Picture a popular sports network that delivers content via an app to phones, tablets, and smart TVs. Each of the network’s millions of users has opted into the service, enabling it to collect granular data on their viewing preferences. But the app-based delivery means the network also has access to location data for every user. That is, for all those users, the network not only has propensity data on content selections, they also have real-time geospatial data on where that content was consumed.
Now consider that most of those app viewing experiences took place on a mobile device. The mobile phone company that provides a continuous cellular connection to the phone also has voluminous, granular data on where that consumer has been, and when.
Today, geospatial data like this is ballooning, as are its value and the range of applicable use cases for it across virtually every industry. From Retail to CPG to FinServ to Health Sciences, geospatial data can be critical to understanding footfall trends, optimizing supply chains and location-specific offerings, and creating and maintaining customer DNA.
How does geospatial data work?
In the simplest terms, geospatial data makes use of an abstract grid system to provide a consistent analytical unit for location-based event analysis. The grid system tiles a simple, uniform shape (e.x., a square, triangle, or hexagon) continuously around the globe, and events are then bucketed into single units of the grid. To enable efficient analysis, unit shapes must be compact and — for minimal distortion and quantization error — feature the same distance between the centerpoint of each shape and those of its neighbors. Grid systems avoid the subjective inconsistency of traditional geographic zones and markers in favor of a globally consistent mathematical space.
While there are a number of geospatial grid systems available today from Google and others, perhaps the leading solution is Uber’s H3, which uses hexagons to “combine the benefits of a hexagonal global grid system with a hierarchical indexing system.” By nesting grids of hexagons, H3 provides 16 levels of resolution, from continent scale down to a square meter, in which to bucket events. That enables extremely precise location of event data.
Who has geospatial data, and how do you use it?
When considering the potential of geospatial data, it helps to think about where it’s coming from. As mentioned, mobile phone providers have lots of geospatial data, as do online apps. Among others, airlines, car rental companies, hotels, retail media networks, and of course ride-hailing and delivery companies like Uber itself all collect geospatial data.
Partnering with these kinds of geospatial data owners for data collaboration can yield valuable insights. And that cuts across industries: advertisers can now compare campaign data to transaction location; retailers can deepen understanding of visitor trends and figure out the best locations for new stores; restaurants can optimize ordering with location metrics; CPG companies can get hard-to-find insights into customer purchasing behavior; enterprises can improve supply chain management with regional consumption data. The list goes on and on.
But wait: geospatial data is sensitive, first-person, event-level data. It’s a record in time of where consumers have been. In recent years, increasing privacy regulation, including GDPR in Europe and CCPA in California, means that owners of this kind of data cannot simply share it. Geospatial data must be handled in a privacy-preserving way — and that requires a data clean room.
You’ll need a capable data clean room
A data clean room is a software-defined, privacy-preserving environment in which companies can securely share and analyze datasets. Because they offer data collaboration that doesn’t expose proprietary data and IP — and doesn’t run afoul of consumer privacy regulations — data clean rooms vastly expand access to data and to data partners.
There are many clean rooms on the market, but most are not equipped to handle geospatial data. In some cases, that’s because they can’t work with the volume and complexity that comes with raw, event-level data. Other clean rooms have restrictive schema that only work with individuals and don’t do well with real-time geospatial data that’s everywhere and all the time.
Habu data clean room software, on the other hand, is ideal for powering geospatial analysis, whether with data from H3 or another grid system. A business-friendly, out-of-the-box solution with a flexible and open schema, Habu data clean room enables data science teams to jump into geospatial analytics and deliver aggregations of location data just like any other data aggregation. That opens up all kinds of opportunities for brands to deepen insights — and for owners to safely monetize data.
Data science teams looking to implement geospatial analytics should note that Databricks recently announced the release of 28 built-in H3 expressions for efficient geospatial processing, and Habu has partnered with Databricks to offer a data clean room solution that powers multi-cloud, multi-party collaboration for advanced analytic, AI, and ML use cases on the Databricks Lakehouse.
Elevate your data collaboration goals
As the example of geospatial data illustrates, Habu data clean room software enables organizations to derive valuable insights from a much broader universe of data, uncovering insights that fuel better business decisions via data collaboration that’s smart, safe, scalable, and simple. With flexible, multi-cloud deployment, high levels of automation, and an intuitive interface, Habu data clean room software empowers companies to accelerate transformation and seize market advantage.