Places API Update, December 2020

By Chris Winfield

Welcome to our December 2020 update, featuring seamless international street searching in Great Britain, Canada and the USA.

When we launched early access for the Places API, we focused initially on Great Britain so that we could tune our new search engine technology. Building a search engine from scratch is a mammoth task, but it gives us the flexibility to create a location search API that's completely tailored to our customers' needs.

We included city-level data outside of Great Britain for the first time in the previous release, and in our December 2020 update we've boosted this coverage from city-level to street-level for autocomplete, geocoding, reverse geocoding and data cleansing in the Places API.

USA street autocomplete

We had great feedback for our international searching in the previous release, and we've enhanced our USA city, state and ZIP searching to now include full street-level coverage, with addresses available shortly (hopefully in our January release).

Although the coverage only currently includes streets, it's possible to find full addresses using our AI building detection.

US address searching, with location biasing towards New York

Combining location biasing and autocomplete finds results in a handful of keystrokes, and every result returns coordinates that can be used to plot places on a map or calculate distances.

Canadian street autocomplete

We can now also offer street and AI address searching in Canada. We've worked to standardise and cleanse the data and to handle both English and French searching, which is important for the bilingual provinces in Canada.

It's possible to restrict searching to just Canada, or to search cross-border with Great Britain and the USA (and other countries in future).

Canadian address autocomplete, with location biasing towards Toronto, ON

Automatic data updates

This isn't something that you'd necessarily notice as an end-user, but we now automatically build, test and deploy all of our data updates to the Places API.

When we first launched the Places API, we only had a handful of datasets which were easy to manually keep up to date, but as we expanded our international coverage we knew that we'd need to automatically update our datasets to ensure quick and consistent updates.

We've written hundreds of tests to ensure that data updates are reliable, and standardised and cleansed data to deliver the best possible data quality.