Research
Free booking links vs. paid Hotel Ads: incremental value and channel efficiency
Free booking links now deliver a third of Google metasearch bookings. At destination hotels, paid Hotel Ads are sometimes optional. At city center hotels they are not, and paid costs more per new booking than in-platform reporting suggests.
Of Google metasearch bookings come from free links
City center bookings after a big drop in paid visibility
Free-link bookings when paid visibility surged
Of added paid bookings were net new
The assumption
Since Google introduced free booking links, a hotel's direct rate can appear in price comparison without paying a cent. That has led many hoteliers to a tempting conclusion: if budgets get cut, travelers will just book through the free link instead.
We tested that assumption across our hospitality portfolio, using more than 18,000 Google metasearch bookings worth $11 million and more than 4 million paid Hotel Ads impressions. It's only half true.
What we found
Free links are essential
They consistently deliver about a third of Google metasearch bookings, quarter after quarter, and the guests are just as valuable.
At city center hotels, free links don't catch what paid drops
When these hotels lost significant paid visibility, metasearch bookings fell 17 to 22%. Free links recovered only part of the loss, and this held for branded and independent city center hotels alike, and even when the hotel was undercut. The data doesn't show why. Corporate and member rates, loyalty programs and business travel are possible factors, and OTAs bidding harder in those auctions may account for part of it. The size of the discount mattered: with a gap above about 20%, losing visibility cost about a quarter of bookings, while with smaller gaps it cost almost nothing.
At destination hotels, they largely do
Moderate visibility losses cost nothing measurable. Free links carry more of the load at destination hotels, which likely explains why. That held in both peak and off-season weeks. But these hotels were also rarely undercut, so their resilience may come from a competitive rate rather than from hotel type. A destination hotel that loses parity shouldn't assume its free links will hold up the same way.
Paid takes some of what free would have won
When paid visibility rose sharply at city center hotels, free-link bookings fell about 13% while total bookings rose about 9%. About three-quarters of the added paid bookings were new. The rest came from travelers who would have booked through the free link anyway.
Being undercut changes the math
At the hotels most often undercut in our portfolio, extra visibility added no bookings. Visibility losses still coincided with fewer bookings. Those losses were driven by auction rank, not budget, and came with lower average CPCs. That's consistent with OTAs bidding more aggressively as their conversion strengthens, alongside hotels lowering bids when return on ad spend fell below target. The safer course is to keep a defensive baseline presence while the rate gets fixed, rather than bidding aggressively for more visibility. For what a parity loss costs beyond metasearch, see When rate parity slips.
The top slot isn't everything
In weekly data, changes in top-position share didn't move free-link bookings in any consistent direction. What matters is whether the ad shows at all. The paid ad and the free link usually lead to the same official-site booking, so when the ad shows, some bookings simply shift from free to paid. Most, though, are new: travelers who likely wouldn't have found or chosen the free link on that search. We couldn't measure click-through by ad position, so this doesn't mean position is irrelevant, only that it didn't show up as a free-link effect week to week.
What hoteliers should do
- Know your hotel before you cut. Destination hotels may have room to trim paid visibility. City center hotels shouldn't expect free links to cover the gap.
- Calculate incremental cost of sale. Evaluate paid on net-new bookings rather than in-platform attribution, since part of what paid reports would have come through free links. Compare that incremental cost of sale with OTA commission when setting bid targets.
- When undercut, hold visibility and fix the rate. Buying more visibility won't win bookings while an OTA is cheaper, but going dark doesn't help either.
- Keep free listings accurate. Correct rates, availability and landing pages keep a third of metasearch bookings flowing at no media cost.
- Report free and paid together. Judging either on its own misreads what metasearch is delivering.
How Brightly helps
- One view of metasearch. We report paid and free booking links together, alongside other metasearch channels, so budget decisions reflect total bookings.
- Visibility tied to price. We track paid visibility and rate parity side by side, so spend goes where the rate can convert it.
- Weekly parity monitoring and alerts. When a hotel starts getting undercut, clients hear about it right away, with the affected dates, lead times and OTAs. See ParityWire.
About the data
Based on more than 18,000 Google metasearch bookings worth $11 million and more than 4 million paid Hotel Ads impressions across city center and destination hotels, January 2025 to September 2026. Each week was compared with the same week a year earlier at the same hotel, grouped by the change in paid impression share, and analyzed separately for city center and destination hotels. Destination Ads were excluded.
This is an observational analysis, not a controlled experiment. Visibility losses were driven by auction rank, which reflects both the hotel's bids and competitors' bids, so some losses may follow booking declines rather than cause them. City center hotels in our portfolio were more often undercut than destination hotels, so the two overlap. Within city center hotels, visibility losses cost bookings whether the hotel was competitively priced or undercut, which points to hotel type more than rate. Destination hotels were rarely undercut, so their behavior when undercut is untested. Read the sizes as directional.