Google Ads for care homes: why you keep paying for job applicants

When a care home concludes that "Google Ads didn't work", the reason is rarely the one the operator expects. Pull the search terms report and, more often than not, somewhere between a third and two thirds of the spend has gone on people looking for a job rather than a bed.

Why it happens

Care is one of the few sectors where the customer and the job applicant search using almost identical language.

"Care home Preston" is a family looking for a bed. "Care home jobs Preston" is somebody looking for work. So far so obvious — you would add "jobs" as a negative keyword and move on. The problem is that the overlap is much wider than the word "jobs":

• care assistant, carer, care worker, support worker
• vacancies, hiring, apply, recruitment, staff
• HCA, healthcare assistant, night carer, bank staff
• "care home near me" typed by someone who wants to work at one
• the name of your own home, typed by an applicant checking you out

And crucially, if the campaign uses broad match — which is the default Google nudges you towards — a bid on "residential care home Preston" can serve against "care jobs in Preston" without you ever seeing the connection unless you go looking.

The result is an account that generates plenty of clicks, a healthy-looking click-through rate, and almost no enquiries. Which reads, from the outside, exactly like "Google Ads doesn't work for care homes".

How to check in ten minutes

You do not need an agency to diagnose this.

1. Open the Google Ads account and go to Insights and reports → Search terms.
2. Set the date range to the last 90 days.
3. Sort by cost, descending.
4. Read the top fifty rows.

You are looking for anything that is clearly a person seeking employment. Add up the cost column for those rows. In a care home account running without a proper negative keyword list, that number is routinely between 30% and 60% of total spend.

If you cannot access the account because an agency holds it, that is a separate and more serious problem — you should own your Google Ads account outright, always.

Fixing it properly

The instinct is to add a negative keyword list and stop there. That helps, but it treats the symptom.

Build a shared negative list, not a campaign one. Employment terms will keep appearing in new forms. A shared list applied across every campaign, reviewed monthly, is the only version of this that stays fixed.

Stop using broad match until the account has conversion data. Broad match works well in accounts with a rich conversion history to learn from. In a new or rebuilt care home account it is an expensive way to discover what you do not want. Exact and phrase first; open up later.

Separate recruitment properly, if you advertise for staff at all. Many groups do want to advertise roles. That belongs in its own campaign, with its own budget, its own landing page and its own conversion action — not competing for the same pound as a family enquiry.

Point families at a page built to take an enquiry. A large share of care home ads land on the homepage. A family who has clicked an ad about dementia care should arrive at a page about dementia care, with fees addressed honestly, a way to book a visit in two clicks, and a form they can complete at eleven at night without ringing anybody.

The structure that works

Once the waste is out, the structure that consistently produces enquiries is unglamorous:

• One campaign per home. Not per group. A group campaign spends the Harrogate budget in Hull, because Google optimises for cheap conversions and does not know that beds are not interchangeable between sites sixty miles apart.
• Ad groups by specialism. Residential, nursing, dementia, respite. Different family, different urgency, different ad, different landing page.
• Tight radius targeting. Most families place a parent close to themselves or close to where the parent lived. A 30-mile radius on a rural home is usually generous; a 10-mile radius on an urban one is often enough.
• Modest budgets, precisely spent. A single home rarely needs more than £600 to £1,500 a month. The maths does not require aggression: at a typical private fee of £1,200–£1,600 a week, one additional self-funded admission is worth £60,000–£83,000 a year.

The bit nobody wants to hear

You can fix all of the above and still see no improvement, because the enquiry is being lost after it arrives.

We routinely find enquiry forms going to an inbox nobody owns, phone numbers that ring out after 5pm, and voicemails returned on Monday for calls made on Friday. A family working to a hospital discharge deadline will have booked two visits elsewhere by Monday.

Before spending more on advertising, it is worth doing something slightly uncomfortable: ring your own home on a Saturday morning as a member of the public, and submit your own enquiry form on a Sunday evening. Time how long the response takes. In a meaningful number of cases, the cheapest available improvement in enquiry volume is not in the ad account at all.

Where ads fit alongside AI search

Google Ads reaches the family searching this week, which makes it the right tool when you have beds to fill now. It works immediately, it costs money every month, and it stops the day you switch it off.

What it does not do is get you onto the shortlist an assistant writes before anybody searches. That is a slower, compounding piece of work on your CQC record, your directory listings and your published content — and the two are complements rather than alternatives. Ads for this quarter's admissions, AI search visibility so that next year needs less advertising than this one.