Two commutes, one address: how dual-earner households should search
Minimize the worse commute, not the average — and don't trust the geographic midpoint. A method for households with two workplaces.
Nearly every commute tool assumes one destination. Most buying households do not work like that: two earners, two workplaces, one address. The single-destination habit produces a specific, predictable mistake, and this piece is about avoiding it.
Minimize the maximum
The naive approach averages the two commutes and picks the area with the best mean. But a household's day-to-day happiness tracks the worst commute, not the average one. An address where one person drives 15 minutes and the other 45 averages the same as one where both drive 30 — and is worse in practice, every week, for the person doing the 45.
So the number to compare across candidate areas is the maximum of the two commutes. Get that under both people's tolerance and the average takes care of itself.
The midpoint fallacy
The other instinct is geographic: draw a line between the workplaces and look in the middle. Geometry is not connectivity. The midpoint of two workplaces can sit on the wrong side of a freeway gap, a river, or a mountain ridge from both of them, while an area farther from the midpoint sits on a corridor that serves both directly.
This is the same lesson as radius versus travel time, doubled: if a circle around one workplace misleads, the midpoint of two is a circle squared.
Direction is half the problem
With one commute you have one chance of fighting peak flow; with two you have two. Phoenix's peaks are directional — the inbound morning direction is often far worse than the outbound — so an address that puts one commute with the flow and one against it can beat a closer address that puts both against it. This never shows up in distance, only in time, and only in time measured at the hours you actually travel.
A worked example
Suppose one of you works at Intel Ocotillo in Chandler and the other at Banner – University Medical Center in central Phoenix. As of late August 2026, Intel had 61 homes for sale within a 30-minute drive and Banner had 107 — but your real search area is the overlap of the two sheds, which is far smaller than either list. (The full six-employer picture is in what it costs to live near Phoenix's biggest employers.)
Computing that overlap by hand is exactly the tedium that makes people give up and guess. On the map, add both workplaces as custom places, and every home draws a route to each with a minute count — judge each candidate by the worse of its two numbers, and the overlap simply appears as the homes where both numbers fit.
When no address clears both budgets
Sometimes the honest answer is that the two workplaces do not share a livable middle: every area busts one budget or the other. Finding that out from a map in an evening is cheap. Finding it out eighteen months into a mortgage is not. If the map says no, the conversation you need is about time budgets, remote days, or one of the jobs — not about neighborhoods.