Questions? We have answers.

Search

  1. MindBridge: English (US)
  2. Get started
  3. Import files
  4. Data checklists

Data checklist: Company card analysis

  • Updated

Summary

Learn which fields are required to leverage the existing configuration and control points for company card analytics.

Data Checklist

Columns Description
Cardholder ID Each cardholder's unique identifier.
MCC Code or description of the "Merchant Category".
Charge Code The code associated with the Business Unit or general ledger account.
Transaction Date Date when the transaction occurred.
Posted Date Date when the transaction was posted to the ledger.
Amount

The monetary value associated with the entry.

Note: All amounts in an analysis must use the same currency.
Group Cardholder Day of the Week

Groups cardholders by activity for the given time period. This field is created by combining "Cardholder ID" with "Day of the Week".
Count of Days from the Transaction to Posted Dates

The number of days between the date the transaction was created and the date it was posted to the general ledger. 
Group MCC Day of the Week

Groups merchants by activity for the given time period. This field is created by combining "MCC" with "Day of the Week".
Group Cardholder Charge Code

Groups cardholders by activity related to the given charge code. This field is created by combining "Cardholder ID" with "Charge code".
Transaction Day of the Week

Day of the week (Monday through Sunday) a given transaction occurred. 
Posted Day of the Week

Day of the week (Monday through Sunday) a given transaction was posted to the general ledger. 
Note: Please do not provide or load PAN (Primary Account Number) or data subject to HIPPA or PCI DDS.

General data expectations

Data must be/contain...

  • In .xlsx, .csv, or another supported format (.pdfs are not supported)
  • Entry based
  • Densely populated (limited blank cells)
  • At least 200,000 entries (rows)
    • Larger data sets are preferred (1+ million rows preferred)
    • MindBridge can analyze smaller datasets (20,000 rows), but some algorithms will not be as effective. Leveraging multiple years of data could add to the data volume.
  • In a consistent format across columns 
  • A healthy variety of values, including:
    • Numerous values for any categorical columns to allow richer categorization and pivoting of visuals
    • Limited synthetic, fake or columnar data

Assumptions

  • There are no missing values for the required fields 
  • Each row represents one unique monetary flow (i.e., the details about one unique charge)
Note: If a row is missing a required field, that row of data will be excluded from the analysis.

Additional columns relevant to your business could be mapped in the analysis for filtering. These values will not be used in risk scoring. Examples of data that would be useful for analysis:

  • Cost Center
  • Department
  • Role/Position

Anything else on your mind? Chat with us or submit a request for further assistance.

Was this article helpful?