ScreenVault / Guide

How to organize screenshots

A system that survives a busy sprint: name once, file once, tag for the rest, and clean up weekly. Written for people who take dozens of screenshots a day.

1. Decide what a screenshot is for when you take it

Most screenshots are one of four things: evidence (a bug, a failed test, a customer’s error), reference (a design you like, a setting you’ll need again), communication (something to paste into a chat right now), or record (a receipt, a confirmation). Communication shots rarely need keeping. The other three do, and they need a name.

2. Name it in words you’ll search for

The OS gives you Screenshot 2026-09-14 at 10.42.17.png. Replace it with the words future-you will type into a search box:

RolePatternExample
QA / supportarea-what-happens-platformcheckout-coupon-500-safari18
Developerservice-error-contextauth-token-expired-staging
Designerproduct-screen-state-versionpricing-hero-mobile-v3
Everyonedate first, only if time matters2026-09-14-invoice-aws

Keep ticket numbers and people’s names out of the title and in a note or tag; they change, and they make titles unreadable.

3. One collection, several tags

Give every screenshot exactly one home: a collection for the project, release or client, nested as deep as helps (QA / Sprint 42 / Safari). Then use tags for the other ways you’ll slice it: type (#bug, #reference), severity (#p1), platform (#ios), status (#fixed). Collections stop the heap; tags stop the duplication. Three or four tags per screenshot is plenty; a tag used once is noise.

4. Make the pixels searchable

The most useful information in a screenshot (an error code, an order number, a line of copy) is locked inside the image. Extract it once and it becomes searchable forever. In ScreenVault, open a screenshot and press “Read text & suggest tags”; the screenshot to text tool does the same for a one-off.

5. A QA evidence workflow

  1. Capture the defect at a known viewport (from the browser or with the OS shortcut and paste).
  2. Mark it up: one box on the defect, numbered steps, and solid redaction over customer data.
  3. File it under the build, tag severity and platform.
  4. Compare against the last good build with a pixel diff and attach the diff.
  5. Close the loop: when the fix ships, capture again, tag both #fixed, and keep the pair as the regression baseline.

Frame-by-frame evidence is a small piece of storytelling: each capture should show one step, in order, so the reader never has to guess what changed between them. Want to see a sequence carried by narration done well? ahaboo takes why the Moon shows phases through narrated steps, and it’s a useful model before your next multi-screenshot bug report.

6. Clean up weekly, back up monthly

Once a week, filter to Unfiled and Untagged, give each survivor a home, and delete communication shots that did their job. Once a month, export a ZIP backup: a vault that lives in one browser should never be the only copy of evidence you may need later.

7. Share without scattering copies

Every upload to a screenshot-hosting service is another copy you can’t delete later. When the audience is known, send the file itself: a ZIP, a PDF, or ScreenVault’s single-file HTML gallery, which opens in any browser with no server. Before you send, strip metadata from photos and shrink oversized captures.

Questions

What is the best way to organize screenshots?

Rename at capture time with a short descriptive title, file each screenshot in one collection by project or release, and add two to four tags for the other axes you search by (type, severity, platform). Review the inbox weekly and delete anything that no longer matters.

Folders or tags?

Both. Folders (collections) answer “where does this belong?” and a screenshot should have exactly one. Tags answer “what is this?” and a screenshot can have several. Using only folders forces you to duplicate files; using only tags leaves you with a flat heap.

How should QA teams name screenshots?

A pattern such as <area>-<what happens>-<platform>, for example “checkout-coupon-500-safari18”, sorts sensibly, reads well in a ticket and contains the words people will search for. Put the ticket number in the note or a tag rather than the title.

How do I find a screenshot by the text in it?

Run text extraction on it once. In ScreenVault, “Read text & suggest tags” stores the text next to the image so the search box matches it. Phones do this natively: iOS Photos and Google Photos both search text found in images.

Other screenshot tools