Missing Level 2 and Level 3 data led to doubled transaction costs at B2B marketplace

A routine reconciliation at a B2B marketplace uncovered a costly oversight in card data collection, resulting in doubled effective transaction rates. The issue was traced to missing Level 2 and Level 3 data, which are essential for qualifying for lower interchange fees, prompting a swift operational fix that significantly cut costs.

A reconciliation exercise at a B2B marketplace uncovered a costly blind spot in card processing: transactions that looked identical on the surface were being charged at markedly different effective rates because one set carried only basic payment data. The issue emerged when the finance team compared fees on commercial card volume and found a visible split between transactions that should have been priced the same under the acquirer contract.

The problem turned out to be the absence of Level 2 and Level 3 card data, the extra information that can help commercial and purchasing card transactions qualify for lower interchange. According to payment industry guides and processor documentation from PaymentCompanies, Corpay, Finix, Payengine and PayPal, these enhanced data sets typically include fields such as tax amount, purchase order number and, at the most detailed level, line-item information including SKU, quantity, unit price and commodity code. When those fields are missing, transactions can be downgraded to a higher commercial rate even if the card type, merchant category code and ticket size are otherwise unchanged.

In this case, the integration had been sending only Level 1 data: the card number, amount and date that were captured by the checkout flow. Because the payment form did not collect the additional fields, every eligible commercial transaction was effectively routed without the information needed to reach the lower interchange tier. The result was expensive: the batch with full Level 3 data settled at about 1.8% effective cost, while the transactions missing the extra fields came in close to 3.6%, enough to become a material expense on six-figure monthly volume.

The fix was straightforward but operationally important. Tax amount and purchase order number were made mandatory before authorisation, and the team pulled line-item detail from the order record already stored in the database. Once those checks were in place, the rate on new transactions fell within a single billing cycle. The lesson, as the original developer post makes clear, is that schema support alone is not enough; merchants need to verify that Level 2 and Level 3 fields are actually populated before the authorisation request goes out.