Writing at Work: Audience, Purpose and Context

A nurse writes an incident note at three in the morning. A project lead sends four lines of email and a frozen budget releases. A student messages a clinical coordinator, hears nothing back, and never learns why. Workplace writing is not schoolwork with a job attached — it is the job. It moves money, schedules, patients, permissions and blame. And it fails for boring, fixable reasons: the reader could not tell what was being asked, or why it concerned them, or what to do by when. This chapter drills the three questions sitting underneath every professional document. Who is actually reading this? What do I want to happen? What conditions am I writing inside? You will build a reader profile in ninety seconds, write a purpose statement that contains a verb and a date, pick a channel on purpose instead of by habit, and rebuild one real message from ignored into answered.

Section 1Writing that has a job to do

What school rewards, what work rewards

Most of us learned to write for a reader who was paid to finish. An instructor opens your paper already committed to the last page, hunting for evidence that you understood the material. That reader is rare and, outside a classroom, largely imaginary. At work your reader has ninety unread messages, a meeting in eleven minutes, and no obligation whatsoever to reach your second paragraph. They are not evaluating you. They are looking for the part that concerns them, and if they cannot find it quickly they will move on and tell themselves they will come back later. They will not come back later.

This changes what counts as good. An essay rewards visible thinking — you show the work because the work is the point. A workplace document rewards a decision made sooner, a task done correctly the first time, a risk recorded before it becomes an event. Reader-centered writing is the name for that shift. You organize the page around what the reader has to do, not around the order in which you personally figured things out.

It is not a softer standard. Compressing a two-week audit into five sentences a director can act on is genuinely harder than writing fifteen pages about it. Anyone can transcribe everything they know. The compression is the skill, and it is the skill this book keeps returning to.

Notice the last row. School writing dies on delivery. Work writing keeps living — forwarded to a person you never met, pasted into a slide deck, pulled from an archive by a surveyor three years on. You are always writing for at least one reader who is not in the room.

Your reader is scanning, not reading

Usability researchers have spent nearly thirty years watching where eyes land on screens. The recurring finding is that people scan in patterns — an F shape on dense text, a layer-cake pattern when headings are strong — rather than reading line by line. Scholars argue, reasonably, about how far laboratory eye-tracking generalizes to a six-line email read on a phone between patients, and the honest answer is that the exact pattern varies with the page. What is not seriously disputed is the direction: the first line and the left edge collect the most attention, and attention decays quickly after that.

So the practical rule writes itself. Put the load-bearing sentence where the eye already is. If your most important sentence is in the middle of paragraph three, you have hidden it, no matter how well it is written. Writers resist this because it feels abrupt, even rude. It is neither. Front-loading is a courtesy: it lets a reader who is not the right person for this message stop reading in four seconds instead of forty.

The same logic governs subject lines, headings and the first sentence of each paragraph. Chapter 4 turns this into page design; here it is enough to internalize the underlying fact. You do not get the reader’s full attention. You get a sample of it, and you choose which sample.

Bottom line up front

Bottom line up front, usually shortened to BLUF, is a habit borrowed from United States military staff writing, where a briefing memo opens with the recommendation and only then supplies the reasoning. Civilian offices adopted it for the obvious reason: it works on busy people. In practice BLUF means your first two sentences answer three questions — what is this about, what do you want from me, and by when.

The single most useful sentence pattern in professional writing is the action request: I need [who] to [do what] by [when] so that [consequence]. Fill in all four slots. A request missing the deadline becomes a suggestion. A request missing the consequence becomes optional. A request missing the name becomes somebody else’s problem, which is to say nobody’s.

Read the right-hand column end to end and notice what it never does. It never explains how the writer became interested in the topic. It never narrates the investigation. It does not apologize for taking up time, which paradoxically takes up time. Every sentence either states the ask, supports it, or lowers the reader’s risk in saying yes.

Section 2Reading the reader

Four readers hide in one distribution list

The person in the To line is not automatically the person who decides. Professional communication tends to travel through a small cast, and confusing one for another is how good writing produces bad outcomes. Your primary reader is whoever must act. Your secondary readers are copied in to stay informed or to object. A gatekeeper stands between you and the decision — an administrative assistant, a compliance reviewer, a supervisor who forwards upward with a one-line verdict attached. And a tertiary reader is the stranger in the future: an auditor, an attorney, a surveyor, the new manager reconstructing why this was done.

Most writers draft for the primary reader alone, which is understandable and incomplete. The gatekeeper is the one who decides whether the primary reader ever sees it. The tertiary reader is the one who decides, years later, whether you behaved reasonably. Both of them read differently from the person you had in mind.

Practically, this means one document usually has to serve several appetites at once: a decision for the primary reader, reassurance for the secondary readers, a clean summary the gatekeeper can forward without rewriting, and enough dated specificity that a stranger can follow the reasoning cold. That is not a contradiction. It is a layered structure — ask first, evidence second, detail in the attachment.

Expertise, stakes and the translation gap

Expertise is not intelligence, and mistaking one for the other produces both of the classic failures. Write above your reader and you get the translation gap: technically flawless prose that produces no action, because the reader cannot convert it into anything they control. Write below your reader and you get condescension, which costs credibility you will need later. The chief financial officer does not know what a rapid response team activation is; she knows exactly what an unplanned transfer to intensive care costs.

The other axis is stakes — what this reader stands to gain, lose, or be held responsible for. Stakes decide attention far more reliably than interest does. A charge nurse cares about staffing minutes; a risk manager cares about the gap between policy and practice; a family member cares whether their mother is safe tonight. Naming the reader’s stake in your opening is not manipulation. It is relevance, stated out loud instead of assumed.

For mixed audiences, the reliable rule is this: write the body for the least technical reader who must take action, and push the depth into an attachment. Experts are not offended by a clear summary; they skip to the appendix. Non-experts confronted with the appendix simply stop.

Three versions, one truth. Nothing in the compliance entry contradicts what the daughter was told; the daughter’s version simply omits what she cannot use. That distinction — between omitting the irrelevant and shading the inconvenient — is the ethical line, and Chapter 10 walks it in detail. Audience adaptation changes emphasis, vocabulary and depth. It never changes the facts.

Build the profile in ninety seconds

An audience profile does not need to be a form. Five questions, answered in your head or in the margin of a notepad, will do more for a draft than an hour of polishing. Who has to act? What do they already know? What do they stand to lose? What will they object to? What do they need in order to say yes?

Take a real case. You want three colleagues released from floor duty for a four-hour training on a new documentation system. Primary reader: the nurse manager who controls the schedule. Gatekeeper: nobody, but the staffing coordinator will be asked. Secondary readers: the three colleagues, and the charge nurse who will be short-staffed. Tertiary: whoever audits training completion at go-live. Prior knowledge: the manager knows the system is coming and does not know that unfinished training carries a documented penalty at go-live. Stakes: her staffing ratio for one shift, versus a unit that cannot chart on day one. Predictable objection: I cannot lose three people at once.

That last answer rewrites your message before you type a word. You now know to open with the go-live date, offer to split the training across two shifts, and name who covers the floor. You have answered the objection inside the request, which is the difference between a proposal and an argument.

Section 3Naming the purpose before you type

Purpose is a verb, not a topic

Ask a writer what their document is about and you will usually get a noun phrase: the new scheduling software, the incident on Tuesday, our quarterly numbers. That is the topic. The purpose is what you want to be different after the reader finishes. Those are not the same thing, and documents that confuse them tend to end with the two most useless sentences in professional life: Let me know your thoughts and Please advise.

Topic: the new scheduling software. Purpose: get Dana to approve two paid hours of training per staff member before the 15 September go-live. Watch what the second version does. It supplies a test for every sentence you are about to write — does this move Dana toward approving two hours? — and it tells you when to stop writing, which topics never do.

A purpose you cannot state in one sentence is usually two purposes fighting. That is worth catching early, because a document doing two jobs typically does neither. The update that is secretly a complaint. The request for information that is secretly a request for backup. The status report that is secretly a bid for more staff. Readers detect the hidden purpose immediately and respond to it rather than to the stated one.

The one-sentence purpose statement

Use a fixed frame, at least until the habit sets: After reading this, [reader] will [specific action] by [date], because [reason that matters to them]. Write it at the top of your draft. Delete it before you send. If you cannot fill every slot, you are not ready to write — you are ready to go find out something.

The test for a good purpose statement is whether an outsider could tell you succeeded. “Raise awareness of hand hygiene” fails that test; nobody can measure awareness, and nothing in the sentence tells you what to write. “Get every day-shift nurse on 4 West to complete the two-minute hand hygiene attestation by 30 June” passes it, and the document practically drafts itself: here is the link, here is the deadline, here is why it matters, here is who to ask if the link fails.

Notice how the fourth example makes life easier for the reader, not just for the writer. “I have a question about tuition reimbursement” obliges the coordinator to open a conversation. The rewritten version can be answered yes or no with a form number, in eleven seconds, from a phone. Requests that are cheap to answer get answered.

Four jobs, and the trouble with mixing them

Nearly all workplace documents do one of four things. They inform (here is what happened, no action needed). They request (do this, decide this, send me this). They persuade (change your position, fund this, adopt this). Or they record (this exists in writing now, dated, in case it is ever questioned). Some documents carry a fifth, quieter job — maintaining the relationship — which is why thanks and acknowledgment are not decoration.

Pick one as primary and let the others ride along. A well-built proposal informs while it persuades. A good incident note records while it informs. Trouble starts when two purposes are equally weighted, because structure has to serve one of them and you end up with a document that buries its request under a narrative, or a record cluttered with argument. If your reader has to ask “is this an FYI or do you need something from me?” you had two purposes and shipped both.

One more discipline, unglamorous and reliable: one document, one ask. If you genuinely need three unrelated decisions from the same person, you will get a faster and better answer from three short messages than from one long one containing a numbered list, because the reader can dispose of each independently instead of waiting until they have bandwidth for all of it.

Section 4Context: channel, constraint and the permanent record

The rhetorical situation, on a Tuesday afternoon

The classical name for all of this is the rhetorical situation. Lloyd Bitzer gave it its modern form in 1968, describing three parts: an exigence (a problem that writing might fix), an audience capable of acting, and a set of constraints shaping what can be said. It remains a useful map. It is also genuinely contested — Richard Vatz argued in 1973 that situations do not summon discourse so much as writers construct situations by deciding what deserves attention, and that argument has never been settled. You do not have to take a side. You do have to notice that both camps agree the writing is shaped by something outside the writer, and that noticing it is part of the work.

On an ordinary Tuesday it comes down to blunt questions. Why now? What just happened that made this necessary? Who has the authority to fix it? What can I not say — because of policy, privacy, hierarchy, or because it is not mine to disclose? What has already been said about this, and by whom? A brilliant memo sent the week after the budget closed is a bad memo.

Timing deserves its own line. The same request lands differently on the Monday of a survey week, the Friday before a holiday, and the morning after a bad outcome. Skilled writers do not send everything the moment it is finished. They ask when the reader can actually do something with it, and sometimes the answer is Thursday.

Choosing the channel on purpose

A channel is the pipe: face-to-face, phone, email, chat, formal memo, ticketing system, the chart. Most people choose by habit or by whatever is open on screen. Choose instead on four variables — how complex the content is, how fast you need a reply, how much emotional load it carries, and whether it must exist as a record afterward.

Two rules of thumb cover most cases. First, bad news and negotiation go to voice, and the record goes to writing. Deliver the hard thing in a conversation where you can hear the response, then follow with a short written summary of what was agreed. Second, the more people who must not misunderstand, the more it belongs in writing. Corridors are excellent for speed and terrible for accuracy at scale.

The row people misjudge most is chat. Because it looks like conversation, it gets written like conversation — and it is retained like email. Speed of typing is not the same as impermanence of storage.

Constraints you cannot write around

A constraint is any condition that limits what you may say, how you may say it, or when. Time is the obvious one. The less obvious ones do more damage when missed. Policy constrains what you are permitted to promise on the organization’s behalf. Privacy law — HIPAA for health information in the United States, FERPA for student records, and comparable regimes elsewhere — constrains what you may include and who may receive it. Hierarchy constrains sequence: telling a director something before her own supervisor knows is a factual message and a political error. Accessibility constrains form, since a document that only works as a scanned image excludes any colleague using a screen reader.

Constraints also protect you. “I am not able to share that in email, but I can walk it through with you on the phone” is a complete, professional answer. New employees often experience constraints as obstacles to good communication. Experienced ones use them as the shape of the message: knowing what cannot go in the email tells you what the email is for.

One constraint deserves naming because it is invisible until it bites: register, the level of formality a given relationship and setting expects. The same words that read as warm and efficient to a peer read as presumptuous to a regulator and cold to a grieving family. Register is not politeness; it is calibration. Chapter 3 gives it a full treatment.

Everything is a document of record

Assume that anything you type at work is a document of record: retained on a server, retrievable, and potentially discoverable — that is, subject to production in litigation, in a public-records request, or in a regulatory review. Deleting your copy does not delete the organization’s copy. This is not a reason to write nothing. It is a reason to write as though a reasonable stranger will read it, because eventually one might.

Two habits follow. First, keep speculation out of records. “Found on floor at 02:14, unwitnessed; alert and oriented, no visible injury” is a record — note that even “fell” is an inference when nobody saw it. “Patient fell, probably because night staff were slammed again” is an opinion that will be read aloud someday by someone unfriendly. Second, never quietly alter a record. Corrections belong in an addendum or a late entry, dated and attributed, which Chapter 13 covers in clinical detail. A visible correction looks careful. A silent one looks like something else.

Section 5Putting it to work: a routine and a rubric

PACT: five minutes before you type

Everything above collapses into four questions you can answer faster than you can make coffee. P — Person. Who must act, who is watching, who inherits this later? A — Aim. One sentence: reader, verb, date. C — Constraints and channel. What limits apply, and what pipe fits? T — Test. What observable thing tells me this worked?

Here it is filled out for the training-release message from Section 2. Person: Dana approves; the staffing coordinator executes; three colleagues and the charge nurse are affected; the go-live auditor reads it later. Aim: after reading this, Dana will release three nurses for four hours of documentation training before 1 June, because untrained staff cannot chart at go-live. Constraints: she cannot lose three people from one shift; the training vendor offers only two dates; it goes by email because it needs a record and an attachment. Test: Dana replies with approved dates within three business days.

That test line is the underrated one. “Dana replies with dates within three days” is checkable; “Dana understands the urgency” is not. Define success as something the reader visibly does, and you will find that your draft gets shorter and your follow-up gets easier, because you know exactly what you are following up on.

One message, rebuilt

Draft 1, as originally written. “Hi Dana, Hope your week is going okay! I wanted to reach out about the documentation training that is coming up. I know everyone is really busy right now with the survey coming and I completely understand if this is not the right time. Marcus mentioned that the vendor has some dates and I was thinking it might be good if some of us could go, since the new system is pretty different from what we are used to and there could be a learning curve. I think it would probably help with the transition. Anyway, no rush, but let me know if you think that could work. Thanks so much!! — J.”

Diagnosis. The purpose is never stated as an action. There is no date, no number of people, no cost, and no consequence for saying no. Three separate sentences apologize for asking. “Some of us” and “some dates” hand the reader the work of finding out. The strongest fact in the writer’s possession — that staff who miss training cannot document on the first day of the new system — is compressed into “there could be a learning curve.” Dana will file this under things to think about, which is a folder with no exit.

Draft 2. “Subject: Approval needed by 14 May — release 3 nurses for documentation training. Hi Dana — I need your approval by Thursday 14 May to release three day-shift nurses for four hours of vendor training, split across the 22nd and the 29th so we never lose more than two people at once. Why it matters: at go-live on 1 June, staff without the training cannot chart in the new system, and the downtime workaround is paper. Coverage: Marcus has agreed to float on both dates, and I have the charge nurse’s sign-off, attached. Cost: zero — the vendor sessions are included in the contract. If losing two at once on the 22nd is still too many, I can send one on the 22nd and two on the 29th instead. Just tell me which split you want and I will book it. — Jordan”

Second draft is barely longer than the first. It is not more assertive by personality; it is more assertive by structure. The ask sits in sentence one, the consequence is concrete, the predictable objection is answered before it is raised, and the close offers a real alternative so that “no” still produces a decision. That last move is worth stealing: give a busy reader two acceptable options and you convert an open question into a choice, which is far cheaper to answer.

The pre-send rubric

Run this on anything that matters. It takes under two minutes and catches the overwhelming majority of failures, which are structural rather than grammatical.

If you want a faster version, use the ten-second test. Hand the draft to a colleague, let them look at it for ten seconds, take it back, and ask what you want and by when. If they cannot say, the document is not finished, however clean the sentences are. The test is brutal because it measures the only thing the reader will actually experience: locating the ask.

Words worth knowing