The renewal invoice from your property management software arrived Tuesday, and you approved it at 9:15 AM the way people sign for a package they did not order: without ceremony, gone from the day in under a minute. It is the third year running you have paid for software you would not recommend to another operator.
Nothing in that is irrational. You know exactly why you stayed, and it is not the software. It is what the software is holding: five years of ledgers, every deposit you owe and to whom, the rent history behind every renewal decision, the paper trail from the one dispute that went to a hearing. The product you would happily leave and the records you cannot lose live behind the same login, and the vendor's best retention feature is that you stopped being able to tell them apart.
So the cost of staying stops being itemized. The workaround your team runs because a screen never got fixed is just "how we do it now." The new hire learns the workaround in their first week, from a veteran who no longer remembers it is one. The renewal price drifts up a little each year, and each year the same math happens in the same half-second: whatever this costs, moving costs more. That math never gets checked. It is the only invoice you approve annually without once pricing the alternative.
There is a simple test for whether the books are yours: what happens to them when you leave. The portfolio passes it; you can sell it, refinance it, or manage it from a legal pad if you have to. The books pass it in a much weaker sense if the only working copy lives inside a subscription, in a format of someone else's choosing, behind an export you have never actually run. You suspect this already, which is why in five years you have never tested that export. Not because it would fail. Because it might.
Your books are only yours if you can take them with you.
What changes the decision is not braver nerves; it is seeing that the move has a shape. An exit done right is a procedure with a proof step: export everything while access lasts, close one final month properly, cut over at a month-end, land the new ledger on the old one's closing balances, and compare a trial balance from each side until they agree to the cent. The fear was never really about losing data. It was about having no way to prove nothing was lost. A matching trial balance is that proof, and once you know it exists, "can we survive the move" quietly becomes "which month-end."
The other thing that changes is what you look for in the next system, because you only make this mistake once. The question that matters is not what the software holds; everything holds everything. It is how your records get in, and how they would get out. Scaalr's answer on the way in is a getting-started checklist that ticks itself off as the portfolio takes shape, plus CSV import for journal entries and bank statements, with each batch previewed and validated before it saves and an opening bank balance seeded so the first reconciliation starts clean; there is no bulk portfolio import, and that boundary is stated before you sign up, not discovered after. The whole procedure, export list included, is in How to Switch Property Management Software Without Losing Your Books.
For the complete playbook, what to export, the month-end cutover, and the four checks that prove the landing, see: How to Switch Property Management Software Without Losing Your Books.
The first 5 units are free, then $0.99 a unit a month. Worst case, your books got organized.