Colquhoun.AI
← All guides

Research Guides

Tell Readers When a Research Export Is Ready

· Colquhoun.shop editorial

An export button can change to a reassuring check mark while the reader still has no usable file. Review a research export as a sequence of observable states, not a single click. This worksheet is for teams testing their own tools; it is not a claim that Colquhoun.shop provides or has certified an export service.

1. Define what each message actually proves

Write down four distinct states: request accepted, file being prepared, file available and request failed. A server accepting a job is not evidence that a file exists. A download link appearing is not evidence that the browser saved a readable copy. Use wording that stops at the fact the application can observe.

For an invented competitor-research export, a useful sequence might be “Preparing the selected 12 rows” followed by “CSV ready: 12 rows, current filters.” Avoid “Saved to your computer” unless the implementation can establish that result. Specify whether the export includes visible rows, all matching rows or the full dataset.

2. Review the status without moving the reader away

W3C explains that status messages in markup should be programmatically identifiable so assistive technologies can present them without taking focus. Its examples distinguish a result list from a brief message announcing the search result count. The guidance also warns against making interfaces too chatty. W3C: Understanding Status Messages

For your implementation, record both the visible message and what a screen reader announces. Check the surrounding context: hearing only “ready” may not tell the reader which export finished. Do not assume a check mark, changing button label or a particular ARIA attribute proves the whole flow works. Test actual behavior with the browser and assistive technology combinations you intend to support.

3. Exercise more than the happy path

In a safe test environment, try an empty selection, a deliberately delayed response and a controlled failure. Change filters while a job is preparing. Confirm that a completed message still identifies the selection that produced that file, rather than silently describing the new view. If a second request is allowed, ensure the two results cannot be confused.

Decide how retries behave before testing them. Does retry create another job or resume one? Is a stale download link still shown after failure? The acceptance record should state the intended behavior and the observed result, not infer success from the absence of an error banner.

4. Inspect the delivered evidence

Download the test file and open it in the intended reader. Compare its row count, identifiers, dates and source fields with the selection recorded before export. Use invented test data rather than personal or confidential records. Check an empty result explicitly so a header-only file is not mistaken for a populated report.

For the structure of the file itself, use the companion guide to exporting a traffic CSV someone else can interpret. Status clarity and file correctness are separate checks; neither substitutes for the other.

5. Keep a short acceptance record

This focused check is not an accessibility certification or a security audit. Its practical goal is smaller: when a research tool says an export is ready, the reader knows which result is ready and can obtain the evidence that message promises.

This guide describes the built-in Web Reader and Keyword Density behavior verified for this publication. Available data depends on the page response; these tools do not measure visits or search rankings.