How we calculate drive times and rank neighborhoods

Drive-time zones (isochrones)

The colored polygon you see on each map is an isochrone. An isochrone is the set of points reachable from a given origin within a given travel time. We generate ours from the Mapbox Directions API, which models the actual road network including turn penalties, highway ramps, and one-way streets.

We generate two variants per employer:

• Off-peak. Modeled for a Sunday 10 AM arrival. Represents free-flow conditions with no congestion.
• Rush hour. Modeled for a Tuesday 8 AM arrival. Represents typical weekday morning conditions.

The delta between the two is frequently twenty to thirty percent at the perimeter of the polygon. That delta is itself a decision input, not an error.

Neighborhood-to-employer drive times

For every pair of neighborhood and employer in our coverage, we compute an approximate drive time at build time. The calculation uses great-circle distance plus a calibrated speed coefficient that accounts for the Ann Arbor road network (local streets at low distance, arterials in the middle band, limited-access highways at longer distance).

We report a range (low and high) rather than a single number. The low end approximates off-peak. The high end approximates weekday morning rush. A single-number estimate lies in this market. A range tells the truth.

When our approximation materially disagrees with Mapbox Directions API output for a specific pair, we use the Directions API result and log the correction.

Listing data

Listings come from a combination of MLS feed (where we have appropriate brokerage agreements) and direct-submit by licensed agents through our admin interface. Every listing carries a first-listed date, a last-verified date, and a status.

Listing counts and prices on programmatic pages refresh hourly via incremental static regeneration. A listing that goes pending disappears from the active count within an hour.

Editorial content

Our long-form guides and employer commentary are written by a local editorial team with first-hand experience driving the routes. Every guide carries a published date and a last-reviewed date. Content is reviewed semi-annually. Programmatic commute descriptions are regenerated from templates that are themselves reviewed quarterly.

We do not publish invented statistics. When we frame prices or volumes, we use qualitative ranges unless we have a specific citation we can stand behind.

What the numbers do not capture

Honest limitations of our model:

• Football Saturdays, commencement weekend, Art Fair, and major Main Street events rewrite the map for a day. We call those out on each employer page in commentary. We do not adjust the isochrone math for them.
• Construction affects specific corridors for months at a time. Our construction watchlist updates quarterly. Check local Michigan DOT alerts for the current week.
• Winter storms add five to fifteen minutes to any drive involving non-primary roads. Plan accordingly.
• Parking at the final destination is sometimes the slowest part of the commute. Our drive-time numbers stop at the employer address, not at a parking spot.

None of these limitations invalidate the numbers. They do mean a twenty-minute drive in our model can easily become twenty-five under real weekly variance. Budget accordingly.

Refresh cadence

• Listings: hourly (ISR revalidation)
• Isochrones: quarterly (January, April, July, October)
• Editorial guides: semi-annual review
• Destination and employer data: as needed, logged in git
• Neighborhood coordinates: as needed, reviewed annually

Corrections

If you see a number or a claim that is wrong, email corrections@commutin.com. Include the URL and what needs fixing. We respond within three business days and log the change with a visible date stamp.