Real offices, real care,
and twelve copies of one page.
51 pages crawled and the key ones rendered in a real browser. The office pages are genuine and well linked. Around them sits a layer of copied pages that no crawled page links to, a mobile load time in the poor band, and no structured data anywhere.
EmmUcare has the material a home care agency needs to win local search: 6 office pages that name real counties, a named founder with a nursing credential, a real FAQ, and history back to 2017. The problem is what has been built around it. 12 "near me" pages carry the same body text with a different title, and 8 town pages are thin copies of one template; none of those 20 pages is linked from any of the 51 crawled pages, a pattern Google's spam policies call doorway pages. The homepage scores 22 on mobile with a 10.7 second main image. No page carries structured data. And the fax number is a tap-to-call link on 22 pages. The first moves are to fold the copies into the office pages, fix the fax link, and get read-only Search Console access so the next decision rests on data.
01Every page, grouped by what it really is
Each of the 51 crawled pages, sorted by template rather than by URL. Red is a page that should not exist in its current form, amber a page that needs work or should leave the sitemap, blue the office pages that carry the business, white everything else.
02Twelve pages, one page
Text overlap between every pair of pages in each group, measured as shared runs of eight words. The negative control is two pages known to be different, so the scale has a floor.
| Page | Title | Words | Links in |
|---|---|---|---|
| /caregivers-near-me-1 | Caregivers Near Me | 1420 | 0 |
| /in-home-care-in-pa-1 | In-Home Care in PA | 1421 | 0 |
| /best-pa-home-care-near-me | Best PA Home Care Near Me | 1423 | 0 |
| /private-care-near-me | Private Care Near Me | 1421 | 0 |
| /senior-care-near-me-in-pa | Senior Care Near Me in PA | 1423 | 0 |
| /personal-care-near-me | Personal Care Near Me | 1421 | 0 |
| /home-care-near-me-in-pa | Home Care Near Me in PA | 1423 | 0 |
| /top-caregivers-near-me | Top Caregivers Near Me | 1421 | 0 |
| /elderly-care-near-me | Elderly Care Near Me | 1421 | 0 |
| /top-home-care-near-me | Top Home Care Near Me | 1422 | 0 |
| /care-at-home-near-me-pa | Care At Home Near Me PA | 1423 | 0 |
| /care-at-home-near-me | Care at Home Near Me | 1422 | 0 |
03The town pages
The right idea, eight real places the Williamsport office serves, built as copies of one short template.
| Page | Words | Description | Links in |
|---|---|---|---|
| /home-care-duboistown-pa | 318 | present | 0 |
| /home-care-williamsport-pa | 318 | missing | 0 |
| /home-care-in-17701 | 319 | missing | 0 |
| /home-care-montoursville | 321 | present | 0 |
| /home-care-montgomery-pa | 318 | present | 0 |
| /home-care-jersey-shore-pa | 320 | missing | 0 |
| /home-care-muncy-pa | 307 | missing | 0 |
| /home-care-hughesville-pa | 322 | missing | 0 |
04Speed on a phone
Homepage, PageSpeed Insights API, mobile strategy, median of three runs.
Time bars share one 12 second scale; blocking time is drawn on a 2 second scale and layout shift on a 0 to 0.5 scale.
05Images without alt text
Rendered in a real browser, counting images with no alt attribute at all.
Across the server HTML of all 51 pages: 279 of 406 images, on 31 pages.
06Structured data, sitemap and the rest
| Check | Result |
|---|---|
| Structured data (JSON-LD and microdata) | 0 of 51 pages |
| Sitemap addresses | 51 of 133 listed on http://, each redirecting to https; robots.txt points at the http sitemap |
| Store product pages in sitemap | 81 listed; one rendered, no Product markup, two H1s. Not crawled in full |
| Account screens in sitemap | 4: /m/login, /m/reset, /m/create, /m/create-account |
| Pages empty in a real browser | 3: /find-care-by-map, /news-and-updates, /store-terms-and-policies |
| Meta description missing | 17 pages, 8 of them near-me duplicates |
| Canonical tags | Present and absolute on all 51 |
| HTTP and www | Both redirect to https://emmucarehomehealth.com/ |
| 404 handling | Unknown URLs return a real 404 |
| AI crawler access | GPTBot, ClaudeBot, PerplexityBot, OAI-SearchBot and Googlebot: HTTP 200 each |
| llms.txt | Present, generated by the builder |
| Mobile horizontal scroll | None at 390px |
| Search Console and Analytics | Not available, no access granted |
| Index coverage | Unknown. Not checked at tier 2 |
| Backlinks | Not checked in this audit |
07Findings
Eleven findings, ordered by what they cost to leave alone. Each carries a severity and one of the four categories used across this report.
12 pages are the same page with a different title
Twelve URLs such as /elderly-care-near-me, /top-home-care-near-me and /private-care-near-me carry the same body text. Measured as overlapping eight-word runs, every pair shares between 99.2% and 99.9% of those runs. Set /caregivers-near-me-1 beside /in-home-care-in-pa-1 and the only words that differ anywhere on the page are "Caregivers Near Me" and "In-Home Care in PA". None of the 51 crawled pages links to any of the twelve; they are reachable through the sitemap. That combination, many near-identical pages each aimed at one search phrase and reachable only through the sitemap or a search result, matches the pattern Google's spam policies describe as doorway pages. It is a risk, not a verdict, but it is the one item here that can pull the rest of the site down with it.
/caregivers-near-me-1 and /about-us returns 6.7%, so the method separates different pages cleanly. Inbound links counted from every anchor on all 51 crawled pages: 0 for each of the 12.The homepage scores 22 out of 100 on mobile and takes 10.7 seconds to show its main image
PageSpeed Insights, mobile, median of three runs: performance 22, Largest Contentful Paint 10.7s, Total Blocking Time 1231ms, Cumulative Layout Shift 0.26, First Contentful Paint 6.6s. Google treats LCP over 4 seconds and CLS over 0.25 as poor, so the page is in the poor band on both. Families searching for care do it on phones. Chrome has not published field data for this site, so the lab figure is the only public measurement available.
No structured data on any of the 51 pages
Not one crawled page carries JSON-LD or microdata: no Organization, no LocalBusiness for any of the 6 office pages, no FAQPage on the one page that has a real FAQ, no Product on the store items. Structured data is how a search engine reads an address, a phone number and an opening time as facts rather than as text, and it is the input for several rich-result types. For a business with offices in two states it is the cheapest way to make each office legible on its own.
itemscope/itemtype: 0 and 0. Confirmed in a real browser on 9 rendered pages, including a store product page: 0.No Search Console and no Analytics access, so index coverage is unknown
This is an outside read. Without Search Console nobody can say which of these pages Google holds, whether the twelve duplicates in finding 01 have already been filtered out, or which queries the site surfaces for. Every click, impression, ranking and conversion figure is absent from this report by design. Read-only access to both turns the next round from inference into measurement.
8 town pages share one template and no crawled page links to them
Pages for Duboistown, Williamsport, Montoursville, Montgomery, Jersey Shore, Muncy, Hughesville and the 17701 ZIP code run 307 to 322 words each, a count that includes the site navigation, and on average share 83.1% of their eight-word runs. 5 of the 8 have no meta description. None of the 51 crawled pages links to any of them. These are the right idea (real towns the business serves), built as thin copies instead of as pages with something only that town would need: the office that covers it, travel area, a local caregiver or client story.
The fax number is a tap-to-call link on 22 pages
The contact block reads "Tel: (570)-666-3856 / Fax: (570)-666-3958", and both numbers are wrapped in a tel: link. On a phone, tapping the fax number dials a fax machine. On a site whose whole job is a phone call, that is a lost lead every time it happens. It sits on all 12 near-me pages, 7 of the 8 town pages and 3 office pages.
href="tel:5706663958": 22 pages. The fax label next to the number was read from the page text. Confirmed in a real browser on the Williamsport office page.279 of 406 images have no alt text, across 31 pages
Alt text is what a screen reader announces and what an image search reads. 231 of the 279 sit on the 12 near-me pages (126) and the 6 office pages (105). In a real browser the Williamsport office page shows 23 of its 51 images without an alt attribute, and the near-me template 12 of 14. The About page is the model to copy: 1 of 25.
The sitemap lists login screens, empty pages and old http addresses
All 50 addresses in the main page sitemap, one more in the booking sitemap, the child sitemap addresses in the index and the sitemap line in robots.txt use http://, and every one redirects to https. The blog and store sitemaps already use https. 4 store account screens (/m/login, /m/reset, /m/create and /m/create-account) are listed as content. 3 pages, /find-care-by-map, /news-and-updates and /store-terms-and-policies, showed only the navigation, footer and cookie notice in a real browser; their only H1 is "This website uses cookies." 17 pages have no meta description.
/news-and-updates shows a blank body between header and footer.The strongest trust signals sit on one page and carry no source
The About page names the founder, "Kayla Borgess, LPN", gives a history back to May 2017 in Williamsport, and says the agency earned "recognition as Central Pennsylvania's Best Home Care Agency for two consecutive years". None of that reaches the service or office pages, and the award names no awarding body or year, so a careful reader cannot check it. Name the source and carry the founder's credential onto the pages families actually land on.
Two snippets that Google will cut
The Contact page title runs 101 characters and the homepage meta description 285. Google trims by pixel width rather than a fixed count, but both are well past what shows. Everything after the cut is invisible in the result.
The basics that are already right
HTTP and www both redirect to the single https address. An unknown URL returns a real 404. GPTBot, ClaudeBot, PerplexityBot, OAI-SearchBot and Googlebot were each served the homepage with HTTP 200, and an llms.txt exists. The homepage has no horizontal scroll on a 390px phone. The builder handles these, and none needs work.
08Who owns what
Established from evidence. The site runs on GoDaddy's Websites + Marketing builder: the server identifies
itself as GoDaddy's DPS, assets load from wsimg.com, the security policy names
godaddy.com and secureserver.net, and /wp-json/ returns 404. The domain
was registered through GoDaddy on 31 May 2024, the same year as the rebrand the About page describes. The fixes in
this report belong to whoever holds the GoDaddy login. Where the builder cannot do something directly, such as
a permanent redirect or custom code in the page head, the fallback is to unpublish the page or to move the site to
a platform that can; that is a decision to make once Search Console shows what the site is earning.
09The plan, cheapest first
- Grant read-only Search Console and Analytics accessTwo minutes, no cost to the site. It answers whether Google has already filtered the duplicates.
- Unlink the fax numberChange the fax line in the contact block to plain text on all 22 pages that carry it.
- Redirect or unpublish the 12 near-me pagesSend each to the office or service page that best answers it, or take it down, and drop it from the sitemap. Fold the search phrases into those pages' copy.
- Take the account screens and empty pages out of the sitemap4 account screens and 3 empty pages. Fill the empty pages or unpublish them.
- Add Organization and per-office LocalBusiness markupName, address, phone and service area for each of the 6 offices, plus FAQPage on the services page.
- Write alt text for the 279 imagesStart with the office pages, where the gaps are largest.
- Lighten the homepageReplace the text-in-image banner with real text over a compressed photo, and cut the image count above the fold.
- Rebuild the 8 town pages or merge themEach needs something only that town would use, or it should redirect to the office that covers it.
- Source the award and spread the founder's credentialName who gave "Best Home Care Agency" and when; add the LPN founder to service and office pages.
- Then decide on content and local search workWith Search Console connected, on data rather than on this audit.
The first four steps are housekeeping.
Access, the fax link, the duplicate pages and the sitemap cleanup are small jobs for whoever holds the GoDaddy login, and they remove the one risk in this report that could affect the whole site. The steps after that are where the actual local search program starts, and what it should be depends on what Search Console shows once it is connected.
Scope and timing are a conversation. If you want to walk through any finding here, get in touch.