Swiss retail banking · Analysis · September 2026
Revolut applied for a full Swiss banking licence on 16 September 2026. FINMA has not named a timeline, and neither has Revolut. Whether approval takes twelve months or twenty-four, the waiting time is the real opportunity: what Revolut prepares now decides how many of its Swiss customers move their salary on day one after approval, rather than years later. The same months are the only window the incumbents get.
Both scenarios keep Revolut growing at its current pace, about 240,000 new Swiss customers a year. They differ in one thing: whether the salary follows. In the first, the licence arrives and main accounts are converted the way Swiss customers switch banks on their own, slowly. In the second, Revolut has scored its base before approval, knows which customers face the lowest barrier, and reaches them with a fitting offer from day one.
Today Revolut serves Switzerland through its Lithuanian bank. A Swiss licence brings a Swiss IBAN, a salary account, eBill, Swiss deposit protection and, later, Pillar 3a and Twint.1 Revolut's Swiss country head described the current position honestly last year: for many users, Revolut is "more of a secondary or tertiary bank".5
of Swiss residents switched their main bank at least once in ten years. The majority stay, regardless of conditions.4
of retail customers plan to switch their main bank in a given year, and about 5% more are considering it, in the Lucerne IFZ survey.6
is how far neobanks got as main bank in the UK between 2019 and 2025. The IFZ expects Switzerland to reach about 5% within five years.3
Those numbers describe the whole market. Revolut's position is better than the average, because its customers have already opened an account, verified themselves and formed a habit. But the same numbers say that a launch campaign aimed at everyone will convert at the market's pace. The question worth answering before approval is which customers face a low barrier, and what offer fits each of them.
Revolut's customer data is not public, so this analysis runs on a simulated Swiss retail market built by Mountain Lion Analytica: 10,000 synthetic residents fitted to federal statistics, with salaries, rent, spending, life events and, crucially, the side Revolut cannot see: the products and balances at their existing bank. Revolut accounts are assigned by age from SNB neobank survey rates. 1,693 of the 6,506 simulated adults hold one.
| Products at existing bank | Lower-barrier | Anchored | Share lower-barrier |
|---|---|---|---|
| 2 | 580 | 7 | 99% |
| 3 | 93 | 689 | 12% |
| 4 | 5 | 280 | 2% |
| 5 | 0 | 37 | 0% |
| 6 | 0 | 2 | 0% |
One illustrative synthetic customer for each. Names, addresses and figures are what the simulation generated. The hypotheses are the ones a launch team would want to test in the first weeks after approval.

Little holds him at the existing bank: two products, a modest balance, most card activity already visible to Revolut. The offer that fits is the one that removes effort.

One retirement product ties her to the existing bank while wallet share and balances say she is ready to move. The offer that fits does not ask her to close anything.

Heavy card use with Revolut, yet the existing bank holds five products and most of the wallet. A salary pitch competes with all of that; a second product does not.
None of this needs the licence. All of it needs the time before it, and the same signals serve whoever reads them first.
RevolutAge, tenure, inflows, spend mix, travel share and canton rank customers by how likely their salary is to move. In the simulation, a model that sees only Revolut-side signals recovers most of the lower-barrier label out of sample.
IncumbentThe mirror image. Two products, a modest balance, a monthly transfer to a Lithuanian IBAN and falling card activity rank customers by how likely their salary is to leave.
RevolutLower barrier: remove the switching effort. Pillar 3a as the only tie: make coexistence explicit. Anchored: lead with a second product, not the salary.
IncumbentThe exposed two-product customer needs a third product that fits, offered before someone else offers a salary account. A retention discount afterwards is the expensive version.
RevolutAdoption, income and the mix of incumbents differ by canton. A ranked launch order with expected main accounts per canton can be finished long before approval.
IncumbentA cantonal bank has one canton and knows it street by street: which municipalities and age bands adopt neobanks first is where the first outreach goes.
RevolutSalary-first versus 3a-first, pricing, canton order: scenarios that cannot be A/B tested on real customers can be run on a simulated base at full scale.
IncumbentThe same twin answers the reverse question: which offer keeps the most salaries, at what cost, and which customers were never going to leave.
RevolutEvery campaign after approval reports main accounts gained per segment and canton, against a control. That plumbing is built before the launch.
IncumbentSalary outflows per segment, monthly, from the day the licence was announced, not the day it is granted. The baseline has to exist before the change.
RevolutReal conversions replace the simulation's labels within months. The model gets better after every campaign, and it belongs to Revolut's own data team.
IncumbentThe same, owned by the bank's own team. The model that predicts who leaves is the model that predicts who moves, trained on the bank's side of the relationship.
Every signal that tells Revolut who might move tells UBS, PostFinance, Raiffeisen and the cantonal banks who might leave. The exposed customer is not the one with the mortgage and the 3a. It is the one with two products, a modest balance and most of their card activity happening somewhere the bank cannot see.
Declining card activity, an outgoing transfer to a Lithuanian IBAN every month, a young customer with a salary account and nothing else. In the simulation, 99% of two-product customers carry the lower-barrier label; among four-product customers it is 2%.
The same scoring, in reverse: find the two-product customers with high external activity, and give them a reason to hold a third product before someone else offers them a salary account. The window is the licensing period, and it is open to both sides for exactly the same months.
Works with the banks that have to answer. The simulation behind this article is his own: 10,000 synthetic Swiss residents, 2.4 million transactions, validated against federal statistics.