A brief story about SafeGraph’s rapid evolution as a Data-as-a-Service company
Only seven years ago, SafeGraph opened its doors as one of the few data-only companies in existence, allowing us to position ourselves uniquely as a Data-as-a-Service (DaaS) provider. While our mission has always been to democratize access to clean, accurate, and highly granular data about physical places—also known as location-based data or geospatial data—at the time, we captured only a few million places across the U.S.
Fast-forward to the present day. SafeGraph’s Places database now includes rich attributes for over 80 million locations across more than 200+ countries and territories worldwide. While expanding the scope and depth of our data has been crucial to our growth, it has not come without obstacles. With each period of expansion, we have continuously reevaluated—and at times challenged—our own understanding of what qualifies as a “place” in order to meet evolving customer needs.
As we continue to progress on our mission, we believe it is important to share some of the insights we have gained along the way and highlight the learnings that have shaped how we define places today.
When we first entered the market, we filled a gap that desperately needed filling. Accessing good, clean, and accurate geospatial data in the U.S. often required hours of manual effort just to make datasets usable. Our initial focus was therefore on ensuring that our data could be used immediately upon delivery.
Once customers experienced how refreshing it was to work with clean, accurate, and regularly updated datasets, they began asking how we could expand our offering to support more use cases. Each question became a catalyst for exploring new ways to scale our dataset—both by adding more physical places and by appending richer metadata to existing records.
For example, we initially focused on places where people spent money, such as retailers and restaurants, within the U.S. We later expanded to places where people spend time but not necessarily money, including parks, schools, offices, warehouses, and manufacturing facilities. We then incorporated small-footprint POIs like electric vehicle charging stations and ATMs. Finally, we began appending new attributes—such as “store ID” (to support integration with transaction data) and “category_tags” (to help isolate places using narrower text descriptors)—and expanded our coverage globally.
It was important for us to move fast and position SafeGraph as the trusted source of comprehensive and accurate geospatial data that captures the full complexity of real world places. We also knew that if we couldn’t provide our customers with the data they needed, they would look elsewhere. As a result, we became relentless in expanding and refining our datasets to ensure we consistently met our customers’ expectations.
This steady expansion mirrors the data transformation in location intelligence, where richer context increasingly drives better decision-making.
Scaling coverage is only one side of the equation. As the dataset grows, it becomes exponentially more complex and must be balanced by rigorous quality assurance to uphold our promise as a high-quality data provider. This reality led us to ask an important question: “What is a real place?”
Because businesses open and close every day, we constantly verify that places in our dataset are truly open and operating in the real world. Adding new data sources also introduces lower-quality records that may not represent legitimate physical places, such as online-only stores or event names mistakenly treated as venues.
The long-term precision of our datasets and our customers’ trust in our ability to deliver high quality data go hand in hand. We cannot have one without the other. We also consider the potential impact, or “cost”, that false positives can create for businesses that rely on our data as a source of truth. For us, those kinds of errors aren’t acceptable. While no dataset can be perfectly accurate at all times, our team consistently goes above and beyond to minimize errors and deliver dependable geospatial data.
If you’ve ever wondered how we build the SafeGraph Places dataset, here is a brief overview of the process—one of many examples of geospatial technology in practice:
As part of this quality assurance process, we use several layers of geocoding to ensure the geographical coordinates (latitude and longitude) are based on a POI’s rooftop, standardize POI names and addresses, and apply machine learning to infer the most appropriate category description based on the associated metadata.
Of course, there is a significant amount of work that happens between each of these steps to ensure the final product is exactly what our customers want and need. But if there is one takeaway, it is this: we make it a priority to go above and beyond to deliver the most accurate and comprehensive geospatial data available.
So, what is a “place”? As this journey shows, the answer has evolved—and will continue to evolve—as customer expectations change. What once seemed like a simple geospatial concept has grown into something far more expansive, reflecting both current needs and future possibilities.
As the need for accurate location intelligence accelerates, our definition of a place will continue to adapt. What will not change is our commitment to accuracy, completeness, and trust.
What’s your definition of a place? Share your thoughts with us.
Location intelligence is the use of highly accurate and comprehensive geospatial data about physicals to drive business insights and decision making.
The real estate data revolution is driven by richer, more accurate geospatial datasets that enable better market analysis, site evaluation, and operational decisions at scale.
Geospatial data provides the spatial foundation for location intelligence by defining where places exist and how they relate to one another. It enables analysis of movement,proximity and patterns in the physical world.
Data quality matters because even small inaccuracies or outdated records can distort analysis, leading to flawed insights and poor decision-making at scale.
POIs provide real-world context by defining what exists at a location, how it functions, and how it relates to surrounding places.
As use cases expand and real estate analytics trends advance, the definition of a place must adapt to ensure geospatial data remains accurate, relevant, and decision ready.
