Wise Business: Which Cross-Border Job Should Lead for a Small Business?

Wise's help centre says it marks a transfer Complete "when we've sent the money to the recipient bank". The same page adds that the receiving bank can take up to a working day to release the funds, and that the amount can look different if that bank converts it again.

That is a definition, not a defect. It does raise a question for anyone paying an overseas supplier: when is the job finished? When the invoice is approved, when Wise says Complete, when the supplier sees the agreed amount, or when the bookkeeper has matched the debit to the bill?

This note is an operator analysis, not a Wise tutorial or a provider ranking. Our argument is that a small business should choose a recurring obligation before it chooses an account pitch. For one segment, Irish businesses funded in euro with recurring approved sterling supplier invoices, we would lead with paying and reconciling the next overseas supplier invoice. That is our prioritisation hypothesis. For a business that mainly sells abroad, receiving may lead instead. We end with a test of the lead message.

What we looked at

We read Wise's Irish receiving-fee page and help pages, two Wise customer stories, Irish bank and Revolut documentation and the National Bank of Belgium's published notices, retrieved on 3 November 2026. We did not use a signed-in account, send a transfer, interview an operator or see Wise's funnel. The customer accounts are selected successes. Nothing here shows an activation or retention problem at Wise. Several help pages show no date, so check them before relying on them.

Four jobs, four reasons to buy

The table is our model, not a maturity ladder. A business can get lasting value from receiving only. We start with the supplier invoice because the business controls when it pays, has an amount and a deadline to meet, and can inspect the result. Receiving also depends on a customer's payment process. That is our reason for the order and not evidence that sending is worth more. A payroll run is a different obligation, with employment and tax questions a payment tool does not settle, so our segment uses ordinary supplier invoices.

What Wise already provides

A recommendation to add invoice import, approvals or batch payments would duplicate existing work. The pieces that matter here:

The Xero limit is a concrete regional boundary. An Irish business should not assume that connecting Xero delivers every listed function, and our test needs no integration, because it can use the bookkeeper's existing process.

The five events inside the supplier job

We separate five events: the invoice is authorised, the business commits funds, Wise sends the money onward, the supplier receives the agreed amount, and the business matches the debit, the fee and the bill. A problem at any step leaves work open for the owner or bookkeeper.

Our definition of a finished job includes the last two events. That does not mean Wise must own the accounting. A clean handoff to the existing bookkeeper can complete the job, and an automatically filled field is not enough if the supplier reports a shortfall or the fee cannot be explained.

Two published accounts, two jobs

Michael Houghton, in Wise's October 2023 account, is a Limerick consultant whose clients pay him in several currencies into local account details. It is an Irish case where receiving is the job. The account also quotes his own savings, setup time and plan details, and we use none of them, because it is dated and written for the UK site.

Pangolia, in Wise's August 2023 account, is a Singapore pet e-commerce company whose founder says he used Wise personally before using it for the business. The account describes batch payments to about 110 remote team members and payments to suppliers in the US and China. Familiarity may explain how it found Wise, and the recurring obligations explain why it kept using it. A company with a team payout run is not a baseline for a five-person Irish firm, and we leave out its savings headline.

Both are Wise-selected accounts. They show mechanisms, cannot rank the jobs, and say nothing about how often a first job leads to repeat use.

Compare the whole obligation

Compare the complete obligation, not the advertised fee. Fix the same net sterling amount, take timestamped quotes from each route, and record the euro required, transfer and funding fees, any deduction at the supplier's end, the delivery estimate and the effort afterwards. This worksheet is ours. Keep cash and time costs separate, and do not count an exchange-rate margin twice when a quote already includes it. We captured no live euro to sterling quote, so we make no claim about price.

Two boundaries matter for an owner asked to move supplier payments. Wise's EU safeguarding page says payment services are not covered by the Belgian Deposit Guarantee Scheme and that customer funds are held separately, in cash at banks and in money market funds. That is safeguarding and not deposit insurance. And no business has to move its working capital to Wise to run a single job.

The National Bank of Belgium supervises Wise Europe SA, which Wise's EEA business agreement names as the contracting party for businesses registered in the EEA. On 25 August 2026 it published two notices about Wise Europe SA under its anti-money-laundering powers: a decision and a second notice. We make no claim here about their findings or current status. They are not evidence that any payment failed, or of how long any customer's verification takes.

A test with one change

Question. For the segment above, does leading with the next approved sterling supplier invoice produce more verified first Wise payment jobs than a broad international-account lead?

Segment. Consenting Irish businesses with no earlier Wise Business supplier payment, a genuine recurring sterling obligation identified before assignment, an authorised operator, enough euro funds and a working fallback.

Audit first. Walk a few consenting businesses through one ordinary invoice and trace it to the supplier and the books. Check the live entry flow too. If it already leads with the supplier invoice there is no contrast, and the test should not run. The walkthrough is a condition before running the test, not a reason to hold back this proposal.

Control. The truthful broad lead the current entry route uses.

Treatment. The same lead module with one difference: it foregrounds paying the next approved sterling supplier invoice. Placement, length, call to action, disclosures, pricing, product access, support and reminders stay the same, with no new integration, payment route or incentive.

Main outcome. The share of all assigned businesses whose identified invoice is paid through Wise, with the agreed net amount confirmed received by the original deadline and the records matched within an interval set in advance. A Complete status alone does not count. This is our measure, not Wise's, and the components are reported separately so the composite cannot hide where failures occur.

Assignment and analysis. Randomise by business, and keep businesses that share an accountant together. Analyse as assigned. A business that never opens the message, fails verification or pays through its bank stays in its group, and effort and cost are counted across every assigned business. Unverifiable receipt is missing evidence, neither success nor failure.

Ordinary work comes first. Nobody delays an invoice, waives an approval or changes a beneficiary check for the study. A payment made through the usual provider is a business success, and we report it separately. Allow time for Wise's normal verification process, including any additional information requests, before relying on the new route for a payment deadline. Wise's verification guide says further verification usually takes up to 10 working days. The full plan is in the methodology appendix.

Reading the result.

• More clicks or funded balances but no more verified jobs: no benefit has been shown.
• More completed jobs but payment accuracy or timeliness outside the predeclared limits: do not expand. If those limits are met, assess whether added support and bookkeeping costs erase the gain.
• Gains only among a small group that gets through verification while many drop out: look at the whole cohort.
• No detectable difference: inconclusive. It does not show the two leads are equivalent.

We would expand the message only if completed first jobs improve by a practically worthwhile amount, payment accuracy and timeliness stay within limits set before assignment, and the extra setup, assistance and bookkeeping do not erase the gain. A good result would justify the message for this segment. It would not show that sending beats receiving, that Wise beats a bank, or any effect on retention or revenue.

The case against, and the decision

• Receiving may be the real constraint. If overseas customers struggle to pay, local receiving details can be worth more than another way to settle bills. Michael's account makes that concrete.
• The incumbent bank can already pay. AIB documents standard and urgent international payments with shared charges, so the full amount may not be credited. Its currency account lets a business pay and receive in one currency without converting to euro, although the page says a pause on new Sterling Current Accounts remains in place.
• A subscription may fit the whole workload. Revolut's Irish plans list monthly prices from €10, €35 and €125, with 0, 5 and 25 no-fee international transfers a month. We captured no overage fees and ran no head-to-head test.
• A message may not remove the bottleneck. Missing authority, unfinished verification, a cash shortfall or a wrong invoice can dominate, and better copy cannot fix them.
• One occasional bill may not justify a new process. Frequency and fit set the limit.

What we do not know: how first jobs are distributed across Wise's customers, how many businesses abandon setup, which job drives repeat use and what each costs to acquire.

The decision follows. Choose a recurring obligation, count receipt and records as part of the job, and compare the real alternative. Lead with the supplier invoice for an Irish business with recurring sterling bills, and with receiving where a customer's payment is the friction.

Next step

Product Garage works through live product and go-to-market decisions with experienced operators. This note takes one of those decisions in public, while the thinking is still provisional.

Subscribe to Garage Dispatch to get the next working note.

FAQ

Does Complete mean the supplier has the money? No. Wise says it marks a transfer Complete once it has sent the money to the recipient's bank. The receiving bank may take up to a working day to release it, and the amount can look different if that bank converts it again.

Does Wise's Xero connection import bills for Irish businesses? Wise's Xero guide limits bill importing to UK businesses and says some advanced settings are not available to new customers. The page shows no date, so check the live version for your account.

Is receiving money with Wise free? Partly. The Irish receiving page lists a one-time €60 fee for account details, free domestic receiving in several currencies including EUR and GBP, and fixed charges for Swift receipts. Fees change, so check the live page.

Are Wise Business balances covered like a bank deposit? Wise says payment services are not covered by the Belgian Deposit Guarantee Scheme, and that customer funds are safeguarded separately from its own money.

Does the test show Wise is cheaper, or that sending beats receiving? No. It can show whether leading with the supplier invoice produces more verified first jobs for one segment. Price and the other jobs are separate questions.

Methodology appendix

This appendix supports the working note "Wise Business: which cross-border job should lead for a small business?". It is a proposal, not a report of a test that has been run. No sample size, observation window, uplift target or waiting interval is specified, because the sources give no baseline to set them from.

The question and what it covers

For the defined segment, does leading with the next approved sterling supplier invoice produce more verified first Wise payment jobs than a broad international-account lead? This is a test of a communication and onboarding message. It is not a head-to-head provider trial, a test of a new payment feature, or a test of whether sending is better than receiving. Other lead jobs need separate, segment-specific tests.

Segment and screen

• Segment: consenting Irish businesses with no earlier completed Wise Business supplier payment, a genuine recurring sterling obligation identified before assignment, an authorised operator, enough euro funds and a working fallback provider. The exact legal-entity, corridor, invoice and readiness rules are set before assignment. Exclude emergencies using an agreed feasibility rule. No standard number of days is inferred from the sources.
• Exclusions: businesses whose main problem is unpaid customer invoices, payroll compliance, emergency remittances or an unsupported account configuration.
• Audit before assignment: observe a few consenting businesses with one ordinary invoice each. Check the invoice is genuine and approved, the beneficiary details are independently confirmed under the business's normal policy, the funding account is usable, and the payment path and bookkeeping handoff can be observed. Do not collect banking data beyond what the check needs.
• Questions the audit answers:
  • Current process: what is paid, by whom, from which currency, with what fee and by what deadline? How much work follows the send button?
  • Alternative: if another provider already handles the job cheaply and reliably, the claim may have little room to help.
  • Entry route: what does the live first-touch communication already say? If it already leads with the supplier invoice, there is no contrast at that touchpoint, and the test should not run there.
  • Configuration: which Wise and accounting settings does the business actually have? Features and regional availability vary, for example the Xero bill-import limit in Wise's guide.
• Feasibility versus assigned-group analysis: businesses that fail the screen are excluded before assignment and reported, with reasons, as a feasibility finding. They are not outcome data. A business that is assigned and then fails verification, abandons setup, switches provider or loses authority stays in its assigned group. Its outcomes, effort and cost remain in the main analysis.

Conditions held equal

• Shared setup: the same factual disclosures, pricing, product access, downstream guidance, support budget and reminder schedule. No new integration, payment route, incentive, permission change or accounting automation in either group.
• Control: a truthful broad international-business-account lead, based on the actual current entry route once audited.
• Treatment: the same lead module, with the same placement, length, prominence and call to action, that foregrounds paying the next approved sterling supplier invoice. The task-framing text is the one assigned difference.
• Prototype note: if a research prototype is used, describe it as a prototype. It can test comprehension and task choice. A static concept cannot establish real payment completion. Testing the operational outcome needs the authorised live experience and consistent access to evidence. Do not claim Wise lacks an onboarding flow because one was not visible without signing in.

Cohort and outcome

• Primary cohort: all assigned eligible businesses, including those that never read the message, fail verification, pay through another route or close the account.
• Primary outcome: the share of the cohort whose identified obligation is paid through Wise, with verified correct beneficiary receipt of the agreed net sterling amount by the original payment deadline, and matching records completed within a separately predeclared reconciliation interval. This is our activation measure, not Wise's definition.
• Components reported separately: account and verification steps, payment made through Wise, correct amount, timely receipt, and reconciliation. The composite must not hide where failures occur.
• Receipt evidence: use redacted evidence or consented confirmation, collected under the same rule in both groups. A platform Complete status alone does not meet the receipt condition, because Wise defines Complete as the point it has sent the money to the recipient bank.
• Deadlines: payment deadlines are the business's own facts, not test targets. The researcher cannot extend them to create a success. Each invoice is allowed its complete follow-up interval.

Design and analysis

• Randomise by business, and keep contacts for the same business together. Where one accountant or operator handles several participating firms and the messages could contaminate their behaviour, use that shared unit as the cluster or redesign recruitment. Stratify where feasible on prior personal Wise use, payment frequency, invoice size, existing provider and baseline bookkeeping setup.
• Qualitative first. Observe a few businesses, record baselines, then size any controlled study from baseline rates, clustering and the smallest practically worthwhile gain. Do not invent a sample size or an uplift target.
• Predeclare the observation window from real invoice cadence, each component, the reconciliation interval, missing-data assumptions, and the workload and error limits.
• Analyse as assigned. Report actual exposure separately. Use cluster-aware uncertainty where businesses share an accountant.
• Missing evidence. A confirmed unpaid or late invoice fails the timely-payment component. Unverifiable receipt is missing evidence, neither a success nor a failure. Report missingness by group and use sensitivity analyses. A result limited to businesses that made a transfer cannot support a claim about the whole assigned group.
• Expansion rule. Expand the framing only if completed first jobs improve by a practically worthwhile amount, payment accuracy and timeliness stay within the limits set before assignment, and extra setup, fees, assistance and bookkeeping do not erase the gain. Include costs for unsuccessful businesses. If the only effect is more clicks or funded balances, an operating benefit has not been shown. A small null result is inconclusive, not equivalence.
• Serious events. A serious payment error or disclosure pauses the affected implementation. Record the event and the exposure, and keep the original assignment for analysis. A small pilot cannot show the absence of rare serious errors.
• What it cannot show. The test estimates the effect of this framing for businesses already selected for supplier demand. It cannot rank receiving against sending across the market, show that Wise beats a bank on every payment, identify the effect of a new integration, or establish causal retention or revenue.

Protection of ordinary work

• Nobody delays an obligation, waives an approval control, changes a beneficiary check or keeps trying Wise to satisfy the study.
• The usual provider remains available. A payment completed through it is not a primary Wise-use success, but it may still be a good business outcome, and it is reported separately.
• Rescue rules and serious-event pauses are identical in both groups and set before assignment.

Measures

Cost worksheet (ours)

• Fix the same net sterling obligation and take timestamped quotes from each route on the same date and payment method.
• Record the euro required, transfer and funding fees, any recipient deduction, the delivery estimate and the operational effort.
• Keep sender cash outlay and recipient shortfall apart from allocated setup, subscription, support, correction and bookkeeping effort. Keep cash and estimated time costs separate.
• Do not count an exchange-rate margin twice when a quoted total already includes it, and do not add fees again when they are already in the quote.
• Show both the marginal cost of this job and the fully allocated cost under a stated convention. Do not allocate an entire subscription to one transfer while assigning none of Wise's setup cost, and do not divide a one-time account-details fee by an invented lifetime payment count.
• For receiving, include payer charges, any merchant fees, the currency retained, the eventual conversion and access to usable funds.

Implementation gate

These are checks to run before real demand, not defects reported in Wise.

• Verify the live path: the euro funding and sterling delivery combination, the beneficiary details, and the arrival estimate for the actual supplier account. Do not assume every supplier account is reachable on identical terms.
• Confirm the live entry charges for the chosen business and the exact feature set. Recheck prices on the day of any publication that relies on them.
• Confirm the accounting configuration in a live walkthrough, including whether bill import or paid-status sync applies to the account.
• Allow time for verification. Allow time for Wise's normal verification process, including any additional information requests, before relying on the new route for a payment deadline. Wise's verification guide says further verification usually takes up to 10 working days. This is a planning allowance, not a prediction for any business.
• What the gate limits. It limits execution. It does not prevent publication of a clearly labelled proposal.

Measurement traps

• Clicks, accounts opened or balances funded: show interest or setup, not a paid obligation.
• Complete status: shows Wise has sent the money onward, not that the supplier has it.
• A feed or paid-status sync: shows data moving, not a matched ledger.
• Transfer volume: can rise because one large invoice arrives, because existing payments move provider or because a supplier splits payments. It does not show more businesses succeeding.
• Account balance: can be useful funds, an uncompleted job or idle money. Do not call it dependence without knowing the underlying obligation.
• Repeat payments: may reflect more supplier demand and not a better experience.
• Use only when the quote wins: can still be a successful customer. Full consolidation is not a necessary definition of value.
• Vendor-selected customer stories: selected, undenominated and not a baseline for another business.