A working playbook for founders, People Ops leads, and HR directors. Three group-capable Lisbon hotels, a day-by-day itinerary, a ballpark budget by tier, and the small things that decide whether the retreat lands or wobbles.
Why Lisbon, and why three nights
Direct flights from US East Coast, London, and most of Europe. A walkable city. A hotel scene quietly built for the distributed-team era. Three nights is the sweet spot, long enough to do real work, short enough that nobody’s partner sends a politely worded email about when they’re coming home.
Where to stay
Memmo Principe Real (15 to 30 people): boutique, full-floor buyouts, rooftop pool.
Tivoli Avenida Liberdade (30 to 60 people): Marriott STARS property, banquet-grade meeting space, rooftop SkyBar.
Four Seasons Hotel Ritz Lisbon (25 to 50 people, leadership tier): Virtuoso property, board-grade meeting rooms, 35-metre indoor pool.
The 3-night itinerary
Day 1, arrival: hotel check-in with a one-page Lisbon cheat sheet, welcome drinks on the rooftop, dinner at Time Out Market.
Day 2, the workshop day: breakfast at tables of four to six, two workshop blocks, working lunch at Prado, a free hour with nothing scheduled, group dinner at Pharmacia.
Day 3, the team-building day: two vans to Sintra, private guided tour of Quinta da Regaleira and Pena Palace, closing dinner at Belcanto or Cervejaria Ramiro.
Day 4, closing sync and departure: closing breakfast and a 90-minute team sync, late check-out.
The small things that make or break it
Build one full free hour into every day. Pre-arrange tipping. One person, one bed. Always book one more workshop room than you think you need. SIM cards on arrival, not before.
What this costs, ballpark
Per person, all-in, for a 3-night Lisbon retreat: Boutique tier $2,200 to $2,800, All-hands tier $2,400 to $3,200, Leadership tier $3,800 to $5,200.
I do not take a cut of your budget. I’m paid by hotel partner programmes when your team’s stay is confirmed, which means my incentive is to put your team somewhere genuinely worth staying.