Skip to content

Creating computed columns

Computed columns derive new transaction values from fields you already import. They can be used in tables, filters, commission rules, Metric Definitions, targets, and reports without requiring another source file or custom code.

For: Authorised Admin and Finance users

Prepare:

  • A clear name and business purpose
  • The imported fields used by the calculation
  • The required output type: Number, Date, or Text
  • Representative normal, empty, and boundary-value transactions
  • A manually calculated expected result
  1. Open Settings > Computed Columns.
  2. Create a computed column and enter its name and description.
  3. Choose the output type: Number, Date, or Text.
  4. Enter a formula. Reference transaction fields using braces, such as {amount}, {region}, or {transaction_date}.
  5. Review the live preview against representative transactions.
  6. Resolve any validation error before saving.
  7. Save the computed column.
  8. Verify its value on imported transactions and in every downstream metric, rule, target, filter, or report that will use it.
Goal Example formula
Convert a deal value {amount} / 1.2
Flag a large deal IF({amount} >= 1000, "High", "Low")
Calculate days to close DATEDIFF(day, {transaction_date}, {close_date})
Build a combined label CONCAT({client_name}, " - ", {region})
Use a fallback amount COALESCE({amount2}, {amount})

Supported functions include IF, ROUND, ABS, FLOOR, CEIL, DATEDIFF, CONCAT, LOWER, UPPER, COALESCE, and NULLIF, together with arithmetic and comparison operators.

Each value is stored on its transaction using the selected output type, so it can be sorted, filtered, and aggregated like an imported field. Perslace recalculates computed columns for new imports and source syncs so dependent views and calculations use the current value.

The computed column returns the expected typed value for representative transactions and is available in the intended filters, rules, metrics, targets, and reports.

  • A computed column does not change the original imported value.
  • Changes can affect every metric, scheme rule, target, filter, and report that uses the column.
  • Test empty values, zero, negative amounts, date boundaries, regional formats, and unexpected text where relevant.
  • Record the formula, test cases, expected results, reviewer, and approval before using it in production calculations.

Check field names, braces, quotation marks, function arguments, and the selected output type. Use only supported functions and operators.

Confirm that the referenced fields exist on the sample transactions and contain values. Use COALESCE when an approved fallback is required.

Confirm that the output type matches the formula. Do not return text such as "High" from a column configured as Number.

Existing transactions show an unexpected value

Section titled “Existing transactions show an unexpected value”

Compare the formula with the stored imported fields, then review the latest import or sync. Test the same record manually before changing the formula.

Last updated: August 2026