I Tried to Walk Out on 10 Popular Apps With My Data. Five Slammed the Door.

Every app swears the relationship is open. You can leave anytime. Your data is yours. Then you actually try to leave, and suddenly you're downloading a 400MB XML file at 11 p.m. wondering why your health history arrives as machine gibberish.

So I ran the test. Ten apps I actually use — notes, tasks, health, money, music, reading — and one question per app: if I quit today, how cleanly can I walk out with my stuff? I timed the whole thing from the moment I hit export to the moment I had a file that opened in something that wasn't the original app.

Here's what came back.

How I graded the breakup

Three scores per app, one to five each.

• Speed. Wall-clock time from clicking export to a usable file. Under 10 minutes is a five. Anything measured in "we'll email you" is not.
• Format quality. Did I get something a human or a spreadsheet could read? Were relationships preserved, or did my nested folders arrive as one flat pile?
• Lock-in friction. Hidden obstacles: attachment flattening, metadata stripped, exports that technically exist but require a support ticket.

A clean breakup needed at least 12 of 15. Half the field didn't get there.

The five that let me go

Notion handed over a zip with HTML and CSV in about four minutes. Nested pages kept their structure. Databases came out as spreadsheets I could actually read. The catch is that HTML isn't a great long-term archive, but it's honest.

Todoist was the fastest of the bunch — under a minute for CSV and JSON. Projects, labels, due dates, all intact. No drama.

Google Keep required Takeout, which took about three hours to land in my inbox, but the output was clean and complete. Slow, not sticky.

YNAB exported transaction history as CSV instantly. Budget structure needed manual reconstruction, but the raw data was pristine.

Apple Health gets points for completeness — the XML export is genuinely everything, years of records — and loses points for producing a multi-hundred-megabyte file that no normal person can open without a script. Complete, unreadable. I gave it a grudging pass.

The messy middle

Strava eventually gave me my full history, but bulk exports crawled, and the GPS and metadata quirks meant a few years of rides needed cleanup before they were useful anywhere else. Nothing hostile — just friction.

Apple Notes exported folders as HTML with attachments in a tangle of subfolders. Everything was technically there. Reassembling it in another app was an afternoon I'll never get back.

Spotify will hand you playlist data as JSON, which is great if you're a developer and useless if you wanted your library to mean something. Your listening history doesn't really come along for the ride.

The two that made me work for it

Evernote exported everything, and then flattened attachments in a way that turned a decade of PDFs into a scavenger hunt. The data left. The context didn't.

Instapaper gave me a CSV of URLs, which is a list of places my saved articles used to live, not the articles themselves. That's an export in the legal sense.

The thing nobody tells you: export isn't exit

Here's the pattern that jumped out. The apps that let me leave cleanly all had one trait in common: they stored my data in shapes that existed before the app did. Plain text, CSV, standard image files. The apps that fought me all invented their own container and then built the export as an afterthought.

There's also a second layer almost nobody tests: deletion. Exporting is the fun half of the breakup. Actually purging your data — and confirming it's gone — is the half where support tickets go to die. I'm not naming names, but two of the apps above still had my account sitting there weeks later, quietly holding onto everything I'd just downloaded.

And one more: subscription math. If you've pre-paid annually, leaving in month three is a financial decision, not just a technical one. The exit ramp exists; the refund usually doesn't.

A pre-commitment checklist for your next app

Before you pay for anything — or before you move your whole workflow into it — run this:

1. Find the export button before you find the pricing page. If it's buried behind a help article and a support form, that's the answer.
2. Check the format, not the promise. "You can export your data" means nothing if the file only reopens in the same app.
3. Do a test export in week one. Ten files, not ten thousand. You'll learn more in five minutes than from any marketing page.
4. Ask whether relationships survive. Tags, links, folders, history — those are the parts apps quietly drop.
5. Do the attachment math. Photos, PDFs, and audio are where exports go to die.
6. Confirm deletion works. Export with no delete is a one-way door with a welcome mat.
7. Look for an API or webhooks. A public API is a permanent escape hatch; a proprietary sync protocol is a moat.
8. Write the exit cost into the decision. Not just dollars — hours.

None of this is about being cynical. It's about treating portability the way you treat uptime or price: as a feature you check before you commit, not after you're stranded.

The best tool in your stack is the one you could leave tomorrow and still keep everything you put into it. That's not a threat to the app. It's the reason to trust it.