A Short Pay is when a customer pays less than the full invoice amount, either deliberately due to a deduction or dispute, or in error. Short pays are the operational trigger for most deduction and dispute workflows in B2B AR, and the largest source of variance between invoiced revenue and cash collected.
Short Pays are the financial expression of customer-supplier friction. When a customer pays less than invoiced, they are taking a position: either claiming an agreed deduction, raising a dispute, or in some cases making an error. For sellers, every short pay represents potentially recoverable revenue, and the aggregate impact is meaningful: typical B2B short pay rates of 5 to 15 percent of invoices translate to materially less cash than the income statement suggests. In CPG selling to large retailers, the rate can hit 20 to 30 percent of gross invoices.
Short Pays typically fall into five operational categories.
The split between authorised and unauthorised is the most important operational distinction. Authorised short pays are routine; unauthorised ones drive the bulk of deduction management workload.
Short pays break the simple invoice-to-payment matching that drives straight-through processing. When a customer pays 95 percent of an invoice, the cash application system cannot simply post the payment because:
For this reason, short pay rate is one of the largest single drivers of cash application straight-through processing rates. Operations with high short pay rates rarely exceed 75 percent STP without AI assistance.
Mistake 1: Auto-accepting all short pays. Teams without investigation capacity default to closing residuals as accepted deductions, foregoing 30 to 50 percent of recoverable value.
Mistake 2: No root cause analysis. Resolving short pays one at a time misses the upstream pattern. A customer with chronic 3 percent short pays may point to an invoice quality problem, not customer bad faith.
Mistake 3: Single workflow for all categories. Authorised deductions, unauthorised deductions, pricing disputes, and quality claims each warrant different resolution paths. A unified workflow either over-investigates routine deductions or under-investigates real disputes.
Mistake 4: Slow investigation. Short pays sitting in investigation queues age into bad debt at much higher rates than current AR. Best-practice operations resolve short pays within 7 to 14 days; manual teams often run 30 to 60 day cycles.
AI-native AR platforms address the short pay bottleneck by automating classification and investigation:
Mid-market AR teams typically lift recovery rates on unauthorised short pays from 15 to 25 percent baseline to 35 to 50 percent within 12 months of agentic deployment, recovering 0.5 to 2 percent of annual revenue previously written off as accepted deductions.
A Short Pay is when a customer pays less than the full invoice amount, either deliberately due to a deduction or dispute, or in error. Short pays are the operational trigger for most deduction and dispute workflows in B2B AR and the largest source of variance between invoiced revenue and cash collected.
Five common causes: authorised deductions matching contract or promotional terms, unauthorised deductions requiring investigation, pricing disputes (catalog errors, missed discounts), quality or quantity claims (damage, shortage), and process errors (customer clerical mistakes). The split between authorised and unauthorised drives most deduction management workload.
A short pay is the financial event: customer paid less than invoiced. A deduction is the reason or classification: short pay tied to an agreed allowance, programme, or claim. All deductions show up as short pays; not all short pays turn out to be deductions (some are errors or disputes).
Typical B2B short pay rates run 5 to 15 percent of invoices. In CPG selling to large retailers, the rate often hits 20 to 30 percent due to high deduction activity. The split between authorised and unauthorised varies by category and customer maturity, with 30 to 50 percent of total deductions typically recoverable through investigation.
AI-native AR platforms compress short pay resolution by automating classification (matching against contracts and promotional plans), surfacing investigation context via graph-based retrieval, and routing by category to the right resolver. Mid-market teams typically reduce resolution cycle from 30 to 60 days down to 7 to 14 days and lift recovery rates from 15 to 25 percent to 35 to 50 percent within 12 months.
Investigate above a meaningful threshold; auto-accept below. The economics depend on cost per investigation versus expected recovery value. Manual investigation at 20 to 60 minutes per case sets a high threshold; AI-driven investigation at near-zero marginal cost makes full investigation economical even on smaller short pays. The right threshold is determined by the platform's automation capability.
