← Back to App
Honest answers from the person who built this.
The honest answer: running a map service is surprisingly expensive. Every time you search for a midpoint or a coffee shop, the big map providers send me a small bill. As a solo developer — and not a giant corporation — these ads help keep the lights on and the servers running so I can keep this tool free for everyone. Think of each ad as a tiny “tip” that helps me pay for the next calculation.
It’s not a bug, it’s just the messy reality of maps. While the math tries to find a perfect “0-minute difference,” cities aren’t perfect grids. Sometimes the “fair” midpoint lands in the middle of a bridge or inside a building — the app snaps to the nearest accessible entrance. If one person has an 11-minute walk and the other has 13, it’s usually one-way streets, a tunnel, or a traffic light doing the math.
First of all, sorry! Because this is a nascent venture, I’m still ironing out the kinks. If something looks broken, try refreshing the page. If it stays broken, I’d love to hear about it, just click on the "Contact" link at the bottom. You can’t imagine how much it helps an independent developer when a user takes 30 seconds to report a glitch instead of just walking away.
I’ve curated these based on the most common reasons people meet up: a quick caffeine fix, a place to grind on a laptop, or a nice dinner. Under the hood, the app builds a “mixed bag” of the area’s best places first, then filters locally by your chosen category — so switching vibes is instant and free. If you feel a specific category is missing, let me know!
In some neighborhoods, the local favorites wear many hats — a cafe by day might be the best bistro by night! If you’re looking for more variety, try the “Expand Search” button to look a little further afield.
Absolutely not. I built this to solve a coordination headache, not to track your life. Your start and end points are used only to calculate the midpoint for that specific session. I have no interest in where you live or where you’re going — I just want you to get there at the same time as your friend.
Yes! Once a midpoint is calculated, a “Share this location” button appears above the map. You can copy the link, send it via WhatsApp, email it, or use your phone’s native share sheet. Your friend will land on a page with the pin already placed, so there’s no need for them to re-enter any addresses.
Yes — anywhere Google Maps has road and transit data, halfway.guru should work. I’ve tested it in a handful of cities across North America and Europe, but the algorithm is global. If it behaves oddly in a specific city, it’s most likely a gap in Google’s local transit coverage rather than a problem with the midpoint math.
That’s exactly what the travel mode selector is for. Each person sets their own mode — car, walking, or transit — and the algorithm calculates a midpoint that balances both travel times independently. A driver and a subway rider will get a meeting point that’s fair for both of them, even though their routes look completely different on the map.
Yes. After a midpoint is calculated, tap the “Not good? Manually move midpoint” button to make the pin draggable. Drop it anywhere on the map — perhaps a specific street corner or a park you both know — and the place search will automatically refresh around the new location. It’s a great way to override the algorithm when you have local knowledge it doesn’t.
Cities are not grids. Highways, rivers, one-way streets, and pedestrian-only zones can all push the “fair” point somewhere that looks odd on the map but is genuinely equitable in terms of travel time. If the pin lands in the middle of a park or an industrial estate, use the manual drag to nudge it somewhere more practical — the place search will follow.
Right now, halfway.guru is designed for pairs. The equitable midpoint algorithm balances exactly two travel times against each other. If you’re coordinating a group of three or more, a practical workaround is to run two separate searches — e.g., Person A vs. Person B, then check where that midpoint sits relative to Person C — and find an overlap. True multi-person midpoints are on my roadmap.
The first search for a given midpoint fetches live data from Google’s Places API. After that, switching between categories (Coffee, Dinner, Drinks) is instant because the results are cached for your session — no extra API calls, no waiting. If you move the pin or enter new addresses, the cache clears and fresh data is fetched automatically.
A few things could be happening. The initial search covers a 1.5 km radius around the midpoint — if that area is mostly residential or commercial, specific categories like “Brewery” might simply not exist nearby. Try the “Expand Search” button to widen the net, or try the “Show something different” option to surface results the first pass might have skipped.