9.9% of nursing homes with CMS's lowest staffing rating offer daily pay on their websites.

We started from CMS's Provider Information file for nursing homes, processing date September 1, 2026, which lists every nursing home certified by Medicare or Medicaid. We kept the 14,680 in the 50 states and DC, one per CMS certification number.

For each home we searched Google for its name, city and state. Jev, a model from TypeSafe that answers questions with a probability, picked the home's own website or its page on its operator's website from the results. Directories, review sites and job boards never counted. We read that page and the careers, job list and admissions pages it links to on October 26, 2026, within each site's robots.txt.

Jev first checked that the pages were about this home, by its name with its address or phone. It then answered every question in one pass. A yes counted only when Clairic, Cloudflare's extraction model, copied the exact sentence from the page and Jev, reading that sentence, agreed. Jobs, bonuses and pay on an operator's careers site counted only when the page tied them to this home. A general invitation to join the team never counted as an open job.

Review agents then checked a sample of yes answers against the pages. They agreed on 40 of 50 nursing jobs, 44 of 50 sign-on bonuses, 45 of 50 pay figures and 48 of 50 daily pay offers. We scaled each of those shares by its agreement rate, so they are estimates. They checked every agency, admissions and closing answer, and the page counts only the ones they confirmed (12 of 15, 37 of 40 and 2 of 5). In a separate check of 60 homes, reviewers found a few offers our reading missed (2 of 3 sign-on bonuses, for example), so the shares are more likely low than high.

Of the 14,680 homes, 5,022 had no website we could find. 133 had only their operator's general page, and for 489 the pages turned out to be about another home or couldn't be tied to this one. 1,372 had a site we couldn't read (946 blocked us, 338 didn't answer, 49 returned an error, 35 asked robots not to read them and 4 were empty or parked). 9 weren't searched or read to the end. That left 7,655 homes whose own pages we read, 4,629 on their own website and 3,026 on their operator's site. Shares on this page are of those 7,655.

Staffing ratings, chains and turnover come from CMS's same file. A home is in a chain when CMS gives it a chain ID. Turnover is CMS's total nursing staff turnover, split into thirds of the homes we read, at 38.7% and 51.2%. Shares by rating and turnover are scaled by the same agreement rates. 87 homes we read have no staffing rating and are left out of that chart.

A website shows what a home advertises, not how well it's staffed or how many jobs are open. Homes whose sites we couldn't find or read are counted above, not dropped. Offers on an operator's chain-wide job board counted only when tied to the home, so homes in chains may advertise more than we count. We also asked whether sites offer free nurse aide training or pay toward nursing school, and left it out because the review didn't support the answers.

The CSV has the counts behind the page by CMS staffing rating, chain, ownership, turnover, bed count and kind of site, plus the funnel. Its counts are confirmed answers before the review's scaling. It's free to use under CC BY 4.0 with attribution to Fulkerson Advisors.

Fulkerson Advisors (2026). What nursing home websites say about staffing: 14,680 US nursing homes, October 2026. https://www.fulkersonadvisors.com/research/nursing-home-staffing

The charts and the CSV are free to use with a link to this page.