Key Takeaways
- Among users of the cash application software The Hackett Group assessed, the median auto-match rate is 70% and a third exceed 80% (The Hackett Group, 2025).
- The same 11,000 payments read anywhere from 26.7% to 91.8% under seven defensible definitions, so a rate without its definition cannot be compared.
- Five choices move every rate: the unit, the denominator, the numerator, the timing and who counts as a touch.
- Report the rate with five companions: correction rate, unapplied share, time to post, value-based rate and exception ageing.
- Set targets per payment segment from a three-month baseline of your own, not as one blended number.
In This Article
- Key Takeaways
- What is a good cash application auto-match rate?
- Why can one month of payments read 27% or 92%?
- Which five choices change the number?
- Which cash application metrics belong next to the rate?
- How do you set your own target?
- What should you ask when a vendor or report quotes a rate?
- What data do you need to measure it?
What is a good cash application auto-match rate?
The Hackett Group puts the median auto-match rate at 70% among users of the cash application software it assessed, with a third above 80% (2025).
Auto-match rate definitions: what each counts and what it hides
These are this page’s working definitions. We found no definition that every report shares, and reports use these names for different formulas.
Hackett’s 2025 evaluation covered 15 vendors of cash application software, so its figures describe their end users, not every finance team. The same study notes that “a surprising number of organizations are accepting automation rates below 40%” (The Hackett Group, 2025). We found no 2026 figure, so this 2025 study is the latest neutral one we could verify.
Hackett itself scores software on “the touchless automated match rate, remittance-less match rates, application cycle times and the reduction of unapplied cash”: a scorecard, not one rate.
Higher “typical” figures circulate in AI answers; they come from vendors’ own pages without a stated definition, so this page does not repeat them.
Takeaway: a good rate beats the 70% median under a definition you state, with a low correction rate beside it.
Why can one month of payments read 27% or 92%?
The same 11,000 payments produce seven rates, from 26.7% to 91.8%, because each view counts a different numerator, denominator or unit.
The month below is illustrative and fictional. The bank statement shows 12,000 incoming lines; 1,000 of them are not customer payments (interest, cash pool sweeps, supplier refunds). Of the 11,000 customer payments:
- A: 6,300 matched and posted with no human action (USD 24.0M)
- B: 1,200 where an analyst accepted the system’s proposal with one review click (USD 16.0M)
- C: 600 where an analyst edited the proposal (USD 12.0M)
- D: 2,000 matched manually (USD 30.0M)
- E: 900 still unapplied at month end (USD 8.0M)
Value received is USD 90.0M. Later in the month, 63 of the 6,300 automated matches in A are reversed.
One month, 11,000 customer payments, seven match rates (illustrative)
All seven are honest, and a team reporting 91.8% and one reporting 26.7% may describe the same month. The 91.8% says nothing about the 2,600 payments people matched or edited by hand.
By value, this month reads lower than by count because the manual group D averages USD 15,000 a payment while the touchless group A averages about USD 3,800. That direction depends on your payer mix. If your large payers send structured remittance, the value-based rate reads higher than the count-based one, as in the example in our touchless cash application guide. If your largest payments are the complex ones, it reads lower, as here. Report both and the mix shows itself.
Takeaway: a match rate is a definition first and a number second.
Which five choices change the number?
Five choices move any match rate: the unit, the denominator, the numerator, the timing and who counts as a touch.
Five choices behind every rate, and what each did to the worked month
The numerator moves the rate most. Read it as a ladder: system proposed, proposal accepted unchanged, posted with nobody involved, applied by anyone by month end. Each rung is a different metric.
The denominator is where “eligible” hides: it can quietly drop the payments with no remittance, which are the ones that make the work. Customer payments are the cleaner base.
Timing matters because matches get reversed: 63 of the 6,300, or 1.0%, after the first read. Read at month end and move reversed matches back to touched.
Who counts as a touch has to be written down. Our touchless guide counts a payment as touched when anyone searched, keyed, split, reassigned or overrode anything in its match, and treats approving a batch of matches that already passed posting validation as a control, not a touch. Whether a per-payment review click counts is your decision: here it moves the rate by 10.9 points. Time those reviews separately either way.
Takeaway: write the five choices in one line next to every rate, for example “count, customer payments, posted untouched, month end after reversals, batch approval not a touch: 56.7%”.
Which cash application metrics belong next to the rate?
Report the auto-match rate with five companions: correction rate, unapplied share at period end, time from receipt to posting, the value-based rate and exception ageing.
Companion scorecard for the auto-match rate
Hackett’s study gives two reference points for the time and unapplied rows: 61% of assessed end users “are matching payments on the same day they are received”, while at the weak end organizations take “up to three days to post payments” and carry “up to 6% of unapplied cash” (The Hackett Group, 2025). The illustrative month’s 8.2% unapplied share is above even that weak-end figure, which a 91.8% total applied rate would never show.
Two patterns to watch. A rate rising with the unapplied share means payments are being parked; the unapplied cash backlog guide shows how to work that pile. A rate rising with the correction rate means the matcher got bolder, not better; AI payment matching accuracy covers false matches in depth.
Unapplied cash also distorts DSO: invoices a payment should have cleared stay open, as the guide to when DSO misleads explains.
Takeaway: a high rate with a high correction rate is rework, not automation.
How do you set your own target?
Set targets per payment segment from your own baseline, because the reachable rate depends on how much remittance detail arrives with each payment.
Payment segments: what limits the rate and how to measure it
No-remittance payments deserve their own line: 51% of the end users Hackett assessed said their software delivers “remittance-less match rates for up to 60% of their cash application volume” (The Hackett Group, 2025). Blended in, they hide where the work sits.
The method:
- Take three months of payments as the baseline and tag each payment with its segment.
- Report the strict touchless rate, the correction rate and the unapplied share per segment, by count and by value.
- Fix the cause in the worst segment first: usually remittance capture (remittance advice sources), unknown payers (matching a payment with no reference) or deductions (partial payments and deductions).
- Set next quarter’s target per segment against its own baseline, then re-read monthly with the five choices fixed.
We found no neutral benchmark per segment, so this page sets no target values; your baseline is the reference.
Takeaway: a blended target hides the segment that drives the workload.
What should you ask when a vendor or report quotes a rate?
Ask seven questions of any quoted rate, starting with whether it counts payments or value and which payments sit in the denominator.
- By count or by value? In the worked month, the same matches read 57.3% by count and 26.7% by value.
- Which payments are in the denominator? Customer payments, all incoming bank lines or “eligible” payments: 57.3% against 52.5%, and “eligible” can drop the hardest payments altogether.
- Does a per-payment review count as a touch? 57.3% if it does, 68.2% if it does not.
- Is posting included, and by whom? Proposals alone read 73.6%, untouched postings 57.3%, everything applied by anyone 91.8%.
- When was it read? A pilot month, a best month or a twelve-month average at month end are three different numbers.
- Are reversals removed? 57.3% before, 56.7% after the 63 reversed matches.
- Which payers are in the sample? A sample of structured-remittance payers reads higher than the full book; ask for the split by segment.
Our own figure: ClearMatch reaches a 95% auto-match rate, and nothing posts without human sign-off (Transformance customer data, 2026). Do not set it next to the Hackett median; the populations differ. The fair test for any vendor’s rate, ours included, is your own payments under your written definition, which is what a proof of value does; it runs 4-5 weeks on one entity.
For the payments the rules leave open, Vero, the AI agent, investigates the exception from Customer 360, its working memory for each customer, and fetches internal and external information to find more matches. It proposes the match with the evidence attached, a person approves, and postings follow your approval thresholds. The rules vs AI comparison shows which payments each approach handles.
Takeaway: a rate quoted without its five choices is a claim, not a benchmark.
What data do you need to measure it?
You need three records per payment: the bank statement line, the open items it cleared in the ERP, and who or what made the match and when.
- Bank statement lines: amount, value date, payer and remittance text per incoming line, from camt.053, BAI2 or MT940 (bank file formats). They give the denominator and separate customer payments from sweeps and interest.
- ERP clearing records: which open items each payment cleared, when, and any residual or on-account posting. This gives the numerator, the unapplied share and time to post.
- A matcher and approver log: for every payment, whether the system or a person matched it, what was changed, who approved and when. This decides strict touchless against accepted proposal.
Join the three on the bank line and every rate on this page falls out of one query. The third record is the hard part. Check whether your ERP records more than who posted the clearing; if it cannot show whether a person searched or edited the match before posting, tag payments by hand for one month before you trust a strict touchless figure.
In Transformance, the audit log stores the timestamp, what the system did, what the user did, who approved, the input and the confidence. The audit trail and reversals guide shows how such a log is used.
Takeaway: you cannot report a strict rate you do not log.
Who Transformance is for
- ERPs: SAP S/4HANA (on-premises, private or public cloud) and SAP ECC, NetSuite (including OneWorld subsidiaries), Microsoft Dynamics 365 Finance & Operations and Business Central, and Odoo, all in production with customers; Oracle and Infor supported.
- Size: mid-market to large enterprise, 250 to 2,000+ employees, including groups with several entities and shared service centers.
- Payment volume: from under 1,000 to over 100,000 payments a month.
- Industries: manufacturing, consumer goods and food and beverage, logistics, e-commerce, agriculture, energy and professional services.
- Scope: cash application alone, or the full order-to-cash (cash application, deductions, AR collections, cash forecasting).
Conclusion: define the rate before you compare it
A good auto-match rate beats the 70% median of the users Hackett assessed, under a one-line definition, by count and by value, with the correction rate and unapplied share beside it. Measure per segment, log who touched what, and ask every quoted rate the same seven questions. Transformance works with customers across Europe, the United States and Australasia, including Syngenta, Eberspächer and PwC.
Key terms
- Cash application: matching each incoming payment to the open invoices it pays and posting the result.
- Touchless cash application: payments matched and posted with no manual work in the match.
- Straight-through processing: a transaction handled end to end without manual intervention.
- Remittance advice: the customer’s list of the invoices and deductions a payment covers.
- On-account cash: customer money posted to the account but not yet matched to an invoice.
Related guides
- Which ERP systems does Transformance integrate with?
- What is touchless cash application?
- Rule-based vs AI cash application: which payments go where?
- Why does unapplied cash pile up, and how do you clear it?
- Where does remittance advice come from, and how is it captured?
- How accurate is AI payment matching, and which payments still need a human?
- What does manual cash application cost per payment, and how do you build the ROI?
Frequently Asked Questions
What is a good cash application auto-match rate?
Among users of the cash application software Hackett assessed, the median is 70% and a third exceed 80% (The Hackett Group, 2025). A good rate beats it under a stated definition, with a low correction rate.
How do you calculate the auto-match rate?
Divide the payments matched and posted with no human action by the customer payments received in the same period, and state whether you count payments or value. Move matches reversed before month end back to touched.
Is auto-match rate the same as straight-through processing?
Not necessarily. A straight-through rate usually requires posting without intervention and often uses a denominator of payments in scope, while some auto-match figures count proposals only. Ask which definition a number uses.
Which ERPs and what payment volume does this apply to?
The measures work on any ERP that records open items and clearing. Transformance connects to SAP S/4HANA, SAP ECC, NetSuite, Microsoft Dynamics 365 Finance & Operations, Business Central and Odoo, all in production with customers, supports Oracle and Infor, and handles from under 1,000 to over 100,000 payments a month.
How long until we can measure our own baseline?
Three months of payments give a baseline by segment. With Transformance, the ERP connection typically takes 2-3 days and your IT team may be involved; a proof of value runs 4-5 weeks on one entity, and a full rollout takes up to 4-8 weeks.
What does it cost to improve the rate?
It depends on scope and the number of entities, so there is no standard figure. Book a Call and we size it for your set-up.
What does ClearMatch’s 95% auto-match rate mean?
ClearMatch reaches a 95% auto-match rate, and nothing posts without human sign-off (Transformance customer data, 2026). Apply the seven questions to it as to any rate, and test it on your own payments in a proof of value.
SAP, SAP S/4HANA, Microsoft, Dynamics 365, Business Central, Oracle, NetSuite, Odoo and Infor are trademarks of their respective owners (SAP SE, Microsoft Corporation, Oracle Corporation, Odoo S.A., Infor). Transformance is not affiliated with or endorsed by them.
