Most cleaning companies set one bid and walk away. It feels tidy. One number, one campaign, one line on the bill.
It's also a bid that's wrong twice. It's too high for most house cleaning searches and too low to show up for office cleaning at all.
In our example market, Seattle, the price of being among the top ads for an office or janitorial search runs about two and a half times the price for a general house cleaning search. Nationally the gap is narrower, but still wide. Those two customers don't cost the same, don't book the same way and aren't worth the same. They shouldn't share a bid.
This chapter covers how to set the first bid for each service, which report columns tell you when to move it, how to see who else is bidding, and how the numbers you collect become the target Google's automation will chase later.
Office and House Cleaning Are Priced Differently, So Each Needs Its Own Bid
Keyword Planner prints two prices beside each cleaning search and labels the pair "top of page bid." The higher price approximates what past advertisers paid for a spot among the top ads. It's an estimate, not a bill, but it's the best public read on what a search costs to win.
We sorted the example market's Keyword Planner ideas by service, kept the ones with at least 30 searches a month, cleared out job, product, price and franchise searches, and took the median high end for each service. Then we did the same for the whole country:
- Office, commercial and janitorial: $41.20 in Seattle, $20.21 nationally
- Move-out and move-in: $25.63 in Seattle, $19.93 nationally
- General house cleaning and maid service: $17.07 in Seattle, $12.03 nationally
- Deep cleaning: $15.99 in Seattle, $13.07 nationally
Seattle shows the office gap at its widest, about 2.4 times. Across the country it's about 1.7 times. Either way, the most expensive cleaning click on the list is an office click, and the second is a move-out click. Single searches can stretch further: "office cleaning services" tops out at $58.77 in Seattle, against $18.40 for "house cleaning near me."
That's why the rollout in this guide starts with a campaign we call Isolation, BrandRocket's name for it, not a Google setting. It runs exact match keywords with bids you set yourself, which Google calls Manual CPC. You choose the most you'll pay for a click, and you pay nothing when nobody clicks.
Isolation keeps each service in its own ad group from Chapter 5, so each one gets its own bid. Office cleaning bids like office cleaning. House cleaning bids like house cleaning. Nothing averages them into a number that fits neither.
Your First Bid Is an Educated Guess, Priced Off Keyword Planner's Range
A new account has no history, so your first bids come from the range Keyword Planner shows for each ad group's searches. Three things about that range matter before you type anything.
- It's built from broad-match data. Your Isolation keywords are exact match, so treat the number as a neighborhood, give or take about 30%.
- Start toward the high end. A new account has no track record for Google to judge its ads and pages by, so early on it takes more bid to earn a good spot. Once each keyword earns a Quality Score, you can come down.
- Check it against what a customer is worth. Chapter 3 worked out a per-click ceiling for a recurring house cleaning customer: roughly $12 to $15, depending on close rate. The Seattle median high for house cleaning is $17.07. So the top of that auction is often more than the math supports. Bid up to your ceiling and let a competitor overpay for the rest.
Google offers two levels for Manual CPC bids. An ad group default bid applies to every keyword in the group. A keyword bid overrides it for one search. Google's own example is a bakery that bids more on apple fritters because those searchers buy more.
Start with the ad group default. Give a keyword its own bid only when the numbers say it earns one. For a cleaning company, that's often a service plus your busiest suburb, or the one move-out search that keeps filling the calendar. Don't hand out keyword bids on day one. You don't know yet which searches book.
Office cleaning gets its own ceiling. Chapter 3's math was built on a recurring house cleaning customer, and an office contract is priced differently. Use your own contract numbers. If a janitorial account is worth several times a house cleaning customer to you, a click that costs twice as much can still be the better buy.
Impression Share Shows How Often a Cleaning Customer Searched and Never Saw You
Once the ads run, bids move for one reason only: a column told you to. Add these to your keyword and ad group views.
- Search top impression share. Of all the times you could have shown among the top ads, how often you did. Top ads usually sit above the first organic results, though Google says they can sometimes appear below them. On the searches that matter most, like your main house cleaning search or move-out cleaning in your best suburb, aim for 75% to 90%.
- Search absolute top impression share. The share of chances where your ad sat in the first slot of all. Worth watching on two or three signature searches, expensive to chase everywhere.
- Search lost top impression share (rank). Chances you lost to Ad Rank, Google's ranking of your ad against the others. The cause is the bid, the ad and page quality, or both.
- Search lost top impression share (budget). How often you missed because the money ran out. Raising the bid here makes it worse, because each click costs more and the budget runs out sooner.
- Quality Score, by keyword. Google's 1-to-10 grade of your ad, your page and how often people click, compared with other advertisers. Google calls it a diagnostic, not part of the auction, but a low one explains a lot of lost rank.
Read them together. Lost to rank with a good Quality Score means the bid is short: raise it in small steps, and never past your ceiling. Lost to rank with a poor Quality Score usually means the ad or the page doesn't match the search, so fix that before paying more. Lost to budget means you're winning auctions you can't pay for. Cut bids on the weakest ad groups or move money from a service that isn't booking.
If a column shows a dash, Google doesn't have enough data yet. That's normal in the first weeks and on small searches. Wait for it.
Molly Maid, Thumbtack and Angi Bid on Your Searches. Auction Insights Shows When to Leave.
We looked up the big cleaning franchises and home-services marketplaces in Google's Ads Transparency Center. Only two of them weren't running search ads. Molly Maid's ads ran under 103 different advertiser names, most of them franchise owners and their agencies. Thumbtack and Angi both hit the limit of how many ads the Transparency Center would list. A local independent walks into the same auction as all of them.
Google's auction insights report shows who you actually meet. Each rival it spotted gets six figures. The three a cleaning owner reads first: overlap rate, the share of your impressions where their ad appeared too; position above rate, how often theirs sat higher; and outranking share, how often yours beat theirs or showed while theirs didn't. The others are impression share and the two top-of-page rates. It needs some volume: at an impression share under 10%, it shows nothing.
Run it by ad group, not just for the whole account. The office cleaning auction and the house cleaning auction usually have different crowds. National marketplaces tend to sit on the broad searches with big budgets. They're much thinner on a service plus a suburb, which is where a local company can win for less.
Here's the part most owners never do: leave. If a marketplace or a franchise keeps pushing an auction up and your cost to win a customer there climbs past what that customer is worth, stop paying for it. Chapter 3's example ceiling was about $611 to win a recurring house cleaning customer. When a service's ad spend per booked customer passes your number for three or four weeks, cut that ad group's bid to where it pays, or pause it and move the money to a service that books cheaper.
Walking away from an auction isn't losing. It's refusing to pay retail for a customer somebody else is happy to overpay for.
A Bid Raise Is Only Worth It if the Calls Behind It Book Cleans
Cost per conversion can lie. A keyword with cheap conversions might be producing calls from people asking if you're hiring, people who want a price and hang up, or people asking if you sell supplies.
Chapter 1 set primary conversions to real quote requests, calls that pass Google's check, and bookings. Even so, some calls that count aren't customers. Before you raise a bid, listen to a sample of the calls that keyword or ad group produced. Your call recordings, or your call tracking tool if you use one, make this a ten-minute job.
Then judge each ad group on cost per booked clean, not cost per call. That's ad spend divided by the jobs that actually landed on the schedule. If your bookings flow back into your ad account, as Chapter 1 set up, it's a column. If not, it's a spreadsheet you update weekly from your scheduling software.
A keyword that costs $60 a call and books half of them is cheaper than one that costs $30 a call and books one in ten. Only the calls tell you which is which.
What a Booking Costs on Your Hand-Bid Keywords Becomes Broad Match's Target
The bids in Isolation are doing a second job: they're finding out what a customer really costs you. That number becomes the instruction for the next campaign.
Exploration, our other name, is the broad match campaign from Chapter 6. It runs on Maximize Conversions, a bid strategy where Google sets each bid, with a target cost per conversion as the leash. Since June 2026, Google's bid strategy menu calls that pairing "Target CPA." The campaign gets switched on when the hand-bid keywords are bringing in roughly 30 booked cleans and real quote requests a month. For most cleaning companies that's around the second or third month. Companies spending more get there sooner.
Take the target from Isolation's real results. Google's own recommendation works the same way: with conversion history, it suggests the average cost per conversion over the last 30 days, and it advises judging performance over at least 30 conversions. And Google's help page is blunt about aiming low: a target below reality can mean passing on clicks that would have booked.
Chapter 6 made the case for starting the target at what Isolation actually pays, not at a number you'd prefer. Two things help once it's running. First, move the target in small steps and give each change a couple of weeks of results before you judge it. Second, add Google's "Avg. target CPA" column. It shows what the strategy actually aimed for over the dates you pick, which is how you check that a change you made is the one Google is working to.
The same discipline applies across services. If you run office cleaning in Exploration later, it gets its own target from its own Isolation numbers, not the house cleaning average. The service that justified separate bids in the first place still needs a separate target.
Next: Cleaning Ads That Sound Like Your Five-Star Reviews
A bid gets you among the top ads. The ad itself decides whether anyone clicks. Chapter 8 covers writing cleaning company ads from the words your own customers use in their reviews, and how to see what competitors are saying before you write a line.
Earlier chapters: counting every call, text and booking, where cleaning customers search, what a recurring customer is worth, one page per service, competitor keywords and known searches before broad match. All sixteen are on the guide's home page.
In our Google Ads management for cleaning companies, every service gets its own bid and its own ceiling, and we walk away from auctions that cost more than the customer is worth.
