Daily Consumer Spend by NAICS
Overview
Company Data for Academic Finance Research Covering Transactions aggregated by US Government NAICS Code. The tables in this dataset are also aggregated by Channel (Online, Offline, Unknown) and Region (West, Midwest, Northeast, or South)
| Data Information | Value |
|---|---|
| Refresh Cadence | Weekly |
| Historical Coverage | 2018-Present |
| Geographic Coverage | United States |
Schema
Create a Dewey Data login and navigate to the Table Structure view in the product for a description of each variable.
Daily Consumer Spend by NAICS
Key Concepts
This is a multi-table dataset
Navigate to the drop down menu below the
Download Samplebutton to review the various tables in the dataset. Review our resources for support accessing multi-table datasets via API .
Data Filters
| Fiter | Description |
|---|---|
| Date Range | No table reflects data associated with transactions received earlier than 1/1/2018. |
| Returns Excluded | Return transactions are not included; this includes all transactions with values < $0. |
| Non-U.S. ZIPS | Cards which do not have a valid U.S. ZIP code and associated transactions are not included. Non-U.S. ZIP codes include US territories and military ZIPs. |
| Transactions >$10K | Transactions greater than $10K are not included |
| Cardholders spending > $50K/month | Cardholders who have spent more than $50K in any calendar month of our data history are excluded. |
| High-Volatility Financial Institutions | Any financial institution (including within a processor) that has seen a change in cards or transactions per card that exceed volatility filters; any institution without transactions going back to the beginning of the dataset history. |
NAICS Codes
NAICS Industries are delineated by US Government-defined 3-digit NAICS codes. NAICS codes are manually mapped to individual brands for transactions tagged to a main brand. For other transactions, MCC codes have been mapped to NAICS codes and then the transactions are mapped to NAICS codes via their MCC code where available.
Panel Information
The number of cards in the database is growing over time, both due to macro trends in credit card usage and due to bank-specific customer acquisition trends. In order to normalize for this card growth.
This panel goes back to 1/1/2015 for USA1 data; it goes back to 1/1/2018 for USA2 and Combined USA1-USA2 data.
Matching Merchant Descriptions to Brands:
- Merchant names in the raw data are matched to brands.
- Additional data like MCC codes (industry codes), transaction sizes, and proprietary fields are used to refine this matching.
- Exceptions are tracked to ensure that subsidiary brands are included while non-revenue transactions are excluded.
Handling Complex Scenarios:
Special cases like "store-within-a-store" setups are carefully managed to credit revenue accurately to the appropriate brand. This ensures proper revenue attribution to the brand benefiting from the transaction.
Channel Identification:
- Transactions are categorized as either ONLINE or OFFLINE/UNKNOWN based on tagging processes.
- Catalogue and call center purchases are included in the ONLINE category.
- If not all online transactions for a brand can be confidently identified, they are flagged as OFFLINE/UNKNOWN.
If there’s insufficient confidence in the data, the channel is marked as NA.
Attributing Multi-Brand Transactions:
- A single transaction can relate to multiple brands. For example: "PayPal DoorDash Wendy’s" might tag Wendy’s as the Main brand, DoorDash as the Delivery brand, and PayPal as the Payment brand.
- Delivery brands’ revenue (e.g., fees and tips) is excluded from Main brand sales to avoid discrepancies caused by differing revenue capture rates.
Revenue Reporting for Delivery Brands:
Separate files (e.g., day_delivery and period_delivery) break out revenue related to Delivery brand transactions when a Main brand can be identified.
Updated about 8 hours ago