Freight Invoice Automation
Definition, process and key metrics of automated freight invoice auditing in logistics and forwarding
Freight invoice automation (also freight audit) is the systematic matching of incoming freight invoices against agreed conditions and the transport services actually rendered. Invoices are checked against tariffs and rate cards, against surcharge agreements (diesel, toll, currency) and against transport documents such as the CMR consignment note or customs papers. Automated freight invoice auditing performs this reconciliation at line level — without manually screening every invoice.
Key takeaways
- Freight invoices are highly variable: base freight plus diesel, toll, currency and peak-season surcharges plus accessorials — every line can deviate.
- The core of automated auditing is the 3-way match: invoice ↔ tariff/rate card ↔ transport evidence.
- At a German logistics group, feld.ai reconciles 96% of freight invoices correctly on the first pass; the German go-live was followed by a worldwide rollout (references).
- The context keeps growing: Germany processes around 7 billion invoices per year, and roughly 560 billion invoices were exchanged worldwide in 2024 (source: Billentis Report 2025) — logistics invoices are among the most audit-intensive of them.
Why is freight invoice auditing so labour-intensive?
A freight invoice is rarely “one number”. It is composed of:
- base freight per lane, weight/volume and mode,
- variable surcharges: diesel floater (often adjusted monthly), toll, currency surcharges, peak-season surcharges,
- accessorials: demurrage, storage, customs clearance, notification, dangerous goods,
- contract-specific rules per customer, carrier and period.
The reference data is scattered: rate cards and surcharge tables as Excel files in hundreds of formats, contracts as PDFs, shipment data in the TMS. Manual line-level matching quickly costs several minutes to a quarter of an hour per invoice — at thousands of invoices per month, that is barely economical. In practice, many companies audit only samples, or not at line level at all — and deviations go unnoticed.
How does automated freight invoice auditing work?
The flow in an AI-powered pipeline:
- Capture. Freight invoices arrive as PDF, scan or email attachment and are read without templates — header data and all line items, even on unseen layouts.
- Ingest reference data. Rate cards, surcharge tables and carrier statements — often heterogeneous Excel files — are captured in structured form via AI table processing.
- 3-way match. Every invoice line is checked against the agreed tariff and the transport evidence (CMR, shipment data, customs documents such as EUR.1): do lane, weight, surcharge rates and accessorials agree?
- Decision. Matching invoices are released for payment (dark processing); deviations land in the review screen with the source location highlighted (visual grounding) and the difference amount computed.
- Learning. Every human correction feeds back into the model — the exception rate falls over time.
Worked example: what does automation save?
An illustrative calculation for a shipper or forwarder with 3,000 freight invoices per month and 10 minutes of manual checking time per invoice (line-level match against the rate card):
- Manual: 3,000 × 10 min = 500 hours per month — roughly 3 full-time roles.
- Automated at a 96% first-pass rate: 120 exceptions × 10 min = 20 hours, plus approvals — realistically under 50 hours per month.
- In addition, deviations are found systematically rather than by sampling: miscalculated surcharges and double-billed accessorials show up under 100% auditing, not just in the sample.
The order of magnitude is proven in production: at the German logistics group Logwin, 96% of freight invoices are correctly reconciled on the first pass — followed by a worldwide rollout (references).
Frequently asked questions (FAQ)
What is freight invoice automation?
The automated matching of incoming freight invoices against agreed conditions and services actually rendered — against tariffs, rate cards, surcharge tables and transport documents such as the CMR consignment note or customs papers.
Why is freight invoice auditing so labour-intensive?
Because many variable components come together (base freight, surcharges, accessorials) and the reference data sits in heterogeneous formats. Manual line-level matching is barely economical at high volumes.
What does 3-way match mean for freight invoices?
The three-way reconciliation of freight invoice, agreed tariff (rate card/contract) and transport evidence (e.g. CMR consignment note, shipment data). Only when all three agree is the invoice released automatically.
What recognition and straight-through rates are realistic?
feld.ai achieves header-field recognition rates above 96%; at a German logistics group, 96% of freight invoices are correctly reconciled on the first pass.
Does automated auditing work with Excel rate cards?
Yes — AI-based table processing reads heterogeneous rate cards and surcharge tables without templates and provides them to the match as structured reference data.
Further reading: The solution page Freight Invoice Automation with AI shows the 3-way match with customs and freight documents in detail — EU-hosted, live in under 4 weeks.