Payment reconciliation is matching the money that arrives in your bank account to the quote, pro forma or invoice it was meant to pay. In most small businesses the slow part is not the matching. It is getting the bank's information in front of the person who needs it, because the only people who can see the account are the ones with the bank login.
A read-only bank feed inside your own system fixes that. Staff see incoming payments next to their own work, check a customer has paid without asking anyone, and never touch the bank itself. Here is how that works, what it is safe to allow, and what it does not do.
Why payment reconciliation stalls in a small business
The pattern is familiar. A customer is sent a pro forma, says they have paid, and wants their order to go out today. The account manager cannot see the bank, so they email the one or two people who can. Those people are busy, or out, or on holiday. The order waits, or it goes out on trust.
Sharing the bank login is the obvious shortcut and the wrong one. A login that can see payments can usually also make them, and the bank has no idea which person is using it. Most banks' terms forbid sharing security details in any case, which leaves you exposed if something goes wrong.
So the question is not really "how do we reconcile faster". It is "how do we let the right people see what has come in, without giving anyone the ability to send money out".
What a read-only bank feed is
A bank feed is a direct connection that pulls your transactions out of the bank automatically, so nobody has to download a statement. Accounting packages have offered bank feeds for years, but those feeds land in the accounts, which is not where an account manager or a sales team spends their day.
The version that helps the people chasing payments puts the same feed inside the system they already work in: the CRM, the order system, or whatever holds the quotes. We recently built exactly that for a UK business whose customers pay quotes and pro formas by bank transfer into a Starling business account. Their staff now open a page in their own system and see:
- every payment into the account, with the payer's name and the reference they typed
- the amount and when it settled
- the sending account details, when they need to tell two similarly named customers apart
- filters by date and amount, and an export to a spreadsheet for anyone who wants one
Before, checking a single payment meant finding someone with bank access. Now the person dealing with the customer checks it themselves, in seconds, while the customer is still on the phone.
Is it safe to connect your bank account to another system?
It is a fair question, and the answer depends entirely on what the connection is allowed to do. Done properly, it is safer than the alternatives people use today, which are shared logins, forwarded screenshots and statements emailed around the office.
Four things make the difference:
- Read-only permissions. The connection is given permission to read transactions and nothing else. It cannot make a payment, set up a payee or change anything in the account, because the bank never granted it that ability.
- The business creates the key. The access key is created by the account holder, in the bank's own developer settings, not by us. Nobody needs your banking password to set it up.
- The business can switch it off. Because the bank issued the key, the account holder can revoke it from the bank's side at any time, and the feed stops immediately. You are not relying on your software supplier to do it for you.
- Separate permissions in your own system. Seeing payments and managing the connection are two different permissions. Most staff get the first. One or two people get the second. Someone who can view payments cannot see or change the bank key.
Starling is one of a small number of UK banks that offer their own developer access for business customers, which made this one straightforward. With most other banks, the same read-only feed is possible through a regulated open banking provider, which works on the same principle: the business grants read access, for a limited purpose, and can withdraw it.
The small details that make a bank feed trustworthy
A feed people rely on has to get the dull parts right. Two examples from this build.
Customers write the payment reference, not you. Whatever someone types into their banking app arrives in your system as-is. A spreadsheet will treat a reference beginning with an equals sign as a formula, so a careless or malicious reference could turn into something that runs when a member of staff opens the export. The export writes every reference as plain text, so it can only ever be read, never run.
The connection should not quietly expire. Bank keys come in different forms, and some need renewing. The page shows plainly whether the connection is healthy and whether it will need attention, so nobody discovers it stopped working three weeks ago when a customer complains.
What this does not do
It is worth being clear about the limits, so nobody buys it expecting something else:
- It does not replace bank reconciliation in your accounts. Your accountant or bookkeeper still reconciles the books in Sage, Xero or QuickBooks, usually from their own bank feed. This is for the people who need to know a customer has paid, not for closing the month.
- It does not match payments automatically on its own. In this build staff match by eye, using the payer and the reference. Automatic payment reconciliation, where each payment is suggested against the open quote or pro forma it most likely pays, is a natural next step and a separate piece of work.
- It only sees the accounts you connect. A business that takes payments into more than one bank needs a connection for each.
- It shows what the bank shows. If a customer pays without a reference, the feed cannot invent one. It just makes that payment much quicker to find.
Is it worth doing for your business?
It tends to pay for itself where customers pay by bank transfer before you do the work or release the goods, and where the people dealing with those customers are not the people with bank access. Trade suppliers, manufacturers taking deposits, and service businesses invoicing up front all fit.
If every payment comes in by card through a payment provider, you probably already have this through the provider's dashboard. If you have a handful of bank transfers a month, a shared statement may be enough.
The test is simple: count how often this week someone in your team asked "has this customer paid?" and had to wait for the answer. If it is more than you would like, the bank already has the answer. It just needs to be in the right place.
We build connections like this into existing systems as part of our bespoke business software work, alongside integrations with accounting and payment systems such as Sage, QuickBooks and Stripe. If staff in your business are waiting on someone else to check the bank, book a call and we will tell you honestly whether it is worth doing.
