FAQs - Veraset

How is the data sourced?

Anonymized mobile location data comes from raw GPS signals of mobile devices via partnerships with third-party apps, software development kits (SDKs) and data aggregators. This is merged with point of interest data sourced from providers like SafeGraph.

What is Veraset's geographic coverage?

Veraset’s visit data covers most of the US and includes nearly all Census Block Groups, though exact coverage can vary month to month. Broadly, Veraset has consistent coverage across 5,886,791 census block groups (CBGs), representing approximately ~70% of all block groups nationwide—including those without recorded populations. When looking only at CBGs with a recorded population, their coverage rises to roughly 90%.

How does Veraset ensure the privacy of their data?

Privacy is a top priority at Veraset. The company employs a range of measures to protect the privacy of the individuals represented in its data. This includes pseudonymizing and aggregating data to ensure that no personally identifiable information (PII) is included. Veraset also complies with all relevant data protection laws and regulations, and it has robust security measures in place to prevent unauthorized access to its data.

What about visit data to sensitive locations?

Sensiitive locations are excluded from these datasets. This may explain why you are unable to find NAICS codes such as 62 which pertains to healthcare.

Since August 2022, Veraset has been actively filtering its data for these locations. With recent actions by the FTC and other regulatory pressure, Veraset has implemented broader Sensitive Places filtering capabilities. To learn more, visit Veraset's NAICS Filtering resource. Note: locations like the food court within a hospital will also be filtered out.

See full definition of sensitive locations in Dewey Data Partner Terms.

How long can a single device be tracked in Veraset data and what causes its ID to change?

Devices can often be tracked for many months or even years, as long as users keep the same device and app permissions stay active. However, churn rates have been increasing (upwards of 15% month over month), making the monthly data refresh an important part of maintaining accuracy.

What is the horizontal accuracy of the visit data?

The visits clustering algorithm requires at least 4 decimal places (inside 10m) in the latitude and longitude and at least 5 decimal places (inside 1m) for the other.

How is a visit calculated in the Visits and Home/Work Visits datasets?

The Visits dataset has used two separate calculations to attribute a visit:

  • 2019 - June 1, 2024: A visit was counted when a device was pinged twice between 4 minutes.
  • June 1, 2024 - Present: A visit is counted when a single ping is registered.

The current calculation also measures the speed of the device when it is pinged to remove noise (car traffic, bike riders, scooters, etc.).

How does Veraset determine a "Home" and "Work" location?

They use a multi-week look-back period of 6 weeks, analyzing day-over-day and week-over-week observations. Home is determined where a device spends a majority of time during typical home hours (7 PM - 5 AM) and work is determined where a device spends a majority of time during typical working hours (7 AM - 6 PM).

Note if work and home locations overlap for these hours the individual will be flagged as someone who potentially "works from home". In addition, big shifts for a device like a move can be detected.

Does the Visits file include Home and Work visits, since there are also separate datasets?

The home and work visits are included in the master Visits dataset from 2019-2022. In 2023, the home and work visits were moved to separate products due to shifting privacy laws.

If the location for a visit is a designated home location, the Location_namewill = "home" (from 2019-2022).

Why is there a significant change in visits starting ~2022

Privacy requirements have changed significantly from 2019-2022, and the impact of those changes is reflected in the data. Specifically, iOS and Android devices implemented "opt-in" requirements and prompts for tracking through mobile apps.

What is a GEOHASH-5 and why do the Home/Work and Home/Work Visits and Location datasets use that as the lowest level of resolution?

The GEOHASH-5 is 5 characters long, and represents a relatively small geographical area on Earth, providing a moderate level of precision in location identification compared to shorter geohashes, but still encompassing a larger area than a very precise, longer geohash; essentially, it signifies a "medium zoom" level on a map when using geohashing to pinpoint a location.

The dimension for the GEOHASH-5 are 4.89km x 4.89km and is used in these datasets to protect the privacy of the individuals in the panel.

Does Veraset use a POI provider?

Yes, Veraset uses SafeGraph as their POI data provider. As a result, NAICS, Placekey, and SafeGraph ID will all be in the datasets (Home and Work datasets will not have them for privacy reasons). This means researchers on Dewey can easily join to SafeGraph datasets including spend.

How are visits attributed in locations with shared polygons? (e.g. - shopping malls)

For the most part, Veraset is able to differentiate the child location that was actually visited. In some instances where one of the polygons is a sensitive location (like the food court within a hospital, or a glasses store in a mall) the data will be removed causing some inconsistencies.

What year census does Veraset use in their data?

Veraset uses the 2010-2020 census data.

When did Veraset begin using data from the 2020 Census?

Veraset began using the the 2020 census starting in February 2025**.

Does Veraset do any normalization to their data?

Veraset does not do any normalization with their data. While there are many options for normalizing datasets, one approach is to use census data and create a population ratio to tie devices to # of people.

Does Veraset offer their raw movement data?

Yes, however this is not available through Dewey. Please reach out to [email protected] with this request and we will put you in touch with the right contacts at Veraset. This data is available globally.

How does Veraset’s data compare to other sources like Citydata.ai or Cuebiq?

Veraset has data from exclusive providers compared to other panels. They focus on high-quality, consent-based foreground pings and uses a rigorous visit attribution process. What makes them unique is their clustering algorithms, the historical coverage, as well as the ability to join with other datasets.

Why are there observations with negative dwell times

The negative dwell time issue has been fixed. The issue with negative dwell times related to a bug that surfaced when calculating extended visits; effectively, when a device pinged in an area for an extended period, it was subtracting the wrong series of pings resulting in a negative reading. This has been corrected in the latest release (July 2025).