Touchless rate in accounts payable: how to measure it and what benchmark to expect

Most finance directors at SMEs do not know their touchless rate. They know how many invoices come in. They know how long the team takes. But the ratio of automated to manual stays in the grey zone.

Without that figure, any AP automation conversation is blind. You don’t know where you start, you can’t set a target, and when someone says “we’ll reach 80%” you have no idea whether that is ambitious, modest or impossible for your sector.

This article is for finance directors, controllers and admin leads who want to build the metric from scratch, with a realistic benchmark by sector and no marketing targets. Thesis: measuring your touchless rate well matters more than chasing it. And 100% is not the goal.

What the touchless rate is

The touchless rate (or straight-through processing rate) is the percentage of invoices that go through the whole AP flow with no human intervention. From intake to payment-ready. Zero keys pressed.

Formula:

touchless_rate = invoices_processed_without_intervention / total_invoices_received

The numerator is the hard part. An invoice counts as touchless only if it meets the five conditions simultaneously:

• Automatic intake. Arrives at your system without manual forwarding.
• Structured extraction. Header and line data are read without corrections.
• Automatic match. If it has a PO and a delivery note, it is matched line by line within tolerance.
• Validation with no exceptions. No duplicates, no implausible amounts, no unknown supplier.
• Approval and booking. Lands in the input VAT ledger and the ERP without anyone opening it.

If at any of the five steps a human intervenes (corrects a field, forces a match, approves by exception), that invoice counts as manual. The IOFM and APQC definition is strict on purpose: without that strictness the metric inflates and stops being comparable.

How to measure it step by step

1. Define the window. A full calendar month. Not the last week or the full quarter: a month gives enough volume and little variance.
2. Extract the universe. Pull every invoice received with a posting date inside the window.
3. Tag each invoice as touchless or manual. You need a “human intervention” field in your workflow, or reconstruct it from the audit log. If your ERP has no per-invoice audit log, that is already the first finding.
4. Compute the ratio. Touchless divided by total. Also report the breakdown: how many failed by intake, extraction, match, exception.
5. Repeat monthly. The metric is only useful as a time series.

Sector benchmark

The most cited public benchmarks come from IOFM AP Benchmarking, APQC Open Standards Benchmarking and Ardent Partners ePayables. They all measure mostly Anglo-American universes. Translation to the European SME fabric requires adjustment.

This table shows reasonable estimates for an SME with 100-2,000 invoices/month, starting from IOFM/APQC ranges and discounting two factors that weigh in Europe: high supplier dispersion and low penetration of formalised POs.

Sector	Estimated average	Top quartile estimate
Distribution and wholesalers	30-45%	65-75%
Multi-location hospitality	20-35%	55-65%
Construction and subcontractors	10-20%	30-45%
Retail with warehouse	35-50%	65-80%
Professional services	15-25%	40-55%
Light manufacturing	30-40%	60-75%

The bias is clear: the more structured the buying flow (formal PO, receiving with delivery note, recurring suppliers), the higher the rate. Construction is last by the sector’s nature: certifications, withholdings, measurements, lots of “soft” no-PO invoices.

Seven factors that determine your rate

1. Percentage of invoices with a formalised PO. Without a PO there is no three-way matching and the invoice cannot be touchless even if a perfect IDP reads it. If only 30% of your invoices have a PO before the invoice arrives, your absolute ceiling is 30%.
2. Structural quality of supplier data. Native PDF or XML/FacturaE gives clean data. A blurry fax scan is noise. Verifactu and Peppol BIS raise the average bar.
3. Warehouse discipline with delivery notes. If the warehouse does not record receipts with per-line quantities, three-way matching is impossible. The rate drops even if AP runs perfectly.
4. Number of recurring suppliers. One hundred suppliers with five invoices/month is easier than one thousand with one. The long tail concentrates exceptions.
5. Purchase catalogue stability. If line descriptions change every time (supplier concatenates lot, date and reference), description-based matching fails.
6. Configured tolerance. A 2% / EUR 1.50 OR-mode per-line tolerance covers 80% of legitimate noise. Zero tolerance means everything to exception.
7. AP software maturity. Important, but less so than the six above. Excellent IDP on bad data still yields low rates.

Why 100% touchless is not the goal

Reaching 100% means one of two things: either your business is perfectly standard (it is not), or you are approving problematic invoices. The latter is usual when someone brags about a very high rate.

The real trade-off is touchless rate vs post-approval error rate. Raise tolerance until everything matches and both go up: touchless and errors. Lower tolerance and touchless drops while errors drop too. The sweet spot is not 100; it usually sits between 65 and 80% for a well-managed SME.

The mandatory complementary KPI is post-approval error rate: invoices approved automatically that later needed accounting correction, supplier credit note or VAT adjustment. 75% touchless with 0.5% error rate is excellent. 92% touchless with 4% error rate is a disaster in disguise.

How to improve it

Ten actions ranked by leverage. The first four require no new software.

1. Raise the share of purchases with a formal PO before the invoice. Moving from 40% to 70% raises your ceiling by the same magnitude.
2. Force the warehouse to record delivery notes with per-line quantity, not just “received”. No line, no match.
3. Clean the recurring supplier catalogue: unified tax IDs, standard descriptions, POs with unique reference.
4. Define line-level tolerance (percentage OR absolute euros) by spend category, not global.
5. Ask your top ten suppliers to issue FacturaE or send structured XML alongside the PDF.
6. Implement Verifactu QR reading where available: signed data at source, zero ambiguity.
7. Replace classic OCR with IDP that combines QR, XML and PDF; pixel-by-pixel OCR loses value with structured data available.
8. Connect your AP to the ERP via API, not via manual exports. Every manual export is a guaranteed non-touchless invoice.
9. Detect duplicates at intake by cross-checking tax ID + number + amount + date, not by human inspection.
10. Handle exceptions in a shared queue with SLA. Humans only touch what the system flags.

The rate under Verifactu and Peppol BIS

Two regulatory changes lift the European benchmark floor over the next three years.

Verifactu requires the issuer’s invoicing software to generate structured, signed records with a QR. When the issuer operates in verifiable mode, data arrives signed at the Spanish tax authority in real time. For AP, that means clean, verifiable source data without depending on OCR. The effect on the touchless rate is direct: 10-20 points of automatic uplift when the bulk of your suppliers issue Verifactu, simply because extraction stops being a failure point.

Peppol BIS Billing 3.0, the European framework for structured e-invoicing, has a similar effect. Data arrives in standard, parseable XML. Extraction becomes parsing, not OCR.

Related KPIs to watch together

• Cost per invoice. Total AP cost (salaries + software + overheads) divided by invoices processed. APQC puts the top quartile under EUR 2/invoice and the median at EUR 6-10/invoice.
• Cycle time. Hours or days from receipt to approval. Top quartile under 2 days, median 5-7 days.
• Exception rate. Percentage of invoices that stop for exception. The natural complement to touchless.
• Post-approval error rate. As described above. Keep this below 1%.

These four together tell the story. Touchless alone is gameable. The five together are not.

Controller checklist: eight actions for this month

1. Document the touchless definition applicable to your company (the five conditions), publish it and sign it.
2. Pull the first data point for last month with the five-step method. Do not wait for software.
3. Also calculate cost per invoice and cycle time for the same month. Three metrics, one snapshot.
4. Identify the dominant bottleneck: intake, extraction, match or exceptions. One of the four weighs more.
5. Ask the top ten suppliers what percentage already issues Verifactu, FacturaE or structured XML.
6. Audit warehouse delivery-note discipline: how many arrive on time, how many with per-line quantity, how many with PO reference.
7. Define a 12-month target (not 90 days). Realistic: 15-20 points above the starting point.
8. Set up monthly reporting to the finance committee with the five metrics. Without recurring reporting, the metric dilutes.

FAQ

What exactly counts as touchless?

An invoice that goes through the five steps (intake, extraction, match, validation, approval) with no human keying, correcting or manually approving. The IOFM definition is strict on purpose.

What is a realistic rate for an SME?

Depends on sector. The average is 10-30% without modern AP software. Top quartile reaches 60-80% per IOFM/APQC. An SME with PO and delivery-note discipline, modern IDP and recurring suppliers should target 60-70% in 12 months.

Why should I not chase 100%?

Because 100% is only reachable by relaxing tolerances until problematic invoices get approved. The metric inflates but post-approval error rate explodes. Optimum sits between 65% and 80% with error rate below 1%.

What software do I need to measure it?

To measure it, none. An ERP CSV export with an audit field is enough. To raise it, IDP combining OCR, QR, XML/FacturaE, line-by-line three-way matching and ERP export.

How does Verifactu affect my rate?

It raises it. When the issuer operates in verifiable mode, signed data arrives with a QR and, in many cases, attached XML. Extraction stops being a failure point. Expect 10-20 points of uplift.

What are the mandatory complementary KPIs?

Cost per invoice, cycle time, exception rate and post-approval error rate. The five together tell the real story.

What about no-PO invoices?

They cannot be touchless via three-way matching, but they can be via two-way (invoice vs delegated approval) if you have a delegated validation flow. The correct calculation is: % of PO invoices × touchless rate of PO invoices + % of no-PO × touchless rate of no-PO. The latter is usually much lower.

Three things to remember

1. The touchless rate is the percentage of invoices that go through the five steps with no human intervention. Strict definition or it stops being comparable.
2. The benchmark depends on sector. SME average sits at 10-30%; top quartile at 60-80%. Verifactu and Peppol BIS lift the floor.
3. 100% is not the goal: it inflates the metric while exploding post-approval errors. The sweet spot is 65-80% with error rate below 1%.

If you want to see your touchless rate on real invoices, try ininvoice. Also pricing and features.

Related content

• Touchless accounts payable
• Three-way matching
• IDP vs OCR on supplier invoices
• Price variance invoice vs PO
• Supplier invoice exception management