The exit plan to write before you sign with an online business bank

Leaving a branchless provider is cheap when you planned for it and expensive when you did not. The difference is a page of written answers, collected before the first deposit: who holds the balance, how you would be told of a problem, and how the money and the records would leave.

What to take away

The holder of the balance shapes your exit

Which institution holds your money decides what leaving looks like. Some providers are banks. Some are payment firms that hold customer funds at a bank. Some products marketed for business cash are not deposits at all. Ask which applies, and how the account is titled.

Deposit insurance exists at insured banks and insured credit unions, but limits and conditions apply and depend on titling. Take that question to the deposit insurance rules the FDIC publishes for banks, or the share insurance material the NCUA publishes for credit unions, not to the provider's marketing page.

A credit union can be a realistic alternative for some businesses, subject to eligibility, and its service model can be closer to what people miss about branches. Insured credit unions are covered by the NCUA rather than the FDIC. Check who can join first, because membership comes before anything else.

What happens if the provider stops serving you

Providers narrow the industries they accept, exit segments, get acquired and occasionally fail. That is no reason to avoid the model, but it is a reason to know how you would find out, how long you would have, and what happens to a balance meanwhile.

Ask each candidate in writing, and keep the replies. None of these questions is hostile, and a well-run provider answers them without difficulty. The value is partly in the answers and partly in observing which candidates find the questions unreasonable. A provider that will not answer has told you something.

Availability history deserves a question but not a verdict. Every provider has bad days. What separates them is what the business can still do on those days, which is a question about your own redundancy more than about their record.

The five-factor exit plan

Work out how you would leave and write it down before you commit. It takes twenty minutes and changes how you feel about the rest of the comparison. Fill one row per factor, and treat every blank as a question for the provider.

The contract row lives in the fee schedule and the account agreement, so request both before you apply. The documents to request are listed on a separate page, together with what to do when one is withheld.

A provider you can leave in a fortnight with a clean data export is a much smaller risk than one that is slightly better on paper and difficult to escape. The exit terms are the price of being wrong.

Run the same plan on any provider you already use. If you cannot answer these for your current account, that is a finding worth acting on before you shop for a new one. The failure modes it guards against appear in the nine errors that cost most.

Where feature lists fit

Feature lists eliminate candidates well and choose between survivors badly. Features get copied, pricing moves, and a demo interface fades within a month. Use the list to check hard requirements only, then let the exit plan separate whoever is left.

The hard requirements are practical. Does the data come out in a form your bookkeeping accepts? Does the provider serve your entity type and industry, and support every payment type you depend on? Does it offer genuine approval controls rather than a second login, and can cash be handled in a way you could live with? Any candidate failing one is out.

What a list cannot rank is behavior: how a stuck payment is handled, how a review is communicated, whether support has authority. Those decide the relationship, and none of them appear in a table. The page on what a branch quietly does shows where a conventional relationship still holds an advantage.

Testing the exit on a small balance

Test the plan before you need it. Open the account, keep it non-critical, and spend one cycle collecting facts rather than impressions. Export the history, move a small balance out in full, and note how long each step takes against what the provider told you.

Write results down as you go, then compare candidates on identical tests. This is slower than reading a ranking, and it is the only method that produces evidence about your business rather than someone else's. Comparing like with like also shows which provider answered your written questions accurately.

Where this leaves you

Eliminate on hard requirements, qualify on who holds the money, price the exit before you enter, and buy real evidence with one trial cycle. A comparison built that way survives the year in a way that a ranking published this month will not.

Common questions

Should I prefer a provider that is itself a bank?

Not automatically, but know which you are dealing with, because it changes what protects the balance and who you are actually a customer of.

Some businesses run happily on a payment provider for operations while holding reserves at an insured institution, which is a reasonable structure as long as it is deliberate.

What if the best fit is also the smallest provider?

Smaller is not automatically riskier, but it does make the continuity questions more important and the exit plan more valuable. Ask the same questions, insist on written answers, and keep a second relationship live so that one provider's change of policy never stops the business.

How long does the exit plan take to write?

Twenty minutes is enough to fill the five rows for one provider. List what is wired to the account first: payers, processors, direct payments and cards. Anything you cannot fill in becomes a question to send the provider in writing, and you keep the reply.

Can a feature list settle the choice?

Not between survivors. A feature list is good at eliminating candidates and bad at choosing among them. Once every remaining candidate meets your hard requirements, the written answers and the exit plan are what separate them.