Georgia care provider website study.

We could read the home page of 760 sites. Each check below looks at the raw HTML of that one page.

That leaves 156 sites (20.5%) with no phone number in the HTML we read, 265 (34.9%) with no meta description and 286 (37.6%) with no schema. And 56 (7.4%) have no mobile viewport tag. The tag is a strong sign of whether a page was built for phones, but we did not view each site at phone size.

Of 4,759 unique care providers, 2,179 (45.8%) have no website listed in the state directory. We counted a provider by grouping records with the same name and ZIP code. The field is optional on enrollment, so some of these providers may have a site or a Facebook page that we did not look for.

The table counts directory records, not providers. One provider can have several records, so 69.0% for nursing facilities means 240 of 348 records, not 69.0% of facilities. A record counts as no website listed when the field is blank or holds an entry we could not use.

The 130 and the 42 are not a count of broken sites or closed businesses. Some sites block automated visitors, and a few listed addresses are typos or retired.

Among the 791 that loaded, plain http:// redirected to https:// on 656 (82.9%). On 82 (10.4%) it still loaded with no redirect. Of the sites that redirected, 27 (3.4% of the 791) ended on a different domain from the one listed.

We ran Google’s Lighthouse test on a random sample of 100 readable sites, and 95 returned a score. In this simulated mobile test the median performance score was 62 out of 100, with 11 sites (11.6%) at 90 or higher and 19 (20.0%) below 50.

Largest contentful paint, the time until the biggest piece of content on the first screen appears, had a median of 6.4 seconds. It was 2.5 seconds or less on 11.6% of the 95 and over 4 seconds on 76.8%.

These are lab figures: one run per site, simulated slow-mobile throttling, on a computer that was not idle. They run slower than what visitors on fast connections see, and they should not be compared with Google’s field data.

The source is the Georgia Medicaid Active Provider Directory, in a file dated 2026-06-01. We kept Home and Community Based Services (which include residential group homes), Nursing Facilities, Nursing Services, and Agencies limited to home health and hospice, which came to 8,380 records.

We also added the 8 domains from two short lists of care homes in Virginia and Miami that Luziq keeps for sales. They are not Georgia Medicaid records.

We cut those to unique domains and dropped social media pages, shared builder domains and email hosts. Luziq, which builds websites and ran this study, also removed its own clients and the prospects it had built demo sites for. It did not remove everyone on its sales lists, so some sites here belong to businesses we have contacted or plan to contact.

That left 1,344 domains, and a seeded random shuffle (seed 20261105) picked 1,000. Of those, 986 belong to Georgia-enrolled providers and 14 are scattered across 9 other states. 5 of the 14 came from the two added lists. They are the 5 rows typed ALF/group home in the data, and they are counted in every figure on this page.

Each domain got one status request. Sites that answered got one redirect check. Each site that loaded normally got one home page read, and we skipped the 5 whose robots.txt disallowed it.

data.csv has one row per sampled site. Ids are anonymous (A0001 and so on), and the file holds state, provider type and measured fields, with no domain, name, address or phone.

State reads GA or other: this published copy merges the 14 non-Georgia rows into other. Empty cells mean not measured. This study names no provider.

This study checks whether the basics are there. For what else belongs on a care home’s site, read What should a group home website include?

Luziq builds websites for care homes. If you would rather hand this list to someone, see what we do.