-19)

Google has surfaced client-example.invalid roughly 41,000 times across 16 months and earned 870 clicks. The pattern: discoverable, but ranking too deep to convert exposure into traffic — and the bulk of clicks that do land come from people who already know the brand.

28-Day Comparison

Reading the numbers

• Impressions grew 31% — Google is showing the site to more searchers, not fewer. Post-rebuild content IS being crawled and surfaced.
• Clicks effectively flat (+4%) — but impressions grew 31%, so CTR fell. The site shows up; the listings don't get clicked.
• Position degraded ~9 spots — most of the new impressions came from queries where the site sits position 50–60, which generates impressions but few clicks.
• The April 2026 rebuild stabilized organic search performance. It didn't crater rankings, but it also didn't immediately lift them — the bigger leverage is in competitive ranking and URL hygiene, not in crawl health.

90-Day Performance (Feb 20 – May 19)

Caveat: The 90-day window misses long-tail signal. The 16-month view above is what reveals the full demand picture — at 90 days you see roughly 100 queries; at 16 months you see 300+.

Daily Click Trend (Feb 20 → May 19)

Notable spikes:

• 2026-04-20: 13 clicks (day of site rebuild — likely indexation refresh activity)
• 2026-04-21: 8 clicks
• 2026-04-22: 6 clicks
• 2026-04-14: 5 clicks
• 2026-04-30: 4 clicks
• 2026-04-24: 4 clicks

Floor: 0 clicks/day is common; ceiling 1–3 clicks/day is the usual range.

No catastrophic dropoff is visible in the GSC window. The April 2026 rebuild did not destroy click volume — impressions actually grew in the post-rebuild window. The site is in a stable post-rebuild state; the larger story is competitive ranking depth, not crawl damage.

Device Split (90 days)

Observations

• Desktop dominates impressions (87%) — atypical for a home-services business where mobile usually drives 60–70%
• Mobile CTR (2.16%) is better than desktop (1.21%) — when mobile users see the listing, they engage more
• Mobile position is better (40.3 vs 46.8) — Google ranks the site higher on mobile but shows it less often
• The desktop-heavy profile is consistent with bot or internal traffic making up a meaningful share of indexed exposure

Country Split (90 days)

Critical finding

32 of 143 total clicks (22%) come from the Philippines — with an unrealistically high 37.65% CTR and a near-#1 position. These are almost certainly internal users (development partners, freelance team members, or scrapers) consistently searching branded queries from PH IPs. This matches the GA4 traffic from Meycauayan, Philippines (53 sessions, 9 key events).

Practical implication: US-only clicks = 99/90 days, not 51/month. After filtering bot/internal traffic, real US organic traffic is ~33 clicks/month from genuine US searchers.

Coverage caveat: GSC counts only what Google has served. Bing, DuckDuckGo, and other engines aren't in this dataset, so total search traffic is somewhat higher than what GSC reports.

Performance Summary

The site is in a "low-clicks, decent-impressions, deep-page rankings" state:

• Discoverable: Google indexes 97+ pages and surfaces them for 100+ queries (300+ across 16 months)
• Not findable: Average position 45–50 — searchers don't scroll that deep
• Stable post-rebuild: April 2026 rebuild expanded indexation rather than collapsing it; impressions are growing, not crashing
• Branded-dominant: ~66% of clicks come from people already searching the brand name

This is not a "site has been blacklisted" problem. It's a "site exists in Google but doesn't rank competitively" problem — addressable with on-page, internal linking, URL hygiene, and authority work rather than emergency triage.

GSC Top Queries

source: research/gsc/top-queries.md

GSC Top Queries — Authoritative Organic Search Data

Window: 2026-02-20 → 2026-05-19 (90 days) Rows analyzed: 100 queries Total clicks: 143 Pulled: 2026-05-21

GSC is the authoritative source for organic search queries on client-example.invalid. Where third-party tools list a handful of ranking keywords, GSC shows what Google actually surfaced. The 90-day view below captures recent demand; the 16-month view in 02-gsc.md reveals the full long-tail picture (300+ unique queries vs ~100 in 90 days).

Clicks: Branded

Branded subtotal: 52 clicks (36% of all clicks at 90 days; ~66% across the full 16-month window). These are the only queries delivering meaningful CTR.

Clicks: Non-Branded

Non-branded with clicks subtotal: 5 clicks across 90 days. The other ~86 clicks are either anonymized by Google or fragmented below threshold.

High-Impression Queries with ZERO Clicks (Recovery Targets)

These are the queries where Google IS showing the site, but the rankings are too deep to capture clicks. Each is a documented demand signal — Google has the site in its result set, it just isn't surfacing it where people look.

Note: the 16-month view (see 02-gsc.md §4) surfaces additional high-impression recovery queries that are invisible in the 90-day window — including bathroom remodel san diego at 1,413 impressions / position 78 across the full window. The 90-day numbers above are the trailing tip of a much larger long-tail.

"Almost Won" Queries (rank top 10 but no clicks captured)

These are searches where the site is already on page 1 but didn't earn a click during the period — typically because the queries are very low volume, or because local-pack/Maps/ads absorbed the click:

These location pages already rank well — they just need traffic-driving optimization (sharper title tags, meta descriptions, schema, and content depth).

Long-Tail Local Patterns Detected

The site picks up impressions on a wide range of programmatic-looking queries (Rancho Penasquitos, Scripps Ranch, 4S Ranch + budget/affordable + bathroom/kitchen + remodeling). Examples:

• budget kitchen remodeling 4s ranch — 3 imp, pos 26.3
• budget kitchens rancho penasquitos — 11 imp, pos 19.3
• affordable bathroom remodeling scripps ranch — 2 imp, pos 69.5
• bathroom remodeling rancho penasquitos — 6 imp, pos 55.3

The pattern suggests programmatic / templated local pages on the site that produce a wide ranking surface. Cleaned up and properly optimized, these template patterns could become a long-tail traffic engine.

Query Strategy Implications

1. Brand defense is already in place — 6 branded queries deliver 52 clicks at 5–25% CTR. Maintenance only.
2. Real opportunity = 22+ queries currently ranked in position 30–70 with 5+ impressions each. Moving any of these to page 1 multiplies traffic.
3. Quick wins exist — cardiff home remodel (pos 8.4), attic conversion contractors san diego (pos 37), adu portfolio encinitas (pos 36), bathroom remodeling solana beach (pos 37). Specific, low-comp, locally relevant.
4. Generic terms (bathroom remodel, kitchen remodeling) sit deep on page 6–7. These will require sustained authority and on-page work to move and shouldn't be the first priority.
5. The site is invisible to high-intent buyers — nobody searching kitchen remodel san diego finds Client Company today. That's the gap to close.

Caveat: GSC counts only what Google has served. Searches on Bing and other engines aren't in this dataset.

GSC Top Pages

source: research/gsc/top-pages.md

GSC Top Pages — Authoritative Organic Search Data

Window: 2026-02-20 → 2026-05-19 (90 days) Pages with impressions: 97 Total page-level clicks: ~143 Pulled: 2026-05-21

GSC is the authoritative source for organic page performance on client-example.invalid. The 90-day window below captures recent activity; the 16-month view in 02-gsc.md shows 200+ pages with impressions — the full inventory of what's surfacing in Google.

Pages by Clicks

Notes

• The homepage takes 108 of 143 clicks (76%) between the bare URL and the GMB-tagged variant. The site has essentially one ranking destination today.
• The GMB-tagged URL getting indexed (3,227 impressions) is a technical issue — that's a Google Business Profile tracking URL leaking into the index. It should be canonicalized to /.
• 75 of the 97 pages with impressions earned zero clicks in 90 days.

Pages with Strong Position but Low Impressions

These rank well but Google doesn't show them often — usually because the targeted queries have low search volume:

Pages with High Impressions but Deep Position (Recovery Targets)

These get shown a lot but rank too deep to capture clicks. Highest-leverage on-page work targets:

Legacy / Orphan URL Findings

The April 2026 rebuild introduced cleaner URL structures (/services/..., /service-areas/...) but the legacy URLs were never 301-redirected. The result: both versions are indexed, and they compete for the same queries.

This is the URL duplication mess in microcosm: across 25+ legacy paths site-wide (see 02-gsc.md §8 for the full table), both the old and the new versions coexist in Google's index and split the ranking signal.

The site already has the Redirection plugin installed in WordPress — it's just currently unused. Building a clean 301 map there is the highest-leverage technical fix available.

Implied action

Build a redirect map: /[area]/ → /service-areas/[area]/ and /[service-slug]/ → /services/[service-slug]/. Canonicalize one URL per intent.

Image URLs Showing in Results

A few image URLs are indexed and getting impressions:

• /wp-content/uploads/2024/10/NbAjS2GDVnmK-sW6ShKA3-min.jpg — 6 imp, pos 24.2
• /wp-content/uploads/2024/10/Smart-Lighting-and-Climate-Control-Enhance-Your-Homes-Efficiency.jpeg — 4 imp, pos 51
• /wp-content/uploads/2026/03/35-min.webp — 3 imp, pos 18

Indicates an Image SEO opportunity — with proper alt text, captions, and structured image content, the site could earn meaningful Google Images traffic for visually-driven queries like "luxury kitchen remodel san diego."

Caveat: GSC counts only what Google has served. Bing and other engines aren't in this dataset.

GSC Top Movers

source: research/gsc/top-movers.md

GSC Top Movers — Post-Rebuild vs Pre-Rebuild

Comparison window: Current 2026-04-22 → 2026-05-19 (post-rebuild) vs Previous 2026-03-25 → 2026-04-21 (pre-rebuild) Site rebuild executed: 2026-04-20 Pulled: 2026-05-21

GSC is the authoritative source for these movement comparisons. Third-party rank trackers sample at intervals and miss most of the long-tail; the figures below come directly from Google.

Top Growing Queries

Reading the data

• Branded queries improved — Client Company jumped from position 13.4 to 1.5 (a major brand-search win)
• First non-branded movement — kitchen remodeling in san diego moved 7 spots toward page 1 (from 71 → 64)
• New branded discoveries — Client Company home and Client Company construction started ranking from nothing

Top Declining Queries

Reading the data

• kitchen designer san diego got worse post-rebuild (−14 positions) — a content or internal-link signal was lost
• The other two queries actually improved position but lost a click — likely SERP feature cannibalization (an ad, local pack, or AI Overview took the click)
• Net: limited damage from the rebuild itself in this window

Net Takeaway

The April 2026 rebuild did not crater rankings in the immediate window after deployment. In fact:

• Branded queries improved
• One non-branded query moved 7 positions closer to page 1
• One non-branded query declined 14 positions

The dataset is small (143 total clicks over 90 days) so the signal is noisy, but the rebuild appears to be a net neutral-to-positive event in the GSC window. Any older "ranking collapse" signal Ahrefs may report is from a prior period (mid-to-late 2025), not from this rebuild — and Ahrefs traffic estimates undercount Google's actual surface area by roughly an order of magnitude, so anchor any read on this period to the GSC numbers above.

The post-rebuild story is one of stabilization. The bigger lever now is competitive ranking depth and URL hygiene (see 02-gsc.md §8 on the legacy URL duplication issue), not crawl recovery.

What This Doesn't Tell Us

GSC's top-movers data only compares overlapping queries — it does NOT capture:

1. Queries the site used to rank for and now doesn't (those just disappear from the comparison)
2. Queries the site never ranked for and now does (no "previous" baseline)
3. Searches on Bing or other engines (GSC reflects Google only)

To get the full picture, pull a 16-month comparison once that data is referenceable in GSC's UI.

Confidence Level: Medium

The post-rebuild window (Apr 22 – May 19) is only 28 days, which is short for ranking signals to fully stabilize. The next 60–90 days will be more diagnostic. Re-run this comparison monthly.

GSC Indexation & Sitemaps

source: research/gsc/indexation-sitemaps.md

GSC Indexation & Sitemap Status

Pulled: 2026-05-21 Property: sc-domain:client-example.invalid

GSC is the authoritative source for indexation status. The figures below come from Google's first-party crawl and sitemap data.

Sitemaps Submitted to Google — 5 Registered, 4 Should Be Removed

Issues

1. Five overlapping sitemap signals. /sitemap_index.xml (current Yoast format) coexists with /wp-sitemap.xml (default WordPress core sitemap), /sitemap.xml (older Yoast format), and standalone /page-sitemap.xml and /post-sitemap.xml. Google is ingesting overlapping, conflicting URL sets describing the same site.

2. Stale submissions. /wp-sitemap.xml was last submitted March 2025 (14 months ago); /sitemap.xml from February 2025; /post-sitemap.xml from December 2024 — 17 months stale. None were re-submitted after the April 2026 rebuild.

3. One warning on /sitemap.xml — not detailed in the API response; investigate directly in the GSC UI before removal.

4. Current sitemap is being crawled regularly — last download was 2026-05-16. The infrastructure works; the housekeeping is messy.

Action items

• Delete or unsubmit /wp-sitemap.xml, /sitemap.xml, /page-sitemap.xml, /post-sitemap.xml
• Keep only /sitemap_index.xml (Yoast-managed)
• Re-submit the index after cleanup
• Resolve the warning on /sitemap.xml before removal

Crawl & Indexation Health Signals

97 URL-paths surfaced in 90 days against an 88-URL sitemap. The arithmetic implies what the page-level data confirms: legacy URLs are still in Google's index alongside their new equivalents. The April 2026 rebuild added new canonical URLs without retiring the old ones, so Google has duplicates competing for the same queries.

The WordPress Redirection plugin is installed but currently unused — meaning the infrastructure to 301 the legacy URLs is in place; the redirects just haven't been configured.

Indexation Hygiene Sequence

Step 1 — Sitemap cleanup

1. Audit and remove 4 of the 5 sitemap entries (keep only /sitemap_index.xml)
2. Investigate and resolve the /sitemap.xml warning
3. Verify /sitemap_index.xml references current sub-sitemaps only (page, post, category)
4. Submit fresh in GSC

Step 2 — Legacy URL 301 mapping

For every legacy URL still ranking, deploy a 301 redirect to its current canonical (via the Redirection plugin):

Step 3 — URL Inspection batch

After the 301s are live, use GSC URL Inspection on the top 30 priority pages to force re-crawl and re-indexation under the new URL structure.

Step 4 — Canonical tag audit

The site is Yoast-driven; canonical tags should be auto-set, but verify by:

• Sampling 10 service pages (/services/...)
• Sampling 10 location pages (/service-areas/...)
• Sampling 5 blog posts
• Confirming each <link rel="canonical" href="..."> points to itself, not to legacy URLs

What This Tells Us

The site has no manual action, no security issue, no fundamental crawl problem. Google is reading the sitemap weekly, indexing 97+ URLs (200+ across the 16-month window), and showing them in search 10,700+ times per quarter (41,000+ across 16 months).

The bottlenecks are competitive ranking depth and URL hygiene, not crawl health. That's a meaningfully different problem from "the site has been blacklisted" — it means the system is working and what's needed is consolidation, on-page work, and authority-building rather than emergency triage.

Coverage caveat: GSC reflects Google only. Bing and other engines aren't in this dataset.

Google Analytics 4 — Full Report

source: reports/full/03-ga4.md

Google Analytics 4 — Full Findings

Property: properties/[redacted GA4 property] Account: [redacted analytics account] Window covered: 2025-01-21 → 2026-05-19 (16 months) Pulled: 2026-05-21

Read this first — two caveats that change every number below

1. Bot and internal-team traffic is NOT filtered at the GA4 source. Roughly 70%+ of raw GA4 sessions are bots or internal traffic, not real visitors. Of the 12,022 sessions reported over 16 months, only about 5,000 are real human sessions. Every raw number in this report must be read with that in mind. Recommended fix: add bot- and internal-traffic exclusion filters in the GA4 data stream settings (country exclusions + IP exclusions for the Prior Vendor team).

2. Conversion tracking is broken. A generate_lead event has been configured in GA4 since March 5, 2025 — and has never fired. The purchase and s_click_to_call events also sit at zero. Only phone_click (13 fires in 90 days) and email_click (4 fires in 90 days) are working. Form submissions on /contact/ — which received 1,082 sessions and 224 phone/email clicks — are completely invisible to analytics. Every conversion count shown is a floor, not a true number.

1. 16-Month Top-Line Performance

2. Monthly Session Trend

What the trend shows

• September 2025 = peak (1,437 sessions, 149 conversion events). That peak was carried by a Google Ads Performance Max campaign — see Section 8.
• May 2026 = 535 sessions — a 63% drop from September's peak.
• Conversion events fell from 149 (Sep 2025) to single-digits per month afterward.

The conversion collapse happened before the traffic collapse. September 2025 fired 149 events; October 2025 fired 18 — an 88% drop in one month while sessions stayed flat. That pattern is the signature of broken conversion tracking, not collapsed traffic quality. Something stopped firing events around October 2025.

The April 2026 rebuild stabilized site behavior. The bigger ongoing issue is conversion-tracking failure and bot pollution, not the structural health of the site itself.

3. Channel Distribution (16 months, raw)

Channel notes

Cross-network (Google Ads Performance Max) was the highest-converting channel for the months it ran (Aug–Sep 2025):

• Aug 2025: 551 sessions, 10 events
• Sep 2025: 835 sessions, 137 events
• Oct 2025: 6 sessions (campaign stopped)

Then the campaign disappeared from the data entirely. Worth investigating who managed the campaign, what was spent, and why it was stopped.

Paid Search (text ads) was sustained but low-converting:

• Mar 2025: 166 sessions, 0 events
• Aug 2025: 216 sessions, 1 event
• Sep 2025: 139 sessions, 0 events
• Oct 2025: 211 sessions, 2 events
• Nov 2025: 58 sessions, 0 events
• Total ~853 paid search sessions, ~5 conversions → 0.6% conversion rate

That number is suspect given the measurement gaps — the underlying campaigns may have produced real leads that were never captured as GA4 events.

Direct traffic dominates the raw chart at 55%, but most is bot. After bot filtering, real Direct is likely 10–20% of the original total.

4. Top 30 Landing Pages by Sessions (16 months)

What stands out

/contact/ is the conversion engine. 1,082 sessions → 224 conversions = 20.7% conversion rate. This is the page that's working — though even those 224 events are almost entirely phone and email clicks, not form submissions (see measurement gap above).

Legacy service pages convert reasonably:

• /whole-home-renovations/ — 245 sessions → 15 conversions = 6.1%
• /bathroom-remodeling/ — 533 sessions → 7 conversions = 1.3%
• /services/ — 478 sessions → 6 conversions = 1.3%

The newer /services/[slug]/ detail pages convert at 0%:

• /services/bathroom-remodeling/ — 139 sessions, 0 events
• /services/whole-home-renovations/ — 91 sessions, 0 events
• /services/kitchen-remodeling/ — 96 sessions, 0 events
• /services/home-additions/ — 71 sessions, 0 events

These pages get real traffic but have no working conversion path — either the contact form isn't placed on them, or the events aren't firing. This is a lead-capture gap, not a traffic gap.

Portfolio pages show zero conversions despite high session counts (52–97 sessions each). No CTA tracking is in place on individual portfolio detail pages.

/san-diego-home-show/ captured 4 conversions from 52 sessions = 7.7% conversion rate. The event/landing-page combo worked.

5. Top 30 Most-Viewed Pages

6. Country Distribution — The Bot Problem

Bot vs. real breakdown

• Bots: China (2,887), Singapore (2,039), Germany (178), (not set) (133), plus smaller volumes from India, Bangladesh, Indonesia, Vietnam, Ireland, Türkiye, Iraq, and others — together roughly 6,000+ sessions.
• Internal team: Philippines (775) and Serbia (187) — both show abnormally high engagement-per-user, consistent with the previous dev team working in the site.
• Pakistan (436): suspect — could be internal or scraper traffic.
• Real target market: USA (4,573) is the genuine audience. Mexico (1,116) and Canada (100) are partially real cross-border interest.

After filtering bots and the internal Philippines/Serbia team, real human sessions are ~5,000 over 16 months — about 310 sessions/month of real traffic.

Where conversions actually happen

• United States: 93 conversions — the real target market signal.
• Mexico: 126 conversions — either bilingual San Diego–Tijuana cross-border traffic (real and valuable), or spam events firing from click-farms. Worth segmenting by source/medium and language to determine.
• Philippines: 89 conversions — almost certainly internal team testing forms.

After filtering, real conversion count over 16 months is roughly 220 (US + Mexico + Canada combined) — about 14 leads/month, or 1 lead every 22 real sessions. That's a reasonable conversion rate for a luxury remodeling brand, but the absolute volume is small and the underlying tracking is unreliable.

7. Device & Browser (16 months)

Observations

• Mobile share (iOS + Android + Webview): 23% — much lower than typical home-services norms (60%+) because bot traffic on desktop user agents dominates.
• After bot filtering, real-user mobile share should rise toward 50%+.
• Mac users engage at roughly 2× the rate of Windows users — consistent with a luxury homeowner demographic skewing Mac.
• Chrome on Linux at 19% engagement is a bot/scraper signal.

8. Conversion Events Configured

What's broken

• purchase was configured but is irrelevant to a service business and never fires.
• generate_lead is the most important event — and it has never fired since it was created in March 2025. The contact form is not pushing to this event.
• s_click_to_call never fires and appears deprecated.
• phone_click and email_click are the only working conversion events, and even their coverage is patchy across page templates.
• Contact form submissions are not captured as a discrete event anywhere in GA4.

The drop from 149 conversions (Sep 2025) to 18 (Oct 2025) to single digits/month afterward maps to the end of the Cross-network Google Ads campaign + a tracking failure around the same time.

9. Top Event Counts (90 days)

There is no form_submit or contact_submit event firing. Despite 1,082 sessions landing on /contact/ over 16 months and 224 key events from that page, those 224 are phone and email clicks — not form submissions. The consultation form on /contact/ is not tracked.

10. Critical Findings

Finding 1 — Conversion tracking has been broken since at least October 2025

generate_lead has never fired. purchase and s_click_to_call have never fired. Conversion events dropped from 149 → 18 → single-digits month-over-month while traffic stayed flat. The reported conversion count is a floor — true conversions are unknown.

Finding 2 — 70%+ of GA4 raw data is bot or internal traffic

Reports built off raw GA4 numbers overstate real traffic by roughly 2×. Internal traffic filters need to be applied at the data stream level:

• Country exclusions: Singapore, China, Germany, Pakistan
• IP exclusions for the Philippines/Serbia internal team
• Common AWS / data-center IP range exclusions

Finding 3 — A Google Ads Performance Max campaign delivered 147 conversions in Aug–Sep 2025, then disappeared

This is the single most successful marketing event in the 16-month window. The campaign stopped, and the conversion-tracking break that followed makes it hard to evaluate the campaign's true ROI. Questions worth answering:

• Who managed the campaign
• What was the spend
• Why it was stopped
• Whether the underlying tracking attribution was wired correctly

Finding 4 — The new /services/[slug]/ pages have ZERO conversions

The newer service detail pages (/services/bathroom-remodeling/, /services/kitchen-remodeling/, etc.) get real traffic but earn no conversions. The CTAs are missing or events aren't firing on these pages — a higher-priority fix than building new content.

Finding 5 — Mexico shows 126 conversions

Either real bilingual San Diego–Tijuana cross-border interest (a genuine market signal), or click-farm spam. Worth segmenting by source/medium and language to determine which. If real, there's a Spanish-language audience opportunity here.

Finding 6 — Portfolio pages have a zero-conversion floor

Despite engaged visitors (50–250+ second sessions on portfolio pages), zero conversions. The "contact us about a project like this one" CTA is either missing or not tracked.

Finding 7 — Direct traffic appears to dominate, but most is bot

Direct sits at 55% of raw sessions. After filtering, Direct likely drops to ~10–15%, making Organic Search the #1 real channel.

Finding 8 — GMB (Google Business Profile) is a distinct traffic stream worth tracking separately

123 sessions over 89 days come in via ?utm_source=gmb from the Google Business Profile listing. Today this gets blended into "organic" in standard reports — but it's a separate channel with its own management surface (the GBP listing itself) and deserves its own line in reporting.

Finding 9 — AI-engine referrals are showing up

ChatGPT, Claude, and Copilot are each sending small but real traffic to client-example.invalid. This is first-party evidence that content is being cited by LLMs in user conversations — an early-mover signal worth tracking as a discrete channel going forward.

11. Recommended GA4 Configuration Improvements

GA4 Overview

source: research/ga4/overview.md

GA4 Overview — First-Party Site Behavior

Property: properties/[redacted GA4 property] (https://client-example.invalid/) Account: [redacted analytics account] Window: 2026-02-20 → 2026-05-19 (89 days) Pulled: 2026-05-21

Two caveats that frame every number below

1. Bot and internal traffic is NOT filtered at the GA4 source. Roughly 70%+ of raw sessions in this property are bots or internal-team traffic. Every "raw" number below is inflated, and the real-human session estimate is materially smaller. Bot filtering needs to be applied in the GA4 data stream settings.

2. Conversion tracking is broken. The generate_lead event has been configured in GA4 since March 5, 2025 and has never fired. purchase and s_click_to_call also sit at zero. Only phone_click and email_click are firing. Form submissions on /contact/ are completely invisible — the 17 events shown in the 89-day window are a floor, not a true count.

Headline Numbers (89 days, raw)

Reading the numbers

• Real session volume is much higher than Ahrefs estimated (~24 raw sessions/day on average).
• Engagement is poor — 75% bounce rate, only 24% engaged sessions — but a large share of those sessions are bots that drag the average down.
• Reported conversions are minimal — 17 key events in 89 days, and form submissions are not captured at all.

Traffic Quality Is Severely Distorted by Bots

This is the most important finding in the GA4 data. The reported traffic is dominated by bot, scraper, and overseas-developer activity, not by real San Diego homeowners.

Geographic breakdown (top sources)

Filtered estimate (89-day window)

• Bot / non-human sessions: ~1,500+ (70%+ of total)
• Real US sessions: ~250 (12% of total)
• Real San Diego–area sessions: ~140 / quarter ≈ 47 / month
• Internal team activity (PH, ClickUp/Worksection referrals): ~70 / quarter

Why this matters

1. All "site traffic" numbers must be filtered before they're used to make decisions. Raw GA4 sessions vastly overstate real audience reach.
2. Conversion rate cannot be calculated from raw sessions — only from real US human traffic.
3. Bot filtering must be implemented at the GA4 level: exclude internal IPs, exclude known bot-source countries (Singapore, China, Germany, Pakistan), exclude common AWS / data-center subnets.
4. 9 of the 17 key events fired in this window came from the internal Philippines team — those are not real conversions, they're staff testing forms.

Daily Session Trend

Session volume is stable around 20–40/day with occasional spikes:

Rebuild-period spike

April 17–24 shows a huge pageview/session ratio (often 7–25 PV/session vs the normal 1–2) — consistent with the development team testing the new site post-launch. This distorts engagement metrics for that period.

16-Month Context

This 89-day window sits inside a larger 16-month trend (Jan 2025 → May 2026):

• Total raw sessions: 12,022
• Real human sessions (after bot filter estimate): ~5,000
• Peak month: September 2025 — 1,437 sessions, 149 conversion events (driven by a Google Ads Performance Max campaign)
• Conversion break: Around October 2025, conversion events collapsed from 149 → 18 → single digits/month, while raw traffic stayed flat. This signature is broken tracking, not collapsed traffic quality.
• Current bottom: May 2026 (partial) — 535 sessions.

The April 2026 rebuild stabilized site behavior; the bigger ongoing issue is conversion-tracking failure and bot pollution, not site health.

Engagement Quality

When we look at engagement metrics excluding the rebuild week and the bot floods:

These adjusted numbers are reasonable for a service-business homepage, but show the site is not pulling people deep into the content. Service pages, portfolio pages, and contact pages should be earning more downstream traffic from the homepage.

Device Split

Critical observation

A 90/9 desktop/mobile split is the opposite of what a healthy home-services site shows. Industry benchmark: 60–70% mobile for residential remodeling. This split confirms bot/dev dominance — bots run on desktop user agents.

After bot filtering, mobile share should rise substantially. Real-user mobile is likely 60%+, which makes the site's mobile PageSpeed of 53/100 a higher-priority issue than the raw numbers suggested.

Channel Note — Distinct Traffic Streams to Track Separately

Two traffic streams deserve their own line in reporting going forward:

• GMB (Google Business Profile): 123 sessions / 89 days from ?utm_source=gmb. Today this gets lumped into "organic" — it deserves its own channel.
• AI engines (ChatGPT, Claude, Copilot): small but real referral traffic. First-party signal that the site is being cited in LLM conversations — early-mover indicator worth tracking as a custom channel group.

See sources.md for the full channel breakdown.

Comparison to Ahrefs / GSC

The three measurement systems give wildly different answers because they measure different things:

• Ahrefs estimates from public SERP scraping (undercounts everything).
• GSC reports actual Google clicks (most accurate for organic Google search only).
• GA4 reports everything including bots, dev, direct, social, paid, email, etc. (overcounts without filtering).

For reliable reporting, standardize on: GSC for organic-search performance, GA4 (filtered) for total real-user traffic and behavior. Ahrefs is a secondary directional signal only.

GA4 Sources / Channels

source: research/ga4/sources.md

GA4 Traffic Sources & Channels

Property: [redacted GA4 property] Window: 2026-02-20 → 2026-05-19 (89 days) Pulled: 2026-05-21

Two caveats before reading any number

1. Bot and internal traffic is NOT filtered at the GA4 source. "Direct" in particular is heavily inflated by bot sessions from Singapore, China, Pakistan, and AWS subnets. Real-human direct traffic is a small fraction of what the chart shows.

2. Conversion counts are a floor. Form submissions are not tracked (generate_lead has never fired since it was created in March 2025), so any "key events" column understates real conversions on every channel.

Channel Distribution (Raw, Unfiltered)

Key Findings

1. "Direct" dominates the chart (74%) — most is bot

1,583 of 2,146 sessions are tagged "Direct / (direct) / (none)" — almost certainly the same Singapore / China / Pakistan / data-center traffic dominating the geographic report. Real direct traffic (people typing client-example.invalid or clicking unattributed bookmarks) is probably 5–10% of this number.

2. Organic Search Google is the clearest signal — and it converts

204 sessions from google / organic over 89 days = ~68 sessions/month from real Google search. Of these:

• 86 unique users
• 134 engaged sessions (66% engagement) — much higher quality than bot direct traffic
• 12 key events (phone/email clicks) — a meaningful conversion signal

This is the channel that already converts when not held back by tracking gaps.

3. GMB / Google Business Profile is a distinct traffic stream

123 sessions tagged gmb / organic (the homepage URL with ?utm_source=gmb&utm_medium=organic&utm_campaign=website) — these are clicks from the Google Business Profile listing's website link.

• 106 users
• 65 engaged sessions (53% engagement)
• 1 key event

GMB is a separate channel from organic Google search and should be reported separately, not lumped into "organic". It has its own management surface (the GBP listing itself) and its own optimization levers.

Combined, Google search + GMB = 327 sessions / 89 days ≈ 110 sessions/month of real Google-driven traffic — addressable and healthy.

4. AI-engine referrals are present — early-mover signal

Three AI tools are sending small but real traffic:

• chatgpt.com: 3 + 3 + 1 = 7 sessions
• claude.ai: 2 sessions
• copilot.com: 1 session

This is small but meaningful: it's first-party evidence that client-example.invalid is being cited / linked by LLMs in user conversations. Tracking AI referrals as a discrete channel going forward will let you measure whether this stream grows.

5. Social referrals are modest

Houzz visitors are extremely engaged (89% engagement rate vs ~24% site-wide). These are pre-qualified buyers who found Client Company through the Best of Houzz badge.

6. Referrals reveal internal tooling

• app.clickup.com: 13 sessions, 3 users — internal team referencing the site from their PM tool
• project-workspace.example: 7 sessions — another PM/CRM tool
• qr1.be: 28 sessions — a QR code shortener (possibly print marketing or business cards)
• docs.google.com: 1 session — internal docs

The qr1.be source is worth investigating: where is the QR code printed, and what page does it land on? 28 sessions/quarter is non-trivial.

7. Bing exists but is small

20 sessions from Bing in 89 days = ~7 sessions/month. Worth a Bing Webmaster Tools setup but not a primary channel.

Channel Strategy Implications

Highest-leverage channels (real traffic, real engagement)

1. Google Search organic — 204 sessions, 12 key events
2. GMB / Google Business Profile — 123 sessions, 1 key event (track separately)
3. Houzz referrals — 9 sessions, 89% engagement (premium pipeline)

Maintain-only channels

4. Instagram, Facebook — present but low engagement
5. Yelp — small but qualified
6. Bing — minimal volume

Growth channels (untapped or new)

7. AI engines (ChatGPT, Claude, Copilot) — early, growing, worth dedicated tracking
8. Pinterest — not visible in current data despite Client Company having a Pinterest account; unused channel inventory worth investigating

Filter / fix channels

9. "Direct" — must be filtered to extract real direct traffic from bot noise
10. qr1.be and PM-tool referrals — internal; should be excluded from public reports

Configuration Recommendations

1. Set up Internal Traffic filter in GA4 — exclude Philippines, ClickUp, Worksection, internal IPs
2. Set up Bot Filter — exclude Singapore, China data centers, Germany, Pakistan, AWS subnets
3. Tag GMB traffic as its own channel — already utm-tagged; route it into a custom channel group instead of "organic"
4. Add UTM tagging on social channels — Instagram, Facebook, Yelp should have utm tags so social can be properly attributed
5. Track AI engines as a custom channel group — manually classify chatgpt.com, claude.ai, copilot.com, perplexity.ai, etc.
6. Investigate the qr1.be source — find where the QR code is deployed

GA4 Conversions

source: research/ga4/conversions.md

GA4 Conversion Tracking & Events

Property: [redacted GA4 property] Window: 2026-02-20 → 2026-05-19 (89 days) Pulled: 2026-05-21

Headline: conversion tracking is broken

Two facts up front:

1. generate_lead has been configured in GA4 since March 5, 2025 — and has never fired. Not once. The contact form on /contact/ is not pushing this event.
2. 70%+ of GA4 sessions are bots or internal-team traffic, so even the events that do fire are partially polluted by staff testing forms.

Every conversion number in this file is a floor. The true conversion volume is unknown until tracking is fixed.

Configured Key Events (Conversions)

Findings

1. purchase event configured but never fires — likely a template remnant. Service businesses don't need a purchase event. Should be removed or repurposed.

2. generate_lead event configured but has NEVER fired since March 2025. This is the most critical missing conversion. A remodeling firm whose primary CTA is "Book Your Consultation" should be tracking form submissions as generate_lead. The event exists in GA4 but is not wired to the contact form. Likely causes:

   • Missing GTM trigger on form submit
   • Form is rendered via a plugin (Gravity Forms / Contact Form 7) that needs a DataLayer push
   • The form-submit wiring was never completed, or it broke during the April 2026 rebuild

3. s_click_to_call event configured but never fires — the "s_" prefix suggests it was a custom event from a previous tracker that's no longer being pushed. Either it duplicates phone_click (which IS firing 13 times) or it's broken. Recommend deleting.

4. phone_click and email_click ARE firing — these are the only working conversion events. Together they represent 17 total conversion signals in 89 days = ~6/month.

All Event Counts (Last 89 Days)

Observations

• scroll, click, video_progress events are firing — GA4 Enhanced Measurement is on and working
• click event fires only 40 times across 89 days for 24 users — auto-tracked outbound clicks. Could be much richer with proper GTM event tracking
• Video engagement IS happening — 148 video_progress events, 14 video_complete events — but only 4 to 16 unique users. Worth investigating which page hosts the video(s)
• No form_submit event in the list — confirming the form-tracking gap. The consultation form mechanism on the site isn't producing any analytics signal

Conversion Rate (Real, with caveats)

Using filtered estimates and acknowledging the tracking floor:

A 6.8% click-to-call/email rate would be decent for a luxury home-services site, but the absolute volume is so low (17 events / 3 months) that one broken event handler — like form_submit — could easily be doubling the actual conversion volume without anyone seeing it.

Critical Measurement Gaps

Gap 1 — Form submit tracking

No form_submit, contact_form_submit, or fired generate_lead events exist.

The site has at least one contact form (visible at /contact/ and likely embedded across service pages). The 16-month picture is even sharper: /contact/ received 1,082 sessions and 224 phone/email clicks — but zero form submissions captured as events. Form submits must be tracked because:

• Phone clicks (13/quarter) and email clicks (4/quarter) are only the simplest conversion paths
• Most luxury remodeling leads come through a structured consultation request form
• The /contact/ page shows ~45% bounce rate (vs 75% site avg) — clear intent — but no event fires when they convert

Fix: Implement a DataLayer push on form submit via GTM, fire as generate_lead.

Gap 2 — Phone click coverage

13 phone clicks in 89 days = ~4–5/month. Low for a site with the actual traffic levels. Possible causes:

• Phone number CTA only fires on mobile (and mobile traffic is artificially low due to bots inflating desktop)
• Click handler missing on header phone number, sticky bottom CTA, or footer phone
• Multiple phone number elements without consistent event firing

Fix: Audit every clickable phone element on the site to ensure phone_click fires from each.

Gap 3 — Email click coverage

4 email clicks in 89 days = ~1/month. Very low even accounting for the channel.

Fix: Same audit on every mailto: link.

Gap 4 — Outbound/social click tracking

The click event fires 40 times/89 days across 24 users. Enhanced Measurement covers outbound clicks but the volume suggests outbound-link interactions (Yelp, Houzz, Facebook, Instagram from site footer) are happening but not being captured granularly.

Fix: Set up GTM listeners on:

• Yelp link clicks
• Houzz link clicks
• Facebook / Instagram / YouTube / Pinterest social link clicks
• "Book Consultation" button clicks (cross-page)
• Portfolio item clicks
• Service-area card clicks

Gap 5 — Scroll depth & engagement granularity

scroll event fires 1,113 times for 404 users — meaning the average user who scrolls fires the event ~2.8 times. Default scroll tracking only fires at 90%. Custom scroll-depth tracking (25%, 50%, 75%, 100%) would reveal where users drop off on long service/portfolio pages.

Gap 6 — Bot traffic distorting every engagement metric

70%+ of sessions are bots and they inflate page_view, session_start, and first_visit counts. Without proper bot filtering, every engagement and conversion metric reported in GA4 dashboards is materially wrong.

Recommended Event Setup (Priority Order)

What This Means

The site is operating with broken conversion measurement. Any decisions about marketing investment are being made against data that is:

1. Inflated by bot traffic (70%+ of raw sessions)
2. Missing form-submit events entirely (generate_lead has never fired)
3. Tracking deprecated events that never fire (purchase, s_click_to_call)
4. Not capturing the full conversion journey (no outbound, CTA, or scroll-depth granularity)

Fixing measurement is a prerequisite to evaluating performance. Until conversion tracking is reliable, it's not possible to:

• Calculate true conversion rate
• Measure the impact of CRO changes
• Justify or evaluate marketing spend with ROI math
• Compare channels on lead quality (only on session volume)

This is the highest-leverage first move for the analytics stack.

GA4 Landing Pages

source: research/ga4/landing-pages.md

GA4 Landing Pages

Property: [redacted GA4 property] Window: 2026-02-20 → 2026-05-19 (89 days) Total landing pages with sessions: 286 Pulled: 2026-05-21

Two caveats before reading

1. Bot and internal traffic is NOT filtered at the GA4 source (70%+ of sessions). Many of the "sessions" columns below include bot landings — particularly on / and on (not set).

2. Conversion counts are a floor. generate_lead has never fired since it was configured in March 2025. The only working conversion events are phone_click and email_click. Form submissions on /contact/ are completely invisible to analytics, so a page showing 0 key events may have produced real form submits that were never tracked.

Top Landing Pages

Findings

1. Homepage = 44% of all sessions

952 of 2,146 sessions land on / — the homepage is the only meaningful entry point. After bot filtering, this share is likely to drop, but it confirms that deep-page traffic is essentially zero.

2. /contact has the best engagement on the site — but its conversions are invisible

/contact shows:

• 45.5% bounce rate (vs site avg 75%)
• 54% engagement rate
• 18 of 33 sessions engaged

People who land on /contact directly already intend to convert. Yet 0 key events fired from contact-page landings in this 89-day window. That's a measurement gap, not a content failure: form submissions on /contact aren't tracked at all (generate_lead has never fired), so real conversions on this page are happening invisibly.

In the broader 16-month view, /contact/ accumulated 1,082 sessions and 224 conversion events — but those 224 are all phone and email clicks, not form submissions. Form-submit tracking would likely double or triple the captured conversion count on this page alone.

3. Legacy URLs are still landing pages

GA4 confirms the dual URL structure problem:

Each pair splits link equity, indexation budget, and traffic. A 301 redirect pass would consolidate these.

4. New /services/[slug]/ pages get traffic but convert at 0%

The newer service detail pages built during the rebuild get real session volume but produce zero or near-zero conversions:

• /services/bathroom-remodeling: 35 sessions, 3 key events (89-day window) — best of the new set
• /services/whole-home-renovations: 33 sessions, 1 key event
• /services/kitchen-remodeling: 21 sessions, 0 key events
• /services/home-additions: 19 sessions, 0 key events

The 16-month picture is starker:

• /services/bathroom-remodeling/: 139 sessions, 0 conversions
• /services/whole-home-renovations/: 91 sessions, 0 conversions
• /services/kitchen-remodeling/: 96 sessions, 0 conversions
• /services/home-additions/: 71 sessions, 0 conversions

The conversion CTAs are either missing on these pages, or the events aren't wired to fire. This is a lead-capture gap, not a traffic gap.

5. /service-areas has 100% bounce, 0.2s duration

The service-areas index page is broken or empty — 30 sessions land there and immediately leave with 0 engaged sessions. Likely an issue with the page template or content. High-priority audit item.

6. (not set) is 45 sessions of bot traffic

The 45 sessions with (not set) landing page (97.8% bounce, 1s duration) are almost certainly bots that hit the site without a clean landing page record.

7. Long tail of low-engagement service / blog landings

Many service and blog pages received 7–12 sessions each with mostly 0–1 engaged sessions:

Most of these get traffic but fail to retain engagement. Likely cause: thin content, generic templates, no CTAs, no localized photos.

8. Portfolio pages convert when found

• /portfolio-gallery/encinitas-addition-bath-flooring-remodel-project: 7 sessions, 0 engaged
• /portfolio-gallery/escondido-bathroom-remodel-project: 7 sessions, 5 engaged (71% engagement)
• /portfolio-gallery/scripps-ranch-master-bath-remodel-project: 7 sessions, 4 engaged (57% engagement)

Portfolio pages with real project content outperform generic service pages on engagement. Investment in portfolio detail yields a higher return than service-overview copy.

Landing Page Strategy Implications

Immediate actions

1. Audit /service-areas index page — diagnose and fix the 100% bounce
2. Wire event tracking on /contact — fire phone_click, email_click, and generate_lead (form submit) consistently
3. Build 301 redirects for legacy duplicates — /about → /about-us, /san-diego-ca → /service-areas/[appropriate], etc.
4. Add and track CTAs on /services/[slug]/ pages — the new service detail pages convert at 0% in the 16-month window despite real traffic

Medium-term actions

5. Localize service-area pages — they get impressions but no engagement; add real photos, real projects, real testimonials, and neighborhood-specific content
6. Expand portfolio depth — portfolio pages engage 50–70% of visitors when reached; build out more case studies
7. Internal linking from blog → service → portfolio — currently blog visitors don't move through the funnel

Measurement gaps to close

8. Most conversion events are not firing — purchase and generate_lead have 0 fires. Only phone_click (13) and email_click (4) work. Form-submit tracking is broken or missing entirely
9. Map the homepage funnel — where do real users go after landing on /?

WordPress — Full Report

source: reports/full/04-wordpress.md

WordPress Stack Audit — client-example.invalid

Compiled: 2026-05-21 Source: WordPress REST API (client-example.invalid/wp-json/), page source inspection, sitemap analysis

1. Stack & Technology

Detected via REST API namespace probe (/wp-json/)

What stands out

1. Redirection plugin is installed but unused. Zero active 301s are configured. Meanwhile ~25 legacy URLs (/poway/, /encinitas/, /about-us/, /bathroom-remodeling/, etc.) from the pre–April 2026 site remain live and indexed alongside their replacements. The tool to fix this is already in place — it just hasn't been used.

2. LiteSpeed Cache is active but mobile PageSpeed is 53. The plugin is installed but not properly configured for mobile — combine/minify, defer non-critical JS, and critical CSS extraction are not in effect on the mobile render path.

3. Elementor + Elementor Pro + Elementor AI + Brainstorm Force. The full paid page-builder stack is active. This stack is the source of the ~195KB of unused JavaScript flagged by PageSpeed on mobile.

4. NPS Survey plugin is installed but not visible on the live site. Plugin bloat that can be removed.

5. Five sitemaps registered in Google Search Console. Yoast manages the canonical /sitemap_index.xml. The other four entries are residual from earlier configurations and should be removed in GSC.

6. No security plugin is visible. Wordfence, Sucuri, iThemes Security — none detected. For a site with admin accounts and form submissions, this is a gap.

Components inferred but not directly visible

These would require admin login to confirm:

• Form plugin: most likely Elementor Forms (Elementor Pro is active)
• Backup plugin: unknown
• Image optimization: LiteSpeed handles WebP conversion (matches the March 2026 mass WebP upload pattern in the media library)
• CDN: Cloudflare proxy (confirmed)

2. WordPress User Accounts

Three user accounts exist on the site (visible via REST API):

Notes

• The admin slug client-admin is a guessable username. Combined with Application Passwords enabled, this is a known brute-force surface. Standard hardening: rename the admin slug, enable 2FA, audit Application Passwords for unused entries.
• Both Prior Vendor accounts are still active in WordPress. They retain full access until they are explicitly disabled.

3. Content Inventory

From the WordPress REST API:

Page breakdown

Posts (blog)

• All 39 posts authored as user 6 (client-admin)
• 33 created in a one-day batch on Oct 10–11, 2024 (template/import)
• 6 created during the audit window (May 2025 × 2, April 2026 × 4 — 2 of which duplicate page slugs)
• All 33 original posts show last-modified Apr 20, 2026 in a 19-second bulk save (database operation, not editorial work)

Media library

• 1,274 total items
• Author breakdown: ~80% user 10 (Prior Vendor), ~15% user 4 (Content Contributor), ~5% user 6 (admin)
• Format mix: JPEG (60%), WebP (30%), PNG (8%), SVG/MP4/other (2%)
• Roughly 30% of media are WebP duplicates of existing JPEGs — output of LiteSpeed's WebP generation. Significant orphan media is present and should be audited.

4. Plugin Summary

5. Schema / Structured Data

GSC search appearance report confirms:

• Product snippets — 73 impressions in 16 months, 0 clicks
• Translated result — 2 impressions

This indicates Yoast has Product/Service schema active on at least one page type, but the snippets are not earning clicks (likely wrong intent match or weak titles/descriptions).

Schema markup that should exist for this site

6. Sitemap Configuration

Five sitemap entries are registered in Google Search Console, reflecting layered plugin history:

The four residual entries should be unsubmitted in GSC so only /sitemap_index.xml remains.

7. Robots.txt

User-agent: *
Disallow: /wp-admin/
Allow: /wp-admin/admin-ajax.php
Disallow: */init$
Disallow: */params$
Disallow: */migrations$
Disallow: */data$
Disallow: */helpers/*
Disallow: */terms$
Sitemap: https://client-example.invalid/sitemap_index.xml

Standard configuration. The wildcard disallows (*/init, */params, etc.) are unusual and suggest a plugin or app once had endpoint paths needing to be blocked. Worth a review to confirm they're still necessary.

8. URL Structure & Permalink Hygiene

Current state

• Permalinks use /post-name/ format — good for SEO
• The April 2026 rebuild created new /services/... and /service-areas/... paths but left ~25 legacy URLs uncoupled. Both versions are live and indexed for the same content:
  • /about/ (legacy) and /about-us/ (current)
  • /poway/ (legacy) and /service-areas/poway/ (current)
  • /kitchen-remodeling/ (legacy) and /services/kitchen-remodeling/ (current)
  • /privacy_policy/ (legacy, underscore) and /privacy-policy/ (current, hyphen)

Required action

Build a 301 redirect map for all ~25 legacy URLs using the Redirection plugin (already installed). This consolidates Google's ranking signal — right now it's being split between duplicate URLs.

9. Performance Impact

Combining the WordPress detection with PageSpeed results (pulled 2026-05-21):