Most venue lookalikes are decided before a single ad runs, and they are decided in the seed list. Someone picks the page-follower audience or exports the whole email list and asks the platform to find more people like that. It obliges — more followers, more browsers, more people who liked a sunset photo two seasons ago and have never spent a dollar at the venue.
A lookalike is a copy machine. It reproduces the pattern you hand it. Hand it admirers and it brings back admirers.
The direct answer: seed from the people who paid
Build the seed from your bookers, ranked by lifetime value, and send the platform the value along with the person. The list of guests who prepaid a daybed, a cabana, or a table — and what each was worth across every visit — describes your next customer far better than any follower count.
The venues that get this right are not doing anything clever with targeting. They are doing something ordinary with data: the booking system knows who paid, how much, how often, and what they added on, and that goes to the ad platforms instead of sitting in a reporting tab. On Clubtech, bookings fire to Meta, Google, TikTok, and GA4 in real time with revenue values posted back, and lookalikes start from the highest-LTV bookers.
What makes a good seed
Four properties, and size is the last of them.
Recency. A booker from three seasons ago describes a venue that no longer exists — different programming, price ladder, market mix. Use a rolling window tied to your season and let old buyers age out as new ones replace them.
Value. Rank by what the guest has spent across every booking, not by whether a booking happened. Someone who takes the top package twice a season is a different pattern from someone who took a discounted sunbed once; merging them teaches an average that describes neither.
Coherence. A seed of table buyers plus bargain daybed guests is two audiences pretending to be one. Split them and run them separately before deciding either has failed.
Size, last. Each ad platform sets its own minimum and documents it — look it up there, not in an article. When your top spenders fall short, widen the value band a step. Padding the list with non-buyers is how a booker seed quietly turns back into a follower seed.
Why a revenue-valued feed beats a raw event feed
A raw event feed tells the platform that a booking happened. A revenue-valued feed tells it what the booking was worth. With the first, every conversion is identical, so the optimizer will happily find a hundred of your cheapest bookings. With the second, it separates a bed-only reservation from a party package and bids accordingly.
The same data does both jobs: the value that ranks your seed list is the value the platform bids against once the campaign is live. Post bookings back without a value and you can still build a lookalike, but you are optimizing blind on the other side of it. Our guide to the Meta Conversions API for venue bookings covers what to send and why pixel-only setups leak.
How to segment the booker list
One seed of "everyone who booked" beats followers and still underperforms. Four cuts worth running separately:
Repeat bookers. The smallest list and the strongest pattern — guests who came back without being chased. Start here.
High average-order-value groups. The party that books the top tier for eight is a different buyer from the frequent solo lounger, even at matching annual spend.
Add-on buyers. Guests who stacked bottles, cakes, or transfers onto the reservation have told you they buy up. FINNS grew bottle add-ons 67% year over year and average order value 23% — too much signal to leave in a report.
Origin market. The countries your guests fly from are a sharper prospecting list than the country your venue sits in. Clubtech ships 20+ reports covering origin markets and lead time by zone.
One source most venues ignore: the rest of the party. The booker invites their group and each guest supplies their own details, so the marketable audience grows with every reservation, not just with every booker. FINNS has 240,251 guest profiles captured on the platform.
How to judge a lookalike honestly
Against a broad campaign, not against nothing. Run both with the same creative, offer, and value-optimized conversion event, and judge them on booking value per dollar spent, not on conversion counts.
Give each side long enough to leave the learning phase, and check audience overlap first: a lookalike that heavily overlaps your retargeting pool charges you to re-reach guests who were already coming. Then read the two numbers that say whether the mix improved: average value by variant and repeat-customer share. A lookalike that lifts volume while the mix slides down-tier has found more of the wrong guest.
Often the honest outcome is that broad targeting with a clean revenue feed matches the lookalike. Keep the simpler setup and put the effort into the feed that made broad work.
Refresh cadence
Refresh on the venue's clock, not the calendar's. Rebuild after each peak block and again before pre-season bookings open, so the audience that fills your season comes from the guests who filled the last.
Two failure modes. Rebuilding weekly resets learning for no gain — the guest pattern does not move that fast. Never rebuilding is worse: a seed frozen at last season's price ladder prospects for a venue you have since repriced. Keep the definition stable and let the membership roll.
The seed stays yours only if the data does. Because the booking flow runs white-label on your own domain, the list and the values belong to the venue, not to a marketplace. The wider campaign architecture this feeds is laid out in our beach club marketing strategy guide.
Questions operators ask
What is a lookalike audience for a venue?
A lookalike audience is a group the ad platform assembles by finding people who resemble a list you supply. For a venue, the list worth supplying is your bookers ranked by lifetime value. The platform models the pattern in that seed and prospects for more of it, so the seed sets the result.
Should I seed a lookalike from my Instagram followers?
No. Liking your content is a much weaker signal than paying you. A follower seed teaches the platform to find more browsers, so prospecting budget goes to an audience that engages and does not book. Seed from paying bookers and keep followers for what they are good at, which is proof of atmosphere.
How big does the seed list need to be?
Each ad platform sets and publishes its own minimum, so check the current figure in its documentation rather than in an article. The rule that matters is what to do when your best guests fall short: widen the value band to take in the next tier of spenders, rather than padding the list with people who never booked.
How often should I refresh the seed list?
After each peak trading block, and again before pre-season bookings open, which keeps the audience matched to your current programming and price ladder. Avoid weekly rebuilds, which reset learning without changing the guest pattern, and avoid never refreshing, which leaves you prospecting for a venue you have already repriced.
Do lookalikes beat broad targeting?
Not always, so test rather than assume. Broad targeting with real booking values posted back often performs as well, because the platform optimizes against the same revenue signal either way. Run both with identical creative and offer, judge on booked revenue per dollar, and keep whichever wins.
How much does it cost?
Commercials follow discovery. Every venue has a different mix of inventory, zones, payments, integrations, teams, and rollout work, so the proposal comes after we have seen the operation. Book a discovery call. We map the operation and come back with a commercial proposal built around the real scope.
Bring a season of booking data to the call. We will show the booking events posting back to Meta, Google, TikTok, and GA4 with revenue values attached, and the reports that surface your highest-spending guests. Book a 30-minute demo
