Accurat's solutions power your teams to make strategic and tactical decisions that improve decision making and drive ROI & revenues.
To provide our insights we use different sources of data. Let's break down all there is to know about the data we use.
We use two different sources for data:
- Data coming from our self-built SDK that collects data through partnering up with different smartphone applications.
- Data originating from so called data brokers.
- Exact location (GPS signal data), such as latitude, longitude, altitude, speed, heading and accuracy of the location (both horizontal and vertical)
- “Mobile advertising identifier" such as Apple IDFA or IDFV for iOS devices and Android Advertising ID (AAID) for Android devices
- Name of the connected WiFi network (if any) and mobile phone network (carrier) of the User
- Information about the User’s device - brand, model and version of the OS
- Time in date, hour, second, time zone
Accurat has created an SDK (software development kit) that minimizes battery consumption (below 1%) yet captures sufficient data - closely following OS-specific restrictions & possibilities. Data collection (in a simplified summary) works as follows:
- The consent flow requests proper permission (GDPR & location permission)
- The tracking is activated to start collecting coordinates every 2 mins or whenever the user makes a stop (a visit) - even without network connection (4G/wifi)
- We collect fused location data: the best possible coordinate the operating system can provide, whether collected via cell tower, gps antenna, wifi or bluetooth.
- To optimize for low battery consumption, we use additional methods to stop collecting data when stationary but will restart collection when the users move again.
- All coordinates are stored on the device - even when there is no network connection.
Locally stored data is sent to our cloud environment over an SSL-encrypted connection every 3 to 24 hours. - We limit network connections (4g) as 1) the operating system limits usage when in background and 2) it consumes less energy
The local bucket is emptied and new coordinates can be stored locally again.
Accurat performs two distinct types of analysis.
Historical data is used to identify: the visitor’s origin, the visitor’s profile, frequency and trajectory. It is also used to identify the visitor’s social class.
Real-Time data is used to count the number of people at a certain POI (point of interest). This data has a delay of 48hours, given the calculations that have to be done on the incoming data.
Visits are reported on a daily basis, most metrics such as share of visits, penetration, etc. are added on a periodic basis (no penetration per day available).
Data provided by Accurat has an accuracy of approximately 95%. Reporting is done for the entire Belgian and Dutch population, not only for users as tracked by our SDK. Therefore, we developed an algorithm that extrapolates the data for the entire population in a representative manner.
Our data has beenbenchmarked with multiple campaigns with the results of the CIM model. We have a very similar outcome with a small difference in the margin of error. Therefore, we can state that our data is reliable as depicted in the graphs below:
Data is collected 24 hours a day, 7 days on 7 for the Belgian and Dutch territory. Approximately 11 million coördinates are collected each day in our database generated by SDK users.
Accurat’s approach is unique since we work with data that is not provided by our clients in contrast with many of our competitors. Since we have our own data our insights are plug & play. We have access to a bigger panel which enables us to provide also provide solutions for small stores. Furthermore, results are updated on a daily basis instead of monthly or ad hoc reports.
Data is visualized in a variety of dashboards. Accurat delivers access to personalized dashboards through a secure login on our website. Users can select shops separately or get a global overview of a brands performance. Depending on the chosen options users can access more or less dashboards. Detailed descriptions and all insights can be exported in different formats (png, pdf, CSV).
We clean data on row level (coordinate level):
Remove duplicates (sometimes the SDK sends the same coordinate twice to our API).
Remove coordinates with low accuracy (<50 meters)
Remove invalid coordinates (eg. latitude and longitude equal to 0)