How should the distance-from-location radius be determined for a listing-based app?

Question

By "listing-based app," I'm referring to event apps like Eventbrite and restaurant apps like Yelp.

Whereas some apps allow the user to adjust the radius being used, i.e. "I want to see all listings within [#] miles of the location I've selected," other apps, perhaps in the interest of simplicity, don't allow for this level of customization.

In these cases without customization, I assume a fixed radius is being used in all situations? I find this interesting, because a radius that makes sense in one market, e.g. a market with a lot of sprawl like Los Angeles, might not make sense in another market, e.g. a market that's a little more condensed, like New York.

(I'm interested in reading any good articles on this stuff, as well!)

Answer 1

It depends how you sort your listings.

Radius based listing

If you are interested in let's say restaurants in # mile radius, you probably want to find the best ones within walking/driving distance. So you're basically saying: "Find me the best deal in this range. I'll not bother to go any further just to eat." And then you'll choose your preferred ones based on reviews and their offers. The key for sorting are reviews (or some other value).

Distance based listing

In this case reviews or other factors are usually not that relevant. The logic behinds this might be like "I want to go out tonight, what's happening around here?" and then you go and scan through the nearest venues until you find something that you like. The key for sorting is distance. So If there is a big conference or concert happening, you will more likely go the extra mile to get there. It would be limiting not to show you these events just because its couple of miles further.

Answer 2

Distance based listing is always handled in multiple stages:

1. 5 mile radius (5 mile/km is taken as rule of thumb)
2. Based on location (could be a broad category like a city or a much more specific one like area/suburb/post code)
3. A radius that provides relevant result (for example, if the software fails to find enough results in the above mentioned cases then it goes for 5-10 nearest results)

I want to state this beforehand that none of this can be found in a rule-book because the use-cases here are massive and cannot be generalized.

How to pick among the mentioned methods?

As applications are and should be developed keeping the user in mind, an educated guess should do the trick.

You'd want to use the 5-mile rule for emergency/asap services like hospitals, parking lots, food joints but there's more leeway in case of services like hotels, entertainment services, gas stations(I know, it's more or less an emergency service but I highly doubt that you'd get a good result set in a 5-mile radius).

It's pretty difficult customize the criteria depending upon location. You can't hope to get the results in a 5-mile radius in LA but in a 2-mile radius in NY. That's where the manual filtering options come in