Card Customers

Description:
The average number of customers a business has in a day.

See the Merchant Transaction Signals overview page for details about data sources, high-level methodology, and timeliness of this attribute.

Child attributes (and data file structure):

Column Name	Data Type	Description	Example
`end_date`	string	The time index of the features. All features in the row assume this is the final date, inclusive, of the calculation.	2020-08-31
`card_customers__1m__start_date`	string	The date that the 1-month period begins (inclusive).	2020-08-01
`card_customers__1m__average_daily_count`	float		1200.11
`card_customers__3m__start_date`	string		2020-06-01
`card_customers__3m__average_daily_count`	float		1400.77
`card_customerss__12m__start_date`	string		2019-09-01
`card_customers__12m__average_daily_count`	float		3200.80

JSON Sample:

{
        "card_customers": [
    {
            "end_date": "2020-08-31"
        "1m": {
                "start_date": "2020-08-01"
            "average_daily_count": 10.20
        },
        "3m": {
                "start_date": "2020-06-01"
            "average_daily_count": 30.15,
        },
        "12m": {
                "start_date": "2020-09-01"
            "average_daily_count": 70.68,
        }
    }
  ]
}

Other notes and tips:

Enigma uses unique card counts per day, i.e., “daily unique cards,” to estimate customer visits. There are a few limitations with this approach:
If a customer splits a purchase across two different cards, this would show up as two distinct customers. This is because Enigma is estimating customers based on unique cards not unique cardholders.
The count may not be an integer because Enigma applies a projection factor to the aggregated panel counts to estimate the total counts of each merchant
Multiplying the average daily customer by 30 can give you a proxy for the number of customers a business has. Note: this method will overestimate the number of customers because it will not take into account repeat customers in that month.