Most comparisons of multi-PSP reconciliation solutions ask the same question: does the tool match payments? Almost every product can match a settlement total to a bank line. That is not the hard part.
The hard part is what happens to one transaction when it moves through several processors, a refund, a fee and a currency conversion. Where did the money come from, what did each step cost, and which journal did it land in? If a tool cannot answer that for one transaction, it will not answer it for a million.
So here is a better test. Pick one transaction and ask every vendor to trace it from the original order to the ledger entry. Ask them to show each event, each fee, each FX difference and each journal they created. Then compare the trails, not the feature lists.
The example
This is an illustrative example, not a customer case. The figures are used to show the mechanics.
A customer pays €100 for order 12345 by card, through PSP A. The processor takes a €2.50 fee and schedules €97.50 to the bank. Two days later the customer returns part of the order and a €20.00 refund is issued. The processor returns €0.30 of the fee on the refund. The bank should therefore receive €77.80. It receives €77.30.
A tool that works at total level shows a settlement that is €0.50 short. A tool that works at transaction level shows where the €0.50 sits:
?Settlement is €0.50 short. Where does it sit?
- Order event€100.00 against order 12345, in EUR.€100.00
- Payment event€100.00 captured at PSP A.€100.00
- Fee event€2.50 processing fee, booked against the sale.−€2.50
- Expected settlement€97.50 due from PSP A.€97.50
- Refund event€20.00 returned to the customer, with €0.30 of fee returned by the processor.−€19.70
- Expected settlement after refund€77.80
- Bank event€77.30 received.€77.30
- Result: €0.50 unexplained.Flagged against this transaction ID and this journal, not netted into a suspense balance that nobody owns.−€0.50
The first kind of tool gives you a question. The second gives you an answer.
The categories you will meet
PSP reporting
Your processor’s reports are accurate for its own transactions. They cannot see your other processors, your bank or your ledger. They are a good starting point if you run one provider.
Matching engines for finance operations
Simetrik and Solvexia are the best-known names here. Simetrik describes a rules-based core, record-level lineage and a connector catalogue it puts at 5,500 data sources. Solvexia covers bank, GL, payment and intercompany reconciliations and describes audit trails. Both are broad platforms, and both publish match-rate claims.
Askfor the method behind those numbers in writing.
Enterprise close platforms
BlackLine’s transaction matching is built for high-volume work such as cash, card, and intercompany reconciliations. Business users own the matching logic, and manual matches can go through an approval workflow. Its public material says less about how entries reach the ERP and what the audit log records.
Askhow entries reach the ERP and what the audit log records.
Revenue tools for subscription businesses
Leapfin and HubiFi focus on order-to-cash and revenue recognition. Stripe and PayPal are core sources, and NetSuite or QuickBooks are the ledgers. They are strong when your problem starts at the invoice. They are less obviously built for a business that runs several acquirers and payout files.
Mid-market and regulated finance
ReconArt suits mid-market finance teams with bank and balance sheet reconciliation. AutoRek is built for regulated financial services such as safeguarding and client money. It does not publish prebuilt ERP connectors, and its own material says it is built for regulated firms rather than general corporate use. If you are a high-volume merchant, you are probably not its target buyer.
Merchant tools from payment providers
ReconHub is Worldline’s reconciliation product for merchants, with posting to more than 35 ERP systems. If you bank with Worldline it deserves a close look. It is built around that provider’s payment data.
Where Actuals fits
Actuals is built for the point where transactions from many sources have to become accounting. Although we’re connector agnostic, we offer various out of the box options (Adyen, Mollie, Stripe, PayPal). We have also closely known options such as Nuvei, Buckaroo, Checkout.com, Braintree, Riverty, and others. Which provider a business uses, and how many it runs, does not change the approach. The platform pulls every transaction onto one layer, standardises the formats through mapping, matches automatically and reconciles automatically.
- Pullevery transaction onto one layer
- Standardisethe formats through mapping
- Matchautomatically
- Reconcileautomatically
From there, each transaction produces several journal entries. Fees, FX differences, and every event along the way are part of the calculation, not a footnote. The result is a deterministic data layer from origin to end status. The trail in the example above is not a special report we build for demos. It is how the data is stored.
We also believe the general ledger should only receive what has been reconciled.
Raw events do not belong there, and an unexplained difference should stay visible until someone resolves it.
Where Actuals is not the right fit
- One processor, reporting is enoughIf one processor handles all your payments and its reporting already answers your questions, you do not need a multi-PSP layer yet.
- Only bank lines to invoicesIf your only problem is matching bank lines to invoices, with no fee or FX complexity, a simpler bank reconciliation solution will do the job.
- Regulated client moneyAnd if you are a regulated financial firm looking for client money or safeguarding controls, start with tools built for that work.
The checklist
Use these questions with any vendor, including us:
Tick each question once a vendor has answered it with an example.
If a vendor cannot answer these with examples, the feature list does not matter.
Close
Financial truth is not something you claim in a brochure. It is something you can show, one transaction at a time, from the first event to the last. Run this test on every tool on your shortlist, and on ours.
The trail will tell you more than the feature list.
Written by
The Actuals teamAmsterdamWant to run this test on your own data? Talk to us.



