The service holding two years of my data is shutting down, how do people avoid this, and can I get it out?

For choosing the next one, the questions worth asking in order:

Does it export, and what? Look for a documented export before buying, not a promise of one. Check whether it is raw or summarised.

Is there a documented API? If a third party can read your data with your permission, you are much less trapped.

Does it sync to a platform health store? If the data also lands somewhere you control, the company's app becomes one view of your data rather than the only one.

Are the formats standard? Activity files in the common interchange formats are readable by everything. A proprietary blob is not.

Does anything work without the cloud? Some devices are useful offline; some are ornaments the moment the servers stop.

Is a subscription required to see your own history? This is increasingly common and is worth knowing before rather than after.

The honest summary of the category: the sports-focused brands tend to be better on export and standard formats, and the lifestyle and ring products tend to be worse, because their value is the interpretation rather than the numbers. That is a trade-off rather than a verdict: but if portability matters to you it is a real difference and it is visible before you buy.

Do the export properly now, before the deadline, and treat the file you already have as a first attempt rather than the answer.

Check what is actually in it. Exports vary enormously, some give you every raw sample, some give you daily summaries, and some give you a subset with the interesting parts missing. Open it, find one day you remember, and see whether the detail is there.

Look for a second export route. Many services have both an in-app export and a separate account-data request, and the second is often far more complete. Also check whether there is an API you can pull from while the service is still running.

Export more than once, and keep the raw files untouched somewhere safe. You can always parse them later; you cannot re-export after the shutdown.

Check whether it also synced anywhere else. A lot of these push a copy into a platform health service or a third party training site. If it did, that copy is yours already and may be more useful.

On making sense of the file: for most of these formats there are open source tools that convert them into something readable. Worth searching for the specific service name plus export, you are very unlikely to be the first person doing this in the month a service announces its closure.

One distinction worth holding onto, because it decides how bad a shutdown is: what is on the device versus what only exists in the cloud.

Most of these devices store a few days locally and immediately upload. So the historical archive is not on your wrist, it is on their servers, and your access to it is the app. When the app goes, the device usually keeps telling the time and counting steps and loses everything that made it interesting.

Which means the export is genuinely your only copy, and the deadline is real.

Going forward, a habit that costs nothing: export once a quarter and keep the files. Put a recurring reminder in. It takes minutes and it means a shutdown, an account problem or a switch is an inconvenience rather than a loss.

And on the interpretation, since that is what people actually miss when they move: the numbers are portable and the analysis is not. Sleep scores, readiness figures and training load are each company's own formula, and they do not transfer or compare between brands. Keep the raw measurements and accept that the scores end with the service, trying to reconstruct them is the part not worth your time.

A documented export in an open format is worth more than any feature on the box.