It comes up in almost every accounting software decision. The accountant recommends Sage because their practice runs on it. The finance manager who has used Xero wants Xero. Both are arguing from real experience and neither is being unreasonable.
Take the disagreement seriously
The temptation is to treat this as a preference to be overruled. It is not. Your accountant working in an unfamiliar system will be slower, will charge more, and will find fewer of the things you pay them to find. Your finance manager working in a system they find obstructive will do less of the discretionary work that makes the numbers useful.
Both costs are real. The question is which is larger in your specific business.
The test that usually resolves it
Ask two questions.
First: how much of the monthly work happens inside the system, and who does it? If your finance manager is in the ledger every day and your accountant looks at it quarterly, weight the person who lives there. If you have no internal finance function and your accountant does the bookkeeping, weight the accountant.
Second: how much would your accountant charge to work in the alternative? Ask them directly and get a number. Frequently the answer is that the difference is small, and the disagreement dissolves. Occasionally the answer is that they will not take the work, which is also useful information.
When the accountant should win
Where the compliance requirement is genuinely different. A practice filing forty VAT201 returns benefits materially from a system that transfers to eFiling, and that is not a preference, it is a workflow. Where you are audited, an auditor familiar with the system asks fewer questions and charges for fewer hours.
When the team should win
Where the daily experience differs sharply and the compliance outcomes do not. Sage Accounting and Xero produce equally correct VAT201 figures. One transfers to eFiling and one does not. If you file six returns a year yourself, that difference is ninety minutes annually, and it should not outweigh a system your team will actually use well.
The compromise that works
Trial both, with your own data, for a full month, with both parties using them. Not a demo. A real month, with real transactions and a real month end. Then make the decision with evidence instead of with preference.
It costs a month and it settles the argument permanently, which is worth more than the month.