Writing in someone else’s voice is one of the less discussed skills in executive support, but it is one of the more demanding ones. When an executive assistant drafts a communication that will go out under the executive’s name, the output cannot sound like the assistant. It needs to sound like the executive — their cadence, their level of formality, their relationship to the recipient, their characteristic way of getting to the point. Getting that right takes time and deliberate attention.
Learning the Voice
Voice calibration starts with reading everything the executive writes directly. Emails, comments on documents, internal messages, presentations — each one is data about how they communicate. The patterns that emerge across those samples are what an assistant is trying to replicate. Sentence length, how they address people by name, how directly they state a request, whether they soften a no or deliver it cleanly — all of it goes into the model.
The fastest way to learn a voice is to draft something, have it corrected or rewritten by the executive, and then study exactly what changed. The edit reveals the preference more precisely than any style guide could. After enough rounds of this, the drafts require fewer corrections, and eventually the assistant can produce communication the executive sends largely as received.
Context Changes the Register
A good executive does not communicate the same way with every recipient, and a good executive assistant needs to understand those distinctions. A message to a direct report navigating a performance issue is different from a message to a board member requesting information, which is different again from a message to a peer at another company. The assistant needs to know not only how the executive communicates but how they communicate to each category of relationship.
This is where relationship knowledge becomes essential to communication quality. Drafting without understanding the relationship will produce something that reads correctly on the surface but misses the calibration underneath. The executive will recognize the miss even if they cannot articulate exactly what is wrong. Building enough relationship context to calibrate across recipients is a project that runs in parallel with building voice knowledge.
When the Draft Is Not Right
There will be drafts that do not land correctly. The response in that situation is to understand specifically what missed and to update the working model accordingly. An assistant who stops at “the executive prefers something different” without understanding what the preference is and why cannot improve. The correction is useful only if it gets internalized as a more accurate read of how the executive communicates.
Over time, the calibration becomes precise enough that the executive trusts the draft function enough to use it routinely. At that point, the executive’s communication capacity is expanded because responses can go out faster, with less executive time required at the drafting stage. That is a meaningful operational contribution — not just a clerical function.