Why Meta, Microsoft, and Amazon Are Working Together to Disrupt Google Maps
How a free dataset with 59mm locations in the world impacts location-based services, mapping analytics, and physical businesses today
Last week, Meta, Microsoft, and Amazon jointly released a large open-source mapping dataset. This is an important moment in the mapping and geospatial data community. It opens up opportunities to disrupt the long-term dominance of Google Maps (an estimated ~$11bn revenue business).
What does this mean for app developers, data scientists, and physical businesses? Why are Meta, Microsoft, and Amazon doing this? We shared our analyses in this post below.
What is this dataset?
There are four types of data in this dataset:
Places: 59mm points of interest (POIs) globally, e.g. restaurants, airports, orthopedists, insurance agencies
Buildings: footprints and heights of built structures
Transportation: road networks and modes of transportation
Administrative boundaries: boundaries of cities, counties, provinces, countries
Why is this important?
Mapping data is essential for any developer looking to build location-based experiences in their apps, such as Uber (for pick-up/drop-off, navigation, and dispatch), Airbnb (for map-based search), and T-Mobile (for targeted offers based on locations). Google Maps offers one of the most popular APIs on the internet (ranked #8 this year), allowing developers to offer experiences such as searching for a nearby store or listing, validating a delivery address, and planning routes.
Up till now, Google owns the highest quality POI dataset. The challenge with POI dataset is that as businesses open and close, new buildings get constructed, digital maps need to be updated, and this is a laborious process most businesses cannot undertake.
Google has been the gatekeeper of key location information such as POI data, offering it to developers via an API or SDK. Google has placed many restrictions on how the data could be accessed, only allowing 20 results in a query, and banning storing or caching data.
This also means Google Maps’ location data is unsuitable for analytics use cases that need access to the entire dataset, forcing businesses to work with alternatives (e.g. Foursquare) that charge hundreds of thousands of dollars or more for updated POI datasets. While over 70% of Fortune 500 companies work with location analytics, many small to mid-sized businesses could not justify such investment.
Overture’s dataset truly democratizes access to POI data for any developer or business, allowing them to work with this important location dataset for free.
Why are Meta, Microsoft, and Amazon doing this?
While mapping isn’t a part of the core business for Meta, Microsoft, or Amazon, they all need maps.
Amazon needs accurate pick-up/drop-off locations and routes for efficient delivery.
Meta needs location data for business listings and geotagging, and on top of that, many young people are using Instagram and Tiktok as location discovery platforms, allowing Meta to potentially compete directly with Google Maps.
Microsoft operates its own Bing Maps.
In order to access high-quality mapping data, these companies either need to pay Google Maps of the world huge fees or spend significant engineering resources towards maintaining up-to-date maps.
While these companies have contributed to OpenStreetMap, a longstanding open-source mapping previously, this new initiative allows the trio to have the best of both worlds. They can leverage the existing data from OpenStreetMaps to kick-start the project while having more control over the data quality and direction the project is headed to. Their next goal of allowing users and delivery drivers to get from door to door easily shows their ambition to go beyond OpenStreetMap’s offerings to create more tangible business value for the three members as well as the broader open-source community.
Google has established a significant headstart when it comes to maps, through its series of acquisitions and over a decade of data collection. It was unthinkable to dethrone Google in mapping until Overture was released. The contribution of data from Meta (business listings), Microsoft (Bing Maps), and other ecosystem partners helped with the cold-start problem in mapping (i.e. people won’t use bad maps). Overture’s director Marc Prioleau also credited the use of satellite imagery, AI/ML for data extraction for making this possible.
What’s next?
Whether Overture succeeds depends on the adoption of developers and businesses. Marc Prioleau’s hope is that Overture’s map services will be deployed to many apps across different categories, allowing it to receive feedback and improve the map data in a crowd-sourced way.
Today, for smaller app developers, the simplicity of plugging into an API might still outweigh the benefit of using Overture’s data download services, but as Overture’s data quality improves and developer-facing services mature, it may become a strong competitor to Google Maps API.
Overture’s dataset also unlocks many use cases previously cost-prohibitive for small to mid-market physical world companies. Whether it’s planning to enter a new market, selecting a retail site, or analyzing competitive landscape, identifying potential sales targets, a high-quality POI dataset allows businesses to be a lot more strategic and confident in understanding the physical environment they operate in.
This is a great summary, and actually something I’m professionally pretty excited about. Google maps (and the associated API endpoints) are great but they come with a lot of limitations, particularly on the data storage side as you mentioned. This poses a lot of challenges for data-driven work. A more rigorous alternative to OSM is really encouraging!
While the Overture initiative is a fantastic development, categorizing it as a disruptor or threat to Google Maps is inaccurate.
First of all "Google Maps" is a consumer APPLICATION that is available for web and mobile consumers to search for, explore, evaluate and navigate to places. Overture has no aspirations of providing that type of application or service.
If you meant to write that Overture is a potential disruptor of Google's Geospatial Platform Services, that is also incorrect as Overture is a data set - not a platform that provides developers with APIs to access geospatial functions. If a developer wanted to use Overture data, they would need to replicate the huge investment in the highly scalable, feature rich, geospatial platform that Google has built over the past 18 years. I've been with a number of companies who have done this and it's no easy task to compete with Google's platform.
What Overture is a threat to (or a disruptor of) are the commercial map providers (HERE, TomTom, Zenrin, MapMyIndia) who sell their map data to application and geospatial platform companies. But even with that in mind, Overture today is far from providing the breadth of content that these data companies offer (attributes for routing, speed data, imagery, etc).
Overture has made a great start and that gap will close over time. But for now, Overture represents a great source of data for companies who have built (or want to build) their own geospatial platform solely to render great looking, detailed maps and to search for points of interest.