Migrating a Zapier or Make automation to a new geocoding host
No-code automations built on a geocoding step need a different migration approach than custom code does. Here is how to handle that switch.
Migrating from MapQuest Geocoding
MapQuest Geocoding has been around long enough that plenty of integrations built against it predate a lot of the current wave of mapping APIs. Its key, generated through a MapQuest developer account, is sent as a query parameter, and its responses return a results array with a nested locations array, each carrying a latLng object and a broken-out address structure with fields like adminArea5 for city and postalCode for the zip or postal code.
Integrations this old sometimes carry a particular kind of technical debt: the code that calls MapQuest may have been written by someone no longer on the team, with comments that reference decisions nobody remembers making. That is not a reason to avoid migrating, but it is a reason to test thoroughly rather than assume the existing code is well understood by whoever is doing the migration now.
My Geocode's MapQuest compatibility host reproduces the results and locations array structure along with the latLng and address fields, changing only the copyright, terms, and privacy text. Full details are at /compatibility/mapquest/. In most cases, updating the host and the key is the whole change; the nested results[0].locations[0].latLng access pattern should keep returning data in the same shape.
A sensible order of operations for an older integration like this:
- Find every place in the codebase that references the MapQuest host or key, since older code sometimes hardcodes both rather than reading them from configuration
- Move those references into configuration if they are not already, regardless of which provider you land on
- Point a staging environment at the compatibility host and compare a sample of real production addresses against the old results before touching anything live
Authentication on the new side accepts an X-API-Key header, an Authorization: Bearer header, HTTP Basic auth, or a query parameter, so however the old MapQuest key was being sent, there is a matching option here.
Pricing is flat and does not require an account tier decision: 2,500 requests a day free with no key, 2,500 more free per key per day counted per network, then prepaid credit at €0.0001 per request or an Unlimited key at €50 a month. Every endpoint, including this compatibility host, costs the same amount, which removes a variable that older contracts sometimes carried in the form of volume-based tier jumps.
For a MapQuest integration that has quietly kept working for years without much attention, the goal of a migration like this is to keep it that quiet afterward too: same response shape, same code, a different host and key behind the scenes.