Three people edited the same file, and the newest version lost the best work

Three of us were editing a grant application. The shared document was supposed to prevent version confusion. Instead, we created “final,” “final with edits,” and “final use this one” across email, downloads, and the shared folder.

The newest file had the updated budget. It was missing two paragraphs that our program manager had written the day before. Another copy had those paragraphs but an older deadline. Nobody had intentionally deleted good work. We had each edited a different object while believing we were in the same process.

The dangerous temptation was to pick the file with the latest timestamp and call it authoritative. A later save does not prove that it contains every earlier decision, especially when attachments and offline copies have been moving in parallel.

We stopped editing and made read-only copies of every contender. Then we created a comparison table for the consequential fields: deadline, requested amount, program description, commitments, and named owners. Once we had one reconciled version, we moved it to a single shared location and wrote an explicit rule that email attachments were for review, not independent editing.

I also want the document to name an owner for the final merge. “Everyone can edit” had quietly become “nobody knows who decides which version wins.”

The technology did not create the confusion by itself. Our workflow allowed several truths to exist and trusted filenames to settle the dispute.

How do you recover safely from a version collision without losing the strongest material? And what small team rule prevents shared editing from turning back into attachment archaeology under deadline pressure?