THE QUESTION
Does ChatGPT actually save time writing professional emails?
THE TASK
Produce short, routine professional emails that are accurate, complete, appropriately toned, and ready to send after review.
THE TEST
One participant completed 10 manual email scenarios. Total manual time was 25:05, with an average of about 2:30.5 and a median of about 2:31.5. For three comparable AI-assisted scenarios, the participant reported timing from submitting the request through reading and checking a usable draft: 13, 16, and 20 seconds. Their average was about 16.3 seconds and median was 16 seconds.
The original ten timed manual emails were not saved. Three representative manual examples were written afterward and received provisional, non-blind quality scores of 19/25, 20/25, and 16/25. These separate examples cannot establish the quality of the ten timed outputs. The participant considered the AI-assisted drafts ready to send after review.
THE RESULT
Manual median: about 2 minutes 31.5 seconds. AI-assisted median: 16 seconds.
In this test, AI assistance reduced median time-to-usable-output by about 89%. Editing burden was not separately measured. Quality observations came from a separate, non-blind review, not a paired comparison of the timed outputs.
The exact percentage is a result from this experiment. It should not be generalized to every person, email, or workflow.
THE CATCH
The AI timing sample included only three trials, compared with ten manual trials. The scenarios were comparable rather than identical, the test involved one participant, and the quality evaluation was not blind. Times were reported by the participant; independent stopwatch logs and the exact model version were not retained. Scenario preparation was not separately timed. The finding applies to short, routine professional emails—not sensitive communications, complex negotiations, legal language, or high-context work.
THE VERDICT
USE IT.
Use it for drafting, not blind sending. Always verify facts, commitments, and context.