Every offsite city decision comes down to a table with two columns: the departure airport, and how many people fly from it. Build that table first and the shortlist follows from it in about an hour. Build it after someone senior has named a destination and you spend two weeks discovering what it costs to fly 35 people there, by which point the decision is too far along to change.
The five steps below take a roster to two or three priced finalists. Most of the hour is spreadsheet work.
Step 1: Turn your roster into airports (20 minutes)
Export the team roster with locations and map each person to the airport they would actually book, rather than to the metro area. The distinctions matter. Oakland against SFO changes the nonstop map, Newark against JFK changes it more, and the teammate “in Colorado” who lives in Colorado Springs is a different problem from one in Denver.
Then collapse to a count per airport. A 40-person company typically produces something like SFO ×9, JFK ×7, SEA ×5, AUS ×4, ORD ×4, BOS ×3, DEN ×2, and a tail of singles in Boise, Raleigh, Tucson and Berlin.
Pull two numbers out of that table: your core airports, the fewest airports covering about 80% of headcount, usually four to six, and your tail, meaning everyone else.
Step 2: Intersect the core (5 minutes)
Enter the core airports in Midway. It returns every destination with a nonstop flight from all of them, on a map and in a sortable table. Six departure cities is the limit, which fits the core-airport step closely; the limit is deliberate and probably covers 99% of real teams.
For a core of SFO, JFK, SEA and AUS, that is around 40 cities. Add Boston and Chicago and it is still around 33. That is what makes flight-first planning workable: even a six-airport core leaves dozens of viable destinations rather than three.
Leave the tail airports out of the search. A single colleague in Boise should not veto the shortlist, and the tail is handled in step 4.
Step 3: Apply the boring filters (15 minutes)
Only now do the preference filters enter, and they work best in this order.
Flight-time spread. Sort the table by flight time per origin. A city where one office flies 90 minutes and another flies six hours generates quiet resentment, while a spread under about three hours reads as fair. Geography is a poor guide here: Denver is over four hours from New York and under three from San Francisco, so a central city is not automatically a balanced one.
Season. October in Chicago is good and February in Chicago is a risk you are taking on behalf of 40 people. Filter to the month you are actually travelling, because route maps shift seasonally and some destinations only clear the nonstop bar for part of the year.
Ground cost and venue reality. Hotel-block rates and venue availability vary by a factor of two or three across the surviving cities. The ranked list of US offsite cities covers the trade-offs, including Austin’s October blackout. Two or three finalists should be left at the end of this step.
Step 4: Reconcile the tail (10 minutes)
Check each finalist against the tail airports. Some tails resolve for free, since the Raleigh colleague has a nonstop to nearly any eastern finalist. Some do not, and that is a decision rather than a failure: one person connecting through Denver to reach Austin is a known, bounded cost. Midway keeps flight data ready for the 600 busiest airports, 100 per region, so a search from a small tail airport is fetched on the first run and takes longer than the rest.
The rule worth writing down is to aim for 80–90% of headcount nonstop and choose who connects on purpose. If a finalist forces connections on the same office that connected last time, rotate the burden or drop the finalist. Fairness across a run of offsites matters more than perfection at any single one.
An international tail, the colleague in Berlin or São Paulo, usually means a connection whichever US city you pick. If the international contingent is more than a person or two, read the international offsite guide before locking a domestic city, because London or Mexico City sometimes beats any US option on total team travel time.
Step 5: Put a number on each finalist (10 minutes)
For each finalist, pull the cheapest direct round-trip per origin, which Midway shows in the table view, multiply by headcount per airport, and add a 15% buffer. Two or three finalists, one airfare number each, one slide. The cost estimation guide walks through the arithmetic and its failure modes.
That slide ends the debate. “Denver: $14,200. Nashville: $16,900. San Diego: $19,400, and the Boston five fly six hours.” Leadership can pick on cost, on fairness or on preference, and for the first time it is picking between fully priced options.
What this replaces
The manual version of this process is one browser tab per origin per candidate city, a spreadsheet nobody trusts, two weeks of elapsed time, and a decision made by fatigue. Flight search sites price one origin at a time, because they are built for a traveller rather than for a team.
An hour, start to finish: roster to airports, airports to intersection, intersection to two or three priced finalists. The cluster hub explains why this ordering works better than starting from a destination list, and the offsite pillar guide covers everything that happens after the city is chosen.
The tail is the part that stays a decision: which one or two people take a connection, and whether it falls on the same office as last time. Settle that before booking rather than after. Run your core airports through Midway to see the destinations all of them reach nonstop.