Three habits that make leaving cheap
Export your history while you still have access, keep your prompts and instructions in your own notes rather than inside the product, and write a short page on what worked. All 3 cost minutes now and remove most of the cost of any future move.
The time to prepare for leaving is while you are happy. Every one of these habits is impossible to start once a product has announced it is closing.
Export while you can
Do it once now to confirm the export works and what it contains, then repeat occasionally. Export is the first feature to break during a wind-down and the last to be fixed, so an untested export is not a plan.
Keep prompts in your own notes
Anything you tuned belongs in a file you control. Prompts stored inside a product are lost with it, and they are the part hardest to reconstruct because you refined them gradually and never wrote down why.
Write down what worked
Half a page: what this product is good at, what it refuses, how you phrase things for it. On the next product it becomes a checklist of what to re-establish, and it turns weeks of rebuilding into hours.
Prefer plain formats
Text you can read without the product is portable by definition. Anything stored only in a proprietary structure is hostage to the product's continued existence, and choosing plain formats is the cheapest insurance available.
What people ask before switching
How often should I export?
Monthly is generous for most people. Once, ever, to check it works is the minimum and most people have not done it.
What if there is no export?
That is information about the product. Treat everything in it as temporary and keep anything important elsewhere.
Does this apply to a product with no account?
Differently: nothing accumulates to export, so keeping your own notes is the whole practice.
Test the export on whatever you use today.
Open the chat