Two families call your office on the same Tuesday, both from the same ad. One wants a caregiver 30 hours a week and will pay privately. The other qualifies for a Medicaid program that pays your agency the state's hourly rate. Both leads cost you about the same click. They are not worth the same amount, and a bidding setup that treats them as equal will overpay for one and underbid for the other.
This chapter sets the bids. It starts with manual bids in the first campaign, shows which numbers tell you whether a bid is working and who you're bidding against, and then gives each payer its own target so the automated bidding in the second campaign knows what each family is worth.
Dementia Care and Home Health Clicks Start at Different Prices
In our method's first campaign, the exact-match Isolation campaign from Chapter 6, you set bids by hand, one ad group at a time. That is deliberate. A new home care account hasn't produced a single client lead yet, so there is nothing for automated bidding to study. Hand bids let the owner choose how hard to push dementia care versus companion care from day one.
Each service arrives at its own price. In our example market, Chicago, Keyword Planner's estimates ran $4.20 to $16.20 for "in home care," $1.58 to $11.73 for "dementia home care," $3.28 to $12.06 for "home health care near me" and $2.75 to $10.00 for "medicaid home care." The spread inside each range is as telling as the gap between them.
Your market's numbers will differ, but the lesson holds. Give each service its own starting bid, near the upper half of its range while the account is new, and adjust each one separately. A single bid across every ad group will be too high for some services and too low for others.
Impression Share Says Whether the Bid Is Winning, and Why Not
After a week or two, four columns tell you whether a bid is working. Add them to the Keywords and Ad groups views:
- Search impression share: the impressions you received divided by the ones you were eligible for. It tells you whether the family ever saw you.
- Search top and absolute top impression share: the share of those showings that landed above the unpaid results, and the share that landed in first place. For the searches that matter most, our method aims for 75% to 90% at the top.
- Search lost impression share (rank): the share of the time your ad didn't show because its Ad Rank was too low. The fix is a higher bid, a better ad or a better page, as Chapter 4 explained.
- Search lost impression share (budget): the share of the time your ad didn't show because the campaign ran out of money. The fix is budget, not bids. Google reports this one at the campaign level only.
The two "lost" columns are the useful pair. A campaign losing mostly to rank needs a better ad, page or bid. A campaign losing mostly to budget is doing fine and simply needs more money, or fewer keywords sharing the money it has.
An illustration: suppose the dementia care ad group shows 35% top impression share, with most of the rest lost to rank, while the companion care ad group sits at 80%. The dementia group isn't short of money. It is losing auctions, so look at its ad and its page before you raise its bid, and raise the bid only if the page already answers a dementia family's questions. Once each ad group reaches the top most of the time, try lowering its bid a step at a time and watch whether the share holds. As Quality Score builds, the same position usually gets cheaper.
Franchise Offices and Referral Sites Share Your Auction
Google's auction insights report shows who else appears in your auctions and how you compare. Row by row, it lists every agency and referral site that appeared in the same auctions and scores the matchup: overlap rate (you both showed), position above rate (they sat higher), outranking share (you beat them or they didn't show), and each side's top-of-page and first-place rates.
In home care, the names in that report tell a specific story. In our example market:
- Franchise brands bid through many accounts at once. In Google's Ads Transparency Center, Always Best Care ran Search ads through 12 accounts, Assisting Hands through 11 and Visiting Angels through 10, a mix of national and local-office accounts.
- Referral sites bid on the same searches. A Place for Mom ran ads for two domains from one account, and referral-site ads led with cost and "free" help, such as "Free Care Guidance."
- Independents fill the rest, each with its own account and budget.
Run the same report on your brand campaign. If a franchise office or a referral site keeps showing on searches for your agency's own name, you'll see it there first. Keep the brand campaign's bids high enough to hold the top spot on your name, which is usually cheap because your ad is the most relevant one for that search. Bidding on their names in return is a separate decision, made deliberately in its own campaign; in our method it is the exception, not a reflex.
You don't have to beat all of them everywhere. Auction insights shows which competitors outrank you on which ad groups. When the only way to reach the top of a particular auction is a bid higher than a client is worth, our method says step out of that auction and spend the money where you can win.
Each Payer Gets Its Own Ceiling
A ceiling is the most you can pay to win one client. Chapter 3 worked one out for a private-pay client. Swap in each payer's numbers from your own books and the same math prices the others:
- Private pay: take what you keep from each billed hour, multiply by weekly hours and by months of service, and decide what slice of that you'll spend to win the client. Chapter 3's illustration came out at $500 per client.
- Medicaid: the same formula, but the state sets the rate. KFF, a health policy research group, found the median Medicaid rate paid to personal care agencies was $26 an hour. Illinois, our example state, pays $30.80 and requires at least $18.75 an hour for the worker. Less margin per hour means a lower ceiling, unless Medicaid clients stay much longer.
- Medicare home health: priced by the 30-day period of care rather than the hour. Medicare's 2026 national standardized payment is $2,038.22 per period, adjusted for each patient. Your ceiling comes from your margin per period and how many periods a typical patient needs, including the ones who come back.
A Medicaid lead also has to clear a gate before it is worth anything. Illinois's Community Care Program, as one example, serves people 60 and older with non-exempt assets of $17,500 or less who score at risk of nursing facility placement on the state's needs assessment, and enrollment runs through the state rather than the agency. Other states set their own rules. A family that calls about Medicaid-funded care but doesn't qualify is a conversation, not a client, so a Medicaid campaign's target should be built on qualified leads, not every call.
We're deliberately not printing a Medicaid or Medicare ceiling here. Rates, wages, hours and lengths of stay vary too much from state to state and agency to agency for any single number to be honest. Your client records have the real inputs.
One Target CPA Can't Price Three Payers
Once the account moves to what our method calls Exploration, broad match with automated bidding (Chapter 6), the ceilings become targets. Under a Target CPA strategy, you name the cost per lead you want and Google bids to win as many leads as it can around that figure. It treats that number as an average: some conversions will cost more than the target and some less, and it works to keep the overall cost near the number you set.
That is exactly why one target can't serve three payers. Set a single target that blends private pay, Medicaid and Medicare leads, and the bidding treats every lead as worth the blend. It will bid too much for searches that mostly bring Medicaid leads and too little for the private-pay searches that could carry a higher cost.
Google lets you set Target CPA per campaign or share one across several campaigns, and it discourages setting separate targets at the ad group level because that can restrict the bidding. So the clean answer is structural: give each payer its own Exploration campaign, with its own ads, pages and budget, and its own target, seeded from what Isolation actually paid per lead for that payer. Chapter 15 builds those campaigns in full.
Listen to the Calls Before You Trust the Cost per Lead
A low cost per lead can hide a problem. If a campaign's leads are cheap because a share of them are job applicants, families outside your service area or people shopping for a nursing home, the cost per real client is much higher than the report suggests.
So listen. Each week, take five calls from each payer's campaign, using the length and time stamps that Chapter 1's call reporting provides, and find out who was on the line and what they wanted. The pattern tells you whether a target is working or merely cheap. A campaign that looks expensive on paper but fills your schedule with private-pay clients is the one to feed.
If you record calls to do this, choose the tool carefully. Home health agencies, and any agency that bills Medicare or Medicaid electronically, fall under HIPAA, so recordings of patients and families belong in a tool whose vendor will sign a business associate agreement. Have counsel confirm your setup. (This isn't legal advice.)
Next: Families Praise the Office That Calls Back. Your Ads Should Say So.
The right bid gets your ad in front of the family. The ad itself decides whether they click. Chapter 8 writes home care ads from what families actually praise in reviews, from the office that answers the phone to the caregiver who shows up every time.
If you're starting here, Chapter 3 works out what a client is worth, and Chapter 6 sets up the two campaigns these bids belong to. The guide home lists all fifteen chapters.
Under our Google Ads management, a home care account never shares one target across private pay, Medicaid and Medicare.
