Discussions

Hi, I am analyzing the SafeGraph Spend Patterns dataset and noticed significant variations in store/branch coverage for some firms and brands over time.

Here are a few examples of such fluctuations (though there are many similar cases in the dataset):

1/ Stock Symbol: PAG (Brands: Penske Truck Rental, Penske Automotive Group)

Until March 2024, only 1 or 2 stores were covered in the dataset.
Starting in April 2024, coverage dramatically increased to over 300 stores.

2/ Stock Symbol: FRME (Brand: First Merchants Bank)

Around 30 stores were covered up to October 2019.
From November 2019 onward, coverage dropped significantly to fewer than 10 stores.

3/ Stock Symbol: THC (Brands: Tenet Healthcare, MedPost Urgent Care)

Fewer than 10 stores were covered until November 2022.
In December 2022, coverage suddenly increased to more than 40 stores.

Could you provide any insights into why such large shifts in coverage occur? Are there known reasons for changes in the dataset, such as methodology updates, data source modifications, or filtering criteria adjustments?

Thank you!