Bonsai Marketing / Foundation AssessmentXpress built 153 pages to win San Diego. They earn nothing.
Xpress Restoration has done the hard part. The company is licensed, verified, reviewed 192 times at 4.9 stars, and it ranks first in Chula Vista for the terms that matter most in its home city. The technical build is cleaner than most of this market — every crawled URL returns 200, schema is present on every page, and all 8,126 images carry ALT text.
The problem is not effort, and it is not quality. It is architecture. Over 153 service-by-location pages have been built across seven service lines and 21 city slugs. Search Atlas attributes 282 keyword rankings to those pages — and 0 monthly visits. Every one of the 174 organic visits the site earns each month lands on the homepage.
The headline opportunity. Xpress ranks where the demand is not, and is absent where the demand is. Bonita and Coronado — 11 and 30 searches a month between them — return top-20 rankings on 9 and 8 of 10 queries. “San Diego,” worth 2,704 searches a month, returns zero.
Three moves drive most of the upside: (1) collapse the duplicate URL patterns so one page owns each intent, (2) build the four San Diego head terms that carry 2,621 searches a month at $52–$62 a click, and (3) open the mold line, which has 742 searches a month and a single page behind it.
The seven questions, answered
- Are the location pages helping or hurting? Neither, yet — they are inert. 282 rankings, 0 visits.
- Is this a Google penalty? No. There is no evidence of a manual action or algorithmic suppression: pages are indexed, canonicalized and ranking. The pages compete with each other rather than with the market.
- Are the pages duplicate content? Not in the strict sense. Median main-content similarity between two city pages in the same service is 47% — and 62% once city names are removed. They are templated, not copied.
- So what is wrong with them? Information gain. The median location page carries zero of ten genuine local signals, and 32% of the page is unique content.
- Why does Google keep switching URLs? 76 of 182 ranking keywords return more than one Xpress URL. With no hierarchy and no differentiation, Google has no basis to choose.
- Who is actually winning? ServPro takes 82 of the 100; Yelp takes 81. Xpress takes 16.
- Is Xpress visible in AI search? No — 0 mentions across 24 answers, including the Chula Vista question it ranks first for.
[ Data verified ] Full-site crawl of 249 URLs, August 18, 2026 · Search Atlas Site Explorer (site 840971) · Search Atlas rank tracker, live San Diego SERPs, August 18, 2026 · Search Atlas LLM Visibility, 5 platforms · Search Atlas GBP audit 17315.
Two URL systems, one intent, twenty-three collisions.
The site runs two parallel URL patterns for the same thing. One is a folder structure — /emergency-water-removal/coronado/. The other is a flat slug — /emergency-water-removal-coronado-ca/. Both are live, both return 200, both are indexable, and both target the identical search.
Twenty-three service-and-city combinations have two competing pages. These are not near-duplicates that Google will quietly ignore — they are two fully-built, internally-linked, self-canonicalizing pages asking to rank for the same query.
| Service | City | Competing URLs |
|---|---|---|
| Emergency Water Removal | bonita | /emergency-water-removal-bonita-ca//emergency-water-removal/bonita/ |
| Emergency Water Removal | chula-vista | /emergency-water-removal-chula-vista-ca//emergency-water-removal/chula-vista/ |
| Emergency Water Removal | coronado | /emergency-water-removal-coronado-ca//emergency-water-removal/coronado/ |
| Emergency Water Removal | imperial-beach | /emergency-water-removal-imperial-beach-ca//emergency-water-removal/imperial-beach/ |
| Emergency Water Removal | mission-hills | /emergency-water-removal-mission-hills-ca//emergency-water-removal/mission-hills/ |
| Emergency Water Removal | national-city | /emergency-water-removal-national-city-ca//emergency-water-removal/national-city/ |
| Emergency Water Removal | san-diego | /emergency-water-removal-san-diego-ca//emergency-water-removal/san-diego/ |
| Flood Cleanup | bonita | /flood-cleanup-bonita-ca//flood-cleanup/bonita/ |
| Flood Cleanup | coronado | /flood-cleanup-coronado-ca//flood-cleanup/coronado/ |
| Flood Cleanup | imperial-beach | /flood-cleanup-imperial-beach-ca//flood-cleanup/imperial-beach/ |
| Flood Cleanup | mission-hills | /flood-cleanup-mission-hills-ca//flood-cleanup/mission-hills/ |
| Flood Cleanup | national-city | /flood-cleanup-national-city-ca//flood-cleanup/national-city/ |
[ Data verified ] Bonsai full-site crawl, August 18, 2026: 249 URLs fetched, 236 indexable, 233 in sitemap.xml.
The seven-way San Diego overlap
Beyond the city grid, seven separate URLs target San Diego water-damage intent:
/san-diego/— 4,660 words/water-damage/— 4,873 words/water-damage-restoration/— 2,154 words/water-damage-restoration/san-diego/— 2,272 words/emergency-water-removal-san-diego-ca/— 2,787 words/flood-cleanup-san-diego-ca/— 2,827 words/24-7-emergency-water-restoration-san-diego/— 1,330 words, and no other page on the site links to it
Roughly 20,000 words are pointed at one commercial intent across seven URLs. None of them ranks in the top 20 for water damage restoration san diego — the single largest term on this benchmark at 1,289 searches a month.
Ten services. Ten markets. One hundred searches.
We benchmarked ten restoration services against ten San Diego County markets — an exact 100-query grid — and measured Xpress’s position on live, San Diego-geotargeted SERPs. Where the live tracker returned nothing, we fell back to the Search Atlas organic index. Cells marked not verified are queries where neither source places Xpress in range; we do not manufacture a position.
| Service ↓ / Market → | Chula Vista102/mo | San Diego2,704/mo | Downtown San Diego0/mo | National City9/mo | Imperial Beach9/mo | Coronado30/mo | Bonita11/mo | La Mesa23/mo | El Cajon49/mo | Eastlake6/mo |
|---|---|---|---|---|---|---|---|---|---|---|
| water damage restoration | 1index | — | — | 27index | 26index | 15live | 17live | 28index | — | — |
| emergency water removal | 10live | — | — | 18live | — | 5live | 1live | 17live | 19live | 12live |
| water damage repair | 1index | — | — | — | — | 13live | 17live | — | 35index | 16live |
| flood cleanup | 16live | — | 8live | — | 13live | — | 12live | — | — | 5live |
| water extraction | 9live | — | — | — | — | 11live | 6live | 16live | 17live | 8live |
| sewage cleanup | — | — | — | 13live | — | 12live | 8live | 19live | — | 6live |
| fire damage restoration | 1index | — | 14live | 18live | 18live | 12live | 8live | — | — | 16live |
| smoke damage cleanup | — | — | 18live | 17live | 12live | 9live | 4live | 19live | — | — |
| mold remediation | — | — | — | — | — | — | — | — | — | — |
| emergency restoration company | — | — | — | — | — | 16live | 11live | — | — | — |
[ Data verified ] Search Atlas rank tracker project 84994, desktop, geotargeted to San Diego, California, pulled August 18, 2026. Live tracking resolves the top 20; the Search Atlas organic index supplies positions to 100. Volumes and CPC from Search Atlas keyword lookups, US, August 18, 2026.
The column that matters
Read the grid by column and the picture resolves immediately. Xpress performs in the small southern suburbs and disappears in the city.
| Market | Monthly searches | Top 20 | Top 10 |
|---|---|---|---|
| Chula Vista | 102 | 3/10 | 5/10 |
| San Diego | 2,704 | 0/10 | 0/10 |
| Downtown San Diego | 0 | 3/10 | 1/10 |
| National City | 9 | 4/10 | 0/10 |
| Imperial Beach | 9 | 3/10 | 0/10 |
| Coronado | 30 | 8/10 | 2/10 |
| Bonita | 11 | 9/10 | 5/10 |
| La Mesa | 23 | 4/10 | 0/10 |
| El Cajon | 49 | 2/10 | 0/10 |
| Eastlake | 6 | 6/10 | 3/10 |
2,704 of the 2,943 searches on this grid — 92% — are “San Diego” queries. Xpress ranks in the top 20 for 0 of them. Bonita, by contrast, carries 11 searches a month and returns 9 top-20 rankings out of 10. The effort has gone almost exactly where the demand is not.
The service line nobody built
| Service | Monthly searches | Top 20 |
|---|---|---|
| Water Damage Restoration | 1,421 | 2/10 |
| Emergency Water Removal | 11 | 7/10 |
| Water Damage Repair | 484 | 3/10 |
| Flood Cleanup | 30 | 5/10 |
| Water Extraction | 14 | 6/10 |
| Sewage Cleanup | 30 | 5/10 |
| Fire Damage Restoration | 211 | 6/10 |
| Smoke Damage Cleanup | 0 | 6/10 |
| Mold Remediation | 742 | 0/10 |
| Emergency Restoration Company | 0 | 2/10 |
Mold remediation is the second-largest demand pool on the entire grid at 742 searches a month. Xpress names mold in its own schema description and its own service menu — and has exactly one page for it, no city coverage, and 0 of 10 rankings.
The four terms worth building for
| Query | Searches/mo | CPC | Difficulty | Xpress position |
|---|---|---|---|---|
| water damage restoration san diego | 1,289 | $52.51 | 21 | Not ranking |
| mold remediation san diego | 665 | $62.00 | 25 | Not ranking |
| water damage repair san diego | 469 | $61.88 | 9 | Not ranking |
| fire damage restoration san diego | 198 | $0.00 | 3 | Not ranking |
| water damage restoration chula vista | 44 | $0.00 | 6 | 1 |
| mold remediation chula vista | 43 | $0.00 | 9 | Not ranking |
| sewage cleanup san diego | 30 | $0.00 | 13 | Not ranking |
| water damage restoration el cajon | 28 | $0.00 | 5 | Not ranking |
| flood cleanup san diego | 28 | $0.00 | 8 | Not ranking |
| water damage restoration coronado | 25 | $0.00 | 5 | 15 |
The top four San Diego head terms carry 2,621 searches a month at $52–$62 a click, with keyword difficulty between 3 and 25 — Easy to Medium. This is not a market that requires outspending ServPro. It requires one credible page per intent.
When one query returns fourteen of your own URLs.
The clearest evidence that the architecture is working against itself is not a similarity score — it is Google’s own behavior. Across the Search Atlas organic index, 76 of 182 ranking keywords (41.8%) return more than one Xpress URL. Each block below is one competing URL; the number is its position.
[ Data verified ] Search Atlas Site Explorer organic index, site 840971 — 358 keyword-to-URL rows across 182 distinct keywords, pulled August 18, 2026.
Read the top row. For flood damage restoration imperial beach, Google has indexed and ranked 13 different Xpress URLs — the flood page, the water-damage page, the sewage page, the fire page, and pages for entirely different cities. That is not a ranking. That is Google guessing.
The pages are answering the wrong questions
The confusion runs across service lines, not just cities:
- A sewage cleanup page ranks for water damage restoration la mesa.
- A fire damage restoration page ranks for flood damage repair coronado.
- The Chula Vista page is what Google serves for fire damage restoration Eastlake.
- The San Diego sewage page is what Google serves for sewage cleanup Eastlake.
On the live tracker, three of the Eastlake queries resolve to a page for a different city entirely. When a fire page can outrank a flood page for a flood query on your own domain, the pages are no longer distinguishable to a search engine.
What this is, precisely. There is no evidence of a penalty here — no manual action signature, no deindexing, no suppression. What the data shows is textbook intent cannibalization: split relevance signals, split internal authority, split behavioral data, and a crawler spending its budget deciding between your own pages instead of ranking one of them.
Not duplicated. Undifferentiated.
It would be easy — and wrong — to call these pages duplicate content. We measured it properly. After stripping the sitewide navigation and footer that appears on every page, we compared every pair of city pages inside each service cluster, then repeated the comparison with all city names replaced by a neutral token.
| Main-content similarity band (city names removed) | Page pairs | Share |
|---|---|---|
| 90–100% — effectively identical | 1 | 0.1% |
| 80–89% | 11 | 0.6% |
| 70–79% | 28 | 1.6% |
| 60–69% | 935 | 52.6% |
| Below 60% | 803 | 45.2% |
[ Data verified ] 1,778 intra-service page pairs, 6-gram Jaccard similarity on main content after removing 562 sitewide boilerplate shingles. Bonsai analysis of the August 18, 2026 crawl.
The honest verdict: these are not copy-paste pages. Someone wrote genuinely different opening copy for Coronado than for La Mesa. Only 12 of 1,778 pairs exceed 80% similarity. The writing is competent. That is exactly why the real problem is easy to miss.
The real problem: information gain
A location page earns its place by containing something a searcher cannot get from the service page. We tested all 153 location pages for ten categories of genuine local information — measured only against the unique portion of each page, not the shared template.
| Local signal | Pages containing it | Coverage |
|---|---|---|
| Named Neighborhood | 33 / 153 | 22% |
| Property Type | 46 / 153 | 30% |
| Local Risk | 30 / 153 | 20% |
| Zip Code | 28 / 153 | 18% |
| Named Landmark | 11 / 153 | 7% |
| Case Study | 0 / 153 | 0% |
| Response Logistics | 0 / 153 | 0% |
| Named Person | 0 / 153 | 0% |
| Original Stat | 0 / 153 | 0% |
| Unique Review | 0 / 153 | 0% |
The median location page scores zero out of ten. 79 of 153 pages contain not one of these signals.
Not a single page — zero of 153 — contains a case study, a named technician, a documented response time, an original statistic, or a review written by a customer in that city. The pages describe the service. They do not evidence the business in that place.
One template variable, repeated 62 times
The line “We’ve helped 3,100+ Property Owners In [City]” appears on 62 pages, against 22 different cities. Read literally, that claims more than 68,000 customers. The same template carries casing errors — “Chula vista” and “encinitas” both appear in the live copy.
This is the single most damaging line on the site. It is the one place where a careful homeowner — or a quality rater — can see that the page was generated rather than written. Replace it with the real number for the real service area, once.
The winners run one page per city.
We measured every domain appearing in the top 10 across the same 100 SERPs — 203 of them — and ranked by how many of the 100 queries each holds a top-10 position for. This is the real competitive set, not a list of famous names.
| Domain | Type | Keywords in top 10 | Share of the 100 |
|---|---|---|---|
| servpro.com | Operator | 82 / 100 | 82% |
| yelp.com | Directory | 81 / 100 | 81% |
| amresponse.com | Operator | 39 / 100 | 39% |
| whiteglove-restoration.com | Operator | 36 / 100 | 36% |
| facebook.com | Directory | 28 / 100 | 28% |
| pacificflood.com | Operator | 24 / 100 | 24% |
| bbb.org | Directory | 23 / 100 | 23% |
| agnwater.com | Operator | 21 / 100 | 21% |
| quickdry.com | Operator | 20 / 100 | 20% |
| certifiedrestorationinc.com | Operator | 19 / 100 | 19% |
| sandiegowaterrestorations.com | Operator | 17 / 100 | 17% |
| kicrestoration.com | Operator | 14 / 100 | 14% |
| xpress-restoration.com | Operator | 16 / 100 | 16% |
[ Data verified ] Search Atlas rank-tracker competitor visibility, project 84994, 203 domains observed across the 100 geotargeted SERPs, August 18, 2026.
Two things stand out. ServPro holds 82 of the 100 — this market has a national incumbent that actually shows up. And Yelp holds 81: four in five of these SERPs are directory-mediated, which means a review profile is a ranking asset, not a vanity metric.
Pages are not the constraint
The instinct behind 153 location pages is sound — local pages do win in this market. They simply have to be built differently. Compare Xpress with three genuine San Diego competitors:
| Domain | Indexed pages | Ranking keywords | Est. monthly organic visits | Visits per page |
|---|---|---|---|---|
| first2respond.com | 77 | 793 | 1,272 | 16.5 |
| quickdry.com | 55 | 724 | 332 | 6.0 |
| goldcoastflood.com | 125 | 796 | 229 | 1.8 |
| xpress-restoration.com | 100 | 358 | 174 | 1.7 |
[ Data verified ] Search Atlas Site Explorer, one project per domain, August 18, 2026. Visits are Search Atlas organic traffic estimates for the top 100 pages of each domain.
quickdry.com earns 332 monthly visits from 55 pages. Xpress earns 174 from 100. Their Escondido page alone earns 102 visits a month — more than the entire Xpress site.
The difference is the template. Quick Dry runs one combined page per city: /water-flood-damage-escondido/. Gold Coast runs /water-damage-restoration-in-<city>/. First 2 Respond runs /emergency-restoration-company-<city>/. Each competitor gives a city one page that answers every water question. Xpress gives a city up to six service pages, and for 23 combinations, two copies of one of them.
The build is sound. The directives are not.
This is the strongest part of the account and it deserves to be said plainly: the technical foundation is better than most sites in this market. Every one of the 249 URLs we crawled returned HTTP 200. There are no broken canonicals, no noindex leaks, and no missing schema.
The one defect worth fixing this week
99 of 236 pages ship this tag:
The directive name has been substituted into its own value. The correct form is max-snippet:-1, which tells Google to use any snippet length it likes. What is published instead is invalid — and where a parser recovers a value at all, it reads as a one-character snippet limit. Worse, these pages carry two conflicting robots meta tags, one emitted by WordPress and one by the SEO plugin.
This is a single template fix affecting 42% of the site, and it governs how every one of those pages is allowed to appear in the results.
The rest of the on-page picture
| Check | Result | Assessment |
|---|---|---|
| Title length | 215 of 236 over 60 characters | Truncating in the SERP |
| Meta description length | 127 over 160 characters | Truncating |
| Duplicate H1s | 41 pages share 10 H1 strings | Weakens topical signal |
| Multiple H1s on one page | 10 pages | Fix in template |
| Missing H1 | 3 pages | Minor |
| Duplicate meta descriptions | 16 pages | Minor |
| Schema coverage | 236 of 236 pages | Excellent |
| FAQPage schema | 34 of 236 pages | Clear opportunity |
| Image ALT coverage | 100% of 8,126 | Excellent |
| Orphan pages | 14 with zero inbound internal links | Includes a San Diego money page |
| Broken internal link | /sewage-cleanupoceanside/ — 301s to the homepage | Linked from 2 live pages |
| Case-variant URL | /flood-cleanup/Escondido/ resolves 200 | Canonicalized correctly |
Internal linking: everything points everywhere
The site contains 38,230 internal links across 236 pages — an average of 161 outbound links per page. 166 URLs are linked from more than 90% of the site, because the mega-menu lists every service against every city on every page.
This is the mechanism behind the cannibalization. When every page links to 161 others, no page is ever signalled as more important than any other. Combine flat linking with pages that only differ by city name, and Google is given no way to choose a winner — so it keeps changing its mind. The 41.8% cannibalization rate is the visible symptom; the flat link graph is the cause.
Zero mentions in twenty-four answers.
We put eight buying-intent questions to ChatGPT, Gemini and Perplexity — the questions a homeowner actually types when water is coming through a ceiling — and measured whether Xpress Restoration appears in the answer.
[ Data verified ] Search Atlas LLM Visibility prompt simulator, project 2996691f, run August 18, 2026 against ChatGPT, Gemini and Perplexity.
The sharpest finding in this assessment. One of the eight questions was “who should I call for emergency water removal in Chula Vista CA” — Xpress’s home city, where it ranks first in classic organic search. It was not named by any of the three engines. Meanwhile “SERVPRO of Chula Vista” was recommended by name.
What the engines cite instead
| Domain cited in AI answers | Times cited across 24 answers |
|---|---|
| servicemasterrestore.com | 12 |
| servpro.com | 11 |
| pauldavis.com | 11 |
| cslb.ca.gov | 7 |
| puroclean.com | 6 |
| restoration1.com | 6 |
| servprochulavista.com | 3 |
| floodpros.com | 3 |
| servicemasterbybest.com | 3 |
| sandiego.pauldavis.com | 3 |
The pattern is consistent: answer engines reach for national franchise networks with clean entity data, and for license and certification bodies (cslb.ca.gov, iicrc.org) to qualify them. Xpress holds a valid CSLB license and IICRC certification and states both on the site — but not in a form an answer engine can extract and attribute.
Why Xpress is invisible here
- No answer-shaped content. FAQPage schema reaches only 34 of 236 pages. Answer engines lift short, self-contained answers; the pages lead with sales copy.
- No author or expert entity. Not one named person on the site — no owner, no technician, no bio. Engines cannot attribute expertise to an anonymous business.
- No first-party evidence. Zero case studies across 153 location pages. There is nothing specific to quote.
- No third-party corroboration. The link profile is dominated by directories and an inherited network, so there is no independent source describing Xpress as a San Diego restoration company.
- Reviews live only on the profile. With Yelp in 81% of these SERPs, review content is what the engines read — and none of it is on the site.
The strongest asset in the account.
The Google Business Profile is verified, well-reviewed and materially better than the website it points at. It is also the reason Xpress still wins Chula Vista.
| Profile area | Score | Read |
|---|---|---|
| Technical SEO | 100 | Complete |
| Posts & updates | 100 | Active |
| Reviews & reputation | 83 | Strong — 4.9 over 192 |
| Profile optimization | 81 | 3 items need attention |
| Photos & media | 50 | Thin |
| Services & products | 0 | Empty |
[ Data verified ] Search Atlas GBP audit report 17315, audit date 2026-07-27. Listing verified. Address on file: 1547 Jayken Way Ste B, Chula Vista, CA 91911. Citation score 91/100.
Three fixes, all inside the profile
- Phone number mismatch. The Search Atlas audit records a profile phone that does not match the (619) 607-3857 published site-wide and in the site’s own LocalBusiness schema. NAP consistency is a direct local ranking factor. Confirm in the profile before editing
- Services section is empty — scoring 0. This is where the mold, sewage and fire lines should be declared. It is also one of the few places an answer engine can read a structured service list.
- No opening date set. A founding date is an entity-trust signal for both Google and the answer engines. Founding year not stated anywhere on the site — confirm with the owner
Profile and site are pulling in different directions. The profile is a single, verified, well-reviewed Chula Vista business. The website behaves as if it were a 22-city regional operator. Google trusts the profile — which is exactly why the homepage takes all 174 monthly organic visits while 153 location pages take none.
Where the foundation stands today.
Fourteen dimensions, each scored against what we measured rather than what we assume. The teal bar is today. The coral tick is what is realistically achievable in twelve months on this domain, in this market.
4.4 today → ~7.6 achievable in 12 months.
The score is held down by architecture and evidence, not by technical debt or content volume — which is why it moves quickly once the structure is fixed.
Why each score was assigned
| Dimension | Score | Evidence |
|---|---|---|
| Technical SEO | 72/100 | All 249 crawled URLs return 200; 235 of 236 pages self-canonicalize; zero noindex leaks; 8,126 images at 100% ALT coverage. Held back by a malformed robots directive on 99 pages and two conflicting robots meta tags per page. |
| On-Page SEO | 55/100 | 215 of 236 titles (91%) exceed 60 characters and truncate in the SERP; 127 meta descriptions exceed 160; 41 pages share only 10 H1 strings. |
| Content Quality | 48/100 | Median 2,660 words per page — length is not the problem. The body copy is competently written and city-adapted, but it is built from one template and carries no first-party evidence. |
| Information Gain | 18/100 | Median local-signal score of 0 out of 10 across 153 location pages; 79 pages carry none at all. Zero pages contain a case study, a named technician, a response-time claim, or an original statistic. |
| Internal Linking | 25/100 | 38,230 internal links across 236 pages — an average of 161 per page. 166 URLs are linked from more than 90% of the site, so authority is spread almost perfectly flat and no page is signalled as important. |
| Local SEO | 62/100 | Google Business Profile scores 78/100 with a 4.9 rating over 192 reviews and is verified — the strongest asset in the account. Undercut by a phone-number mismatch between the profile and the site. |
| Entity SEO | 58/100 | LocalBusiness schema on 221 pages with consistent NAP and six sameAs profiles. No founder, no founding date, and no author entity anywhere on the site. |
| Topical Authority | 30/100 | Seven service lines are built across 153 location pages, but mold — the second-largest demand pool on the benchmark at 742 searches a month — has a single page and no city coverage. |
| E-E-A-T | 28/100 | No named authors, no technician bios, no project photography, no case studies. The claim "We've helped 3,100+ Property Owners" appears on 62 pages against 22 different cities, which reads as a template variable rather than evidence. |
| Structured Data | 80/100 | BreadcrumbList on 235 pages, LocalBusiness on 221, zero pages without schema. FAQPage reaches only 34 pages — the clearest quick win for answer engines. |
| Conversion Architecture | 55/100 | Phone number, 60-minute response promise and license number are prominent site-wide, and a financing page exists. No location-specific proof, no reviews on the page a searcher actually lands on. |
| Authority / Backlinks | 22/100 | Domain authority 21 with 1,597 backlinks from 311 referring domains, but 710 anchors are other businesses' domain names — the signature of an inherited link network, not earned coverage. |
| GBP ↔ Site Alignment | 45/100 | The profile targets Chula Vista and converts well there. The site's architecture spreads across 22 city slugs with no matching proof, so the profile's strength does not transfer to the organic footprint. |
| AI Search / GEO Readiness | 12/100 | Zero mentions across 24 answers from ChatGPT, Gemini and Perplexity — including the Chula Vista question, where Xpress ranks first organically. Competitors are named instead. |
Do not delete the pages. Consolidate them.
The instinct to build local pages is correct — Quick Dry and Gold Coast prove it works in this exact market. The fix is not deletion. It is giving every intent exactly one page, and giving that page something worth reading.
The target structure
| Layer | What it is | How many | Job |
|---|---|---|---|
| Core service hubs | Water damage · Emergency water removal · Flood cleanup · Sewage cleanup · Fire & smoke · Mold remediation | 6 | Own the non-geographic head terms and define the service for answer engines |
| Market hubs | San Diego · South Bay (Chula Vista, National City, Imperial Beach, Bonita, Eastlake) · East County · Coastal | 4 | Own “water damage restoration san diego” and route to the city pages |
| Service × location pages | One page per service per priority market — not six | ~40 | Win the local pack and the local organic result |
| Evidence layer | Case studies, project galleries, city FAQs, insurance guides | Ongoing | Supply the information gain, E-E-A-T and citable answers every layer above depends on |
What happens to the 153 pages that exist
| Group | Pages | Action | Reason |
|---|---|---|---|
| Duplicate service+city combinations — the winner of each pair | 23 | Merge | Absorb the unique copy from the losing URL, then serve one page per intent. Winner chosen by measured position, then depth. |
| Duplicate service+city combinations — the second URL | 23 | 301 redirect | Two URL patterns for one query is the direct cause of the 41.8% cannibalization rate. Redirect consolidates the split signals. |
| Location pages in the 10 priority markets | 43 | Improve | Keep the URL, rebuild the content against the information-gain checklist: a real project, named neighborhoods, local risk, response logistics, a review from that city. |
| Location pages in North County markets | 64 | Canonicalize | Carlsbad, Oceanside, Vista, San Marcos, Encinitas, Escondido, Del Mar, Solana Beach and Santee sit 30–50 miles from the Chula Vista base with no profile presence and no proof. Point them at the nearest market hub until there is genuine coverage to show. |
| Seven overlapping San Diego water URLs | 7 | Merge 301 | Consolidate into one San Diego market hub plus one page per service. This is the single highest-value consolidation on the site. |
| Blog category archives | 5 | Noindex | Zero inbound internal links, thin, and already disallowed in robots.txt — but still indexable. Align the signals. |
/24-7-emergency-water-restoration-san-diego/ | 1 | 301 redirect | A San Diego money page that no other page links to. Fold its copy into the San Diego hub. |
| Mold remediation | 1 → ~11 | Build | 742 searches a month, 0 of 10 rankings, one page. The largest uncontested opportunity in the account. |
Net effect: roughly 31 URLs redirect into stronger pages, 64 canonicalize, 66 get rebuilt with real local evidence, and a new mold line opens against 742 monthly searches. The indexable count falls from 236 to roughly 147. The number of pages capable of ranking goes up.
Fix the link graph at the same time
Consolidation only works if the internal links stop pointing everywhere at once. Replace the all-cities mega-menu with a hub-and-spoke structure: homepage → service hub → market hub → city page. Keep the sitewide navigation to the six service hubs and the four market hubs, and let each market hub link to its own cities. That alone takes the average page from 161 outbound links to roughly fifteen and restores a hierarchy Google can read.
Sequenced so nothing is rebuilt twice.
Days 1–30 · Stop the confusion
Consolidate · Repair- Fix the malformed
robotsdirective on 99 pages and remove the duplicate robots meta tag from the template. - 301 the 23 duplicate-pattern URLs into their surviving counterparts, absorbing unique copy first.
- Consolidate the seven San Diego water URLs into one market hub plus one page per service.
- Replace the mega-menu with hub-and-spoke navigation; drop average outbound links from 161 to ~15.
- Remove the “3,100+ Property Owners in [City]” line site-wide; replace with one accurate, verifiable figure.
- Fix the broken
/sewage-cleanupoceanside/link and noindex the five category archives. - Correct the GBP phone number and fill the empty Services section.
Days 31–60 · Build for the demand
San Diego · Mold- Build the San Diego market hub and four service pages targeting the head terms worth 2,621 searches a month.
- Open the mold remediation line: one hub plus city pages for the priority markets — 742 searches a month currently unserved.
- Rebuild the 43 priority-market location pages against the information-gain checklist — one documented project, named neighborhoods, local risk factors, response logistics, and a review from that city on every page.
- Canonicalize the 64 North County pages to their nearest market hub.
- Rewrite the 215 over-length titles and 127 over-length descriptions.
Days 61–90 · Become citable
Evidence · AI search- Publish the first six case studies with photography, city, damage type, response time and outcome — the asset class the site has none of.
- Add FAQPage schema across the service and market hubs, currently on 34 of 236 pages.
- Introduce named expertise: owner and technician bios, IICRC certifications, an author entity on every guide.
- Syndicate reviews onto the relevant city pages so the review content lives on the site, not only on the profile.
- Re-run the 100-query benchmark and the AI visibility test to measure movement against this baseline.
The first five moves, in order.
| Action | Effort | Why it matters | Estimated impact |
|---|---|---|---|
| Consolidate the 23 duplicate service+city pairs | Medium | Removes the direct cause of a 41.8% cannibalization rate | Recovers split ranking signals on 76 keywords |
| Build the San Diego layer | High | 92% of benchmark demand, currently 0/10 coverage | Access to 2,621 searches/mo at $52–$62 CPC, difficulty 3–25 |
| Open the mold line | Medium | Second-largest demand pool, one page behind it | 742 searches/mo currently unserved |
| Fix the robots directive | Low | Governs how 42% of pages may appear in results | One template change, 99 pages |
| Add first-party evidence | High | 79 of 153 location pages carry zero local signals; zero carry a case study | The precondition for both information gain and AI citation |
The first five moves
- Redirect the 23 duplicate-pattern URLs and consolidate the seven San Diego water pages into one hub.
- Ship the San Diego market hub and the four head-term service pages.
- Correct the robots directive and remove the duplicate meta tag from the template.
- Replace the “3,100+ Property Owners” template line with one verifiable figure, and fill the empty GBP Services section.
- Publish six real case studies with photography — one per priority market.
Bottom line for Xpress. This is not a site that needs rebuilding. The technical foundation is sound, the schema is thorough, the profile is strong, and the copy is competently written. What is missing is a structure that tells Google which page matters, and evidence that tells a homeowner — and an answer engine — why Xpress is the one to call.
153 pages currently earn 0 visits a month. The same effort, pointed at 92% of the demand instead of 8%, and consolidated so one page owns each intent, is a materially different business — and it is buildable in 90 days.
[ Next step ]
Let’s walk through the battlefield together.
Thirty minutes on a call and we will show you exactly which pages to merge, which to rebuild, and what the San Diego layer looks like in practice.
Book a Call with Bonsai Marketingor reach us directly · fikeshway@bonsaimarketingcompany.com · (707) 593-7539