A brand-new hotel account knows nothing. It has never seen a booking come from an ad, never seen a guest pick a suite over a standard king, never seen a family book four nights for a wedding. Ask Google's automated bidding to optimize that account on day one and it will do its best with no information, which mostly means spending the budget and hoping.
So the first months run on bids you set yourself, by hand, from numbers you can defend. Then, once the account has enough real bookings with real dollar values behind them, the bidding moves to Google, and it moves to the one goal that fits a hotel: not how many bookings, but how much they're worth.
A Hand-Set Bid Is the Only Bid a Hotel Account With No Bookings Can Trust
Manual CPC bidding lets you set a maximum price per click for each ad group. Google's help page puts it simply: you pay only when someone is interested enough to click, up to the ceiling you set. In Isolation, our name for the first non-brand campaign of exact match keywords, that ceiling is yours to choose for every theme: the pet group, the parking group, the balcony group.
Two numbers decide where to start. The first is the break-even price from Chapter 3: what a click can cost before a booking from it costs more than the booking site's commission. At the US average daily rate and an 18% commission, a two-night stay that books from 2.5% of clicks breaks even at about $1.46 a click. The second is what the market charges. In our example market, New Orleans, Keyword Planner's top-of-page ranges for the themes from Chapter 4 looked like this:
- Pet friendly: $0.88 to $2.97
- Parking: $0.97 to $3.27
- Balcony: $1.04 to $2.87
- Near the Superdome: $1.18 to $3.25
- Near the convention center: $1.50 to $5.30
Start each ad group below your break-even and inside the market's range. For most themes on that list, that means opening somewhere in the lower part of the range. The convention center group shows why the math matters: its range starts above the $1.46 break-even, so it only earns a place if your stays there are longer, your rates are higher or your booking rate beats 2.5%. Your own break-even, from your own rates, is the number that counts.
Change hand bids weekly, in modest steps, and let each change run long enough to show a result before you move it again. When Exploration later takes over the open-ended searches, Isolation keeps its manual bids on the themes you've proven. The two run side by side, and you can lean on whichever is booking better. The brand campaign from Chapter 2 runs on manual bids too. Searches for your own name usually book well, so their break-even is higher, and the bid there is about staying above the booking sites, which Chapter 15 covers in detail.
Below 75% Top-of-Page Share, a Hotel Hands Its Best Searches to Booking Sites and Rivals
Then comes the weekly look, ad group by ad group. Start with search top impression share: out of every chance your ad had to sit above the free results, the portion it actually took. On the searches that book, aim for 75% to 90%. Its stricter sibling, absolute top impression share, tracks the single first slot on the page; it's worth watching on your own name and your two or three best themes, and not worth chasing anywhere else. Next to them sit the reasons for every miss, split into share lost to rank and share lost to budget. When the miss is blamed on rank, look at the bid and at the ad and page behind it. When it's blamed on budget, the day's money ran dry while travelers were still searching. Last, read Quality Score and its three parts from Chapter 4, so you can tell a bidding problem from an ad or page problem before you spend more to fix the wrong one.
If an ad group that books is sitting at 50% top-of-page share and losing it to rank, raise the bid, or fix the ad and page first if Quality Score says that's where the problem is. If a group is at 90% and its bookings cost more than the commission, lower the bid; you're paying for position you don't need.
The auction insights report adds the other half of the picture: who you're up against. It lists the other advertisers in your auctions, how often their ads showed when yours did and how often they ranked above you. For a hotel that usually means booking sites and the chains nearby. Google doesn't show anyone whose impression share is under 10%, so the list is the serious competitors, not everyone. When one auction keeps getting more expensive and isn't booking more, walk away from it. There are other searches.
Counting Bookings Teaches Google That One Night Equals a Week
Here is the problem the hand-set phase is preparing to solve. Imagine ten bookings in a week, priced at the US average daily rate of $161.78: three one-night stays, three two-night stays, two three-night stays, one five-night stay and one week. Counted, they're ten bookings. Valued, they're about $4,370, and the single week-long stay is worth as much as seven of the one-night stays.
A bid strategy that maximizes conversions treats all ten as the same win. It will happily find more one-night stays if they're cheaper to get, because by the only number it was given, a one-night stay is as good as a week. That's why Chapter 1 sends every booking to Google with its value. It's also why the goal that fits a hotel is value, not count, once the account has data to work with.
At About 30 Bookings a Month, Google Can Bid on What Each Stay Is Worth
Exploration is where the bidding changes hands. It's the broad-match campaign we build from Isolation's proven keywords, and it gets its own bidding from day one. It launches once the account records about 30 quality bookings a month, which for most independent hotels is around months two and three and sooner for bigger ones. From its first day it bids on value.
The first setting is Maximize conversion value. Google describes it as the strategy that gets the most conversion value from the budget, using what you define as value: here, the booking total. One warning comes with it. Google designs this strategy to spend the full daily budget, so set a budget you'd be comfortable seeing spent every day, and remember from Chapter 3 that a single day can run up to twice the average.
Once the campaign has steady data, add a target return on ad spend. Google's label for this changed in June 2026: the old "Maximize conversion value with a Target ROAS" is now simply "Target ROAS," and you may see either name during the switch. Google requires at least 15 conversions in the past 30 days before a Search campaign can use it. This guide waits for about 30 quality bookings a month before adding it, so the target has something real underneath it.
Where to set it? Start from what the campaign has actually returned over the last month, not from what you wish it returned. Then check it against Chapter 3's benchmark: an 18% commission is the same as earning about $5.60 per advertising dollar, a return on ad spend of about 5.6x. A target well above what the campaign is already doing tends to choke its volume, in our experience, so raise it gradually. Keep one target per campaign; Google's target-bidding guidance warns that separate ad-group targets can restrict its bidding.
Events need one more piece of judgment. Hotel demand rises when people book, not when they stay. In the example market, searches for "mardi gras hotels" peaked in January, weeks before Mardi Gras day itself in February, and "sugar bowl hotels" peaked in December and January. Google offers seasonality adjustments for short spikes of one to seven days, and says they don't suit longer stretches or Travel campaigns. For a hotel, the better move is to raise the budget during the booking window before a big event, let value bidding find the higher-value stays, and bring it back down once the bookings are in.
Value Bidding Keeps Counting Canceled Stays as Money You Never Collect
There's a catch that belongs to hotels. Chapter 1 explained that Google's bidding only hears about a cancellation if it's retracted within seven days of the booking. With guests booking about a month ahead, many cancellations arrive later than that. Value bidding therefore learns from booked value, not stayed value, and a campaign that attracts lots of free-cancellation bookings can look better in Google Ads than it is.
The fix is to judge the campaign on stays. Once a month, pull the stays and revenue from your booking engine or property management system for guests who came from Google Ads, and compare that to the ad spend. If the return on stays is well below the return Google reports on bookings, lower the target or tighten the campaign. Google's number is what the bidding steers by. Your own is what pays the bills.
A Reservations Call Still Needs a Human Ear Before It Counts as a Win
Some guests call instead of booking online, and Chapter 1 counts those calls. Google can now judge calls from ads by their recordings or, failing that, by their length. Both are useful filters. Neither is a guarantee that the call was a booking.
Every week, listen to a handful of calls that came from ads. Note how many were real reservation inquiries, how many were existing guests asking about parking, and how many were wrong numbers. If too many counted calls aren't bookings, the bidding is learning from the wrong thing, and it's worth raising the minimum call length or changing which calls count as primary.
Next: A Hotel Ad Should Sound Like the Guests Who Reviewed the Stay
With bidding set to grow from hand bids into booking value, Chapter 8 turns to what the ad actually says, and where the best lines come from: the reviews guests have already written.
The series so far covers booking tracking, the order of searches, the commission math, ad groups that match, the negative list and exact match before the long tail. The hotel guide's home page lists all fifteen.
When we run a hotel's Google Ads, the bids stay in our hands until the bookings prove what each stay is worth. Running it yourself? Open your top ad group this week and compare its search top impression share to the 75% mark.
