About the visits product
The Tamoco Tamoco Smart Visitation Data uses the vast amount of actively collected location data to understand visits to Points Of Interest (POI). This product combines multiple data points to understand the number, and quality, of visits to a POI.
The visits product provides a visit score as well as extra metadata about the visit and the POI including time, visit duration, POI name, business category, amongst others. A full breakdown of this can be found in the visits structure section of this document.
The Tamoco visit product aims to measure user interactions with points of interest. Our algorithm ascertains if a visit to POI occurs using GPS location data from a mobile device.
This visits methodology contains two main operations: a clustering algorithm and a visit attribution algorithm.
Tamoco generates clusters where data points are proximal in time and space, as we only want to keep user location with a high density of datapoints triggered during a continuous period of time. The algorithm returns a cluster defined by 3 points: the first and the last locations that happened as well as the centroid of the cluster.
At this state, Tamoco is able to determine that a particular device stayed in a unique location for a certain period of time.
In order to transform those clusters into visits, clustered data moves to the second operation, the visit attribution.
Firstly, Tamoco discards clusters related to the user’s home or work location to protect the user’s privacy as well as ensuring that we only generate relevant visit insights.
Secondly, the remaining clusters are attributed to nearby POIs from the Tamoco dataset.
Our dataset is updated on a quarterly basis and improves with each new iteration, as Tamoco includes more accurate opening hours, building boundaries (US only at the moment) and comprehensive store brands and categories.
For a visit to be attributed, its cluster needs to be happening within the POI’s opening hours and its centroid must be inside POI’s polygon.
Lastly, Tamoco estimates the probability of a visit; verifying that a given user actually visited the POI attributed to its cluster. In order to estimate it, Tamoco considers:
- the distance between the cluster centroid and the POI center,
the POI quality score, that represents the level of confidence we have for a POI location and its information.
- the number of potential overlapping POIs nearby.
- the probability that a visit occurs based on the POI business category (for example on a Friday evening if we have a visit overlapping between a bar and a cafe, we should not assign the same weight for both).
Fields used when measuring visits
These are the fields that are used to calculate visits. Each term has a detailed description.
|The unique identifier of the visit
|The unique identifier of the device
|The first timestamp of the visit’s cluster
|The last timestamp of the visit’s cluster
|The duration of the visit
|The timezone of the device
|The local hour based on the first timestamp of the visit’s cluster
|The local date based on the first timestamp of the visit’s cluster
|A universal identifier for a POI
|The name of the POI
|The latitude coordinate of the POI center
|The longitude coordinate of the POI center
|The met area of the POI
|The street address of the POI
|The state of the POI
|The name of the brand related to the POI
|The corporation that owns the POI
|The primary stock symbol of the company
|The stock exchange the company is listed on
|The LEI associated to the POI
|The NAICS code related to the POI
|The name of the category related to the POI
|The name of the sub-category related to the POI
|The visit confidence level determined by Tamoco
You can use the following dashboard to understand POI availability by brand, category or region where Tamoco’s smart visitation dashboard is available.
Click here to access the dashboard.