A spoken method for turning a role brief into a clear description.

Dictating a Prompt for a job description works when the brief is short and specific. Say the role title, the team, the level, the main outcome the person will own, and the three problems the role exists to solve. Add the location pattern and the tools the team already uses. This brief gives the assistant a target and gives you a standard to check the draft against. A description written from a title alone tends to list duties that could belong to any team. Write the brief on a note, then read it before you dictate, so the draft starts from facts rather than assumptions.
Name the audience for the description. A posting read by a candidate, a brief read by a hiring panel, and a summary read by a finance approver need different lengths and levels of detail. Say which one you are writing. If the same text serves two audiences, write the posting first and cut a shorter version from it, rather than trying to satisfy both at once. A description that tries to please everyone usually reads as a committee document, which is the last thing a candidate wants to read. One audience per document, and write the shortest version last, because it is the hardest cut.
State the outcome the role owns, not the tasks it performs. A role that owns onboarding completion is clearer than a role that attends meetings and updates trackers. Ask for the draft to lead with the outcome and use tasks as evidence. This shift makes the description easier to read and easier to interview against, because every task can be tested for whether it supports the outcome. If you cannot name the outcome, the role may not be ready to post, and the brief is where to settle that before an ad goes out. Write the outcome first, and everything else supports it.
List the constraints that shape the role honestly. Say whether the role is new or a replacement, what the team size is, and which decisions the person can make alone. These facts prevent a description that promises autonomy the role does not have. A candidate who joins on a promise the role cannot keep leaves quickly. Honest constraints make the posting less exciting and the hire more durable. Keep the constraints in one short block, so they are easy to find and hard to overlook when the draft is reviewed. Write them in plain language, be specific about limits, and avoid vague phrases.
Ask for responsibilities written as outcomes, not as activities. A line that says own the weekly release process end to end tells a candidate more than a line that says participate in release meetings. When you dictate the responsibilities, say the outcome for each one and let the assistant phrase it. If the result reads like a task list, ask for a rewrite around results. The difference is what separates a useful posting from a generic one. Read each line and ask whether it describes a result or an activity, then rewrite the activities before the draft goes further. This check takes five minutes.
Keep the list short. Five or six outcomes cover most roles, and a longer list dilutes the important ones. When you have ten, group them and ask which three matter most in the first six months. The first-six-months view is more useful to a candidate than a full inventory of duties. A short, ranked list also gives the hiring panel a clear basis for interview questions. Put the top three first, so a reader who stops after one minute still sees what matters most. Move the rest to a second section and label it additional scope. That keeps the main list strong.
Say how each outcome will be judged. A candidate wants to know what good looks like, and a hiring panel needs the same standard. For each outcome, name one signal, such as a metric, a deliverable, or a decision the person is trusted to make. A description with signals is easier to interview against and easier to explain to a candidate who asks what success means in the role. If a signal cannot be named, the outcome may be too vague to hire for, so sharpen it before the posting is approved. One signal per outcome is enough. Write each signal beside its outcome.
Read the responsibilities back and remove the duplicates. Spoken lists often repeat the same idea in different words, and a duplicate reads as padding. If two lines describe the same outcome, keep the clearer one. If a line describes a task rather than a result, rewrite it or cut it. A tight list of outcomes is more persuasive than a long list of activities, because it shows what the role actually changes. Do the cut before the review, so the hiring manager spends time on substance rather than on cleanup. Read the list aloud once, then cut again. Two passes usually catch the rest.
Split the requirements into must-haves and nice-to-haves before you dictate. A must-have is something you would reject a strong candidate for lacking; a nice-to-have is something you would prefer. This test is stricter than it sounds, and most lists fail it. Say the two groups separately in the prompt, so the draft does not blur them into one long list. A clear split helps candidates self-select and helps the panel focus interviews. Write the test at the top of the requirements section, so the rule is visible to whoever edits the draft later. Apply it to every item.
Keep the must-have list short and defensible. Every item should be tied to the outcome the role owns. If a requirement cannot be connected to an outcome, it is probably a preference or a habit. When the list grows past five or six items, look for a requirement that is really a development path rather than a hiring bar. A short list of real requirements attracts more qualified candidates than a long list of wishes. Say which requirement is the absolute minimum, so the panel knows where it cannot bend. Name that one first, because it sets the bar.
Describe nice-to-haves as advantages, not as hidden requirements. A phrase such as experience with a second reporting tool is a plus makes the status clear. If a nice-to-have would actually decide the hire, move it to the must-have list and say why. Candidates read a long nice-to-have list as a set of demands, and some strong applicants will not apply. Honest labeling is both fairer and more effective. Keep the list to three or four items, so the distinction stays useful instead of collapsing into a second requirements list. Fewer items is clearer.
Ask the draft to show how each requirement will be assessed. A line that pairs a requirement with an interview question or a work sample is easier to defend than a bare list. When a requirement has no assessment, either add one or drop the requirement. This step turns the description into an interview plan and keeps the posting honest about what the panel will actually measure. If two requirements use the same assessment, they may be the same requirement in different words, so merge them before the posting is approved. Keep one assessment per line, and then the plan is clear.
Review the draft with the hiring manager before it is posted. Ask them to read the outcomes and mark any that are unclear, inflated, or missing. Then ask them to check the must-haves against the interview plan. A hiring manager who owns the role is the right person to confirm the bar. A review that happens after posting is a correction, not a check. Give them the brief and the draft side by side, so they can see whether the description still matches the role you agreed at the start. Ask for one change at a time, then confirm the next, and keep notes in the draft.
Use Voiskey to capture the review comments in the document. Put the insertion point under the section being discussed, press the shortcut, and speak the comment as the manager gives it. The app removes filler and false starts, so the note is ready to act on. Tag each comment by section, then decide which changes to make. Capturing in the meeting keeps the feedback specific and avoids a second round of emails. When you move to a second machine, download Voiskey there as well, so the review notes live in the same document wherever the conversation happens. Read the tagged list back and confirm each owner.
Ask the manager to name the one outcome the role must deliver in the first six months. If the draft does not make that outcome prominent, rewrite the opening around it. A description that leads with the most important result is easier to read and easier to remember. This single question often reveals that the draft has been organized around tasks rather than around what the role is for. Write the answer in one sentence and use it as the test for every later edit. If an edit does not serve it, cut it. That rule keeps the draft tight, so apply it twice.
When the draft is approved, download Voiskey on the machine where you will publish it and keep the final text in one folder with the interview plan. A last edit by voice is useful when a requirement changes late. Read the changed section aloud to confirm the wording. Keeping the description and the interview plan together makes the next hire easier to run. If a reviewer asks for a change after the posting is live, update both documents in the same pass, so the public text and the interview plan do not drift apart. Check the plan after every edit, then repost the final text, and keep one source of truth.
Save the prompt that produced an approved description. Keep the role brief, the outcome language rule, the must-have test, and the review checklist together, with blanks for the new title and team. The next opening starts from a tested structure. Name the saved prompt by the role family, such as a support lead or a reporting analyst. The name is what makes it findable. Note the date of the last approval, so you can see whether the template reflects the current hiring bar. Update the date after each hire, because it keeps the template honest, and review it twice a year.
Record what the draft got wrong as well as what it got right. A note such as the must-have list was too long or the outcomes sounded like tasks is more useful than a general impression. These notes tell you which rule to change next time. A saved prompt that carries its own lessons improves with use, while one that is filed and forgotten stays as good as it was on the first try. Keep the notes short, one line per lesson, so the template stays easy to read. Date each lesson, group them by rule, and then review together.
Use Voiskey to update the saved prompt by voice when a new requirement appears. Open the prompt, press the shortcut, and speak the change and the reason. The note lands beside the rule, so the next description includes it. Read the changed section once before saving. This keeps the template current without a full rewrite and makes the next description faster to produce. If the change affects the must-have test, update that line first, because the test protects the quality of everything else in the template. Then run one test draft and check the outcomes. Adjust the rule if the test fails.
Share the saved template with other hiring managers on the team. Give them the structure, one approved example, and one corrected example, so the standard is visible. When a colleague sets up a second machine, download Voiskey there and copy the template in the same structure, then open one description to confirm it travels. A shared template makes postings comparable and keeps the bar consistent across teams. Ask each manager to add lessons, because a template that only reflects one team slowly loses touch with the others that rely on it. One shared folder is enough, and keep the notes current.
Begin with the role title, team, level, the outcome the person owns, and the three problems the role solves. Voiskey turns the spoken brief into an AI-ready prompt and removes filler, so the request stays short. Review the brief before asking for the draft.
Yes. Say the outcome for each responsibility rather than the activity, and ask for a rewrite around results. Voiskey adapts phrasing and formatting to the request, which keeps a short ranked list readable. If a line still sounds like a task, ask for it to be restated as a result.
Separate must-haves from nice-to-haves and say why each must-have is needed for the outcome. Keep the list to five or six items and ask how each will be assessed in an interview. Voiskey keeps the spoken sections in order, so the requirement list and the assessment stay together.
Ask the manager to mark unclear outcomes, check must-haves against the interview plan, and name the one result the role must deliver in six months. Capture comments by voice in the same document. Voiskey places the text at your insertion point, so the review notes stay beside the draft.

Download Voiskey for Mac and Windows and speak your next prompt in seconds.
Use it across macOS, Windows, iOS, and Android