--
title: "Why Your Journaling App Should Export Everything, and How to Build It Right"
description: "Why Your Journaling App Should Export Everything, and How to Build It Right"
tags: ["privacy", "journaling", "memory", "mentalhealth"]
canonical_url: https://stillherememory.com/blog/export-your-words-always
Two years ago I switched note apps after four years and 2,000-plus notes, and the export was a disaster. Format errors, lost entries, a full weekend burned. The lesson I took into building a journaling app: export is not a feature, it's the floor. If users can't take their words out, they don't own them.
The design principle is that export must be lossless and boring. Lossless means the output carries everything the app stores: the content, the timestamps, the tags, the links between entries. Boring means it's one click and always available, no request process, no premium tier. If export is hidden behind anything, users correctly assume their data is being held.
Format selection is where most apps get this wrong. One format isn't enough because export serves different jobs, so we ship three.
Markdown is the general-purpose choice. It's readable as plain text, which means it opens cleanly in any editor, and it's portable, which means almost every notes app, blog, and writing tool can import it. If a user plans to keep writing somewhere else, Markdown is the answer.
JSON is the structure-preserving choice. Timestamps, tags, and relations stay intact as structured data. This matters for technical users who want to migrate in bulk or build their own backup system on top of the export.
Plain text is the safety net. Zero format dependency, so it opens a hundred years from now if anything does. It's the answer for users who care about one thing: the words surviving.
The three formats aren't a display of engineering. They're the same promise delivered three ways, give the user back what they wrote, shaped the way they need it.
What users actually do with exports is wider than most teams assume. Local backups, exporting Markdown monthly to a personal drive, so the "cloud-only" assumption goes away. Migration, switching tools without gaps. Printing, turning a year of entries into a PDF and binding it. Review, feeding the JSON into their own script to count words or find when they write most. The point is the export has to be a complete data dump, not a rendered view.
Two implementation details matter. The export must include everything without sanitizing or cherry-picking, a partial export is worse than none because it teaches users not to trust the button. And it has to be free forever, no subscription, no request. Charging for export converts a right into a ransom.
There's also a policy dimension. GDPR frames data portability as a right, which is the correct baseline for any app that stores user writing. And the promise should extend to model training: user data is not training data. That commitment belongs in the export terms, not buried in a privacy policy.
If your app stores user writing and doesn't have one-click full export, that's not a roadmap item. It's the line between holding data and holding memories hostage. The full writeup on our export design: https://stillherememory.com/blog/export-everything-you-wrote
