ATTENTION, PEOPLE WHO HAVE SENT “PER MY LAST EMAIL” WITH GREAT VIGOUR!
PEOPLE WHO HAVE REPLIED “ANY UPDATES?” TO A THREAD CONTAINING ZERO DECISIONS! PEOPLE WHO HAVE WRITTEN “JUST CIRCLING BACK” WHILE DRIVING A TINY CIRCLE AROUND THE ACTUAL QUESTION!
Are you tired of not being asked what you meant?
Has a colleague ever sent a paragraph so polished, so courteous, so exquisitely full of “at this point in time,” that nobody could work out who was doing what by when without convening a small archaeological expedition?
Then congratulations.
You qualify for the Five Number One Rules of Writing When Someone Else Has to Understand You, an all-purpose document-delivery system containing not one, not two, but FIVE NUMBER ONE RULES.
“Wait,” cries the inbox, “they can’t all be Number One!”
That’s exactly what Big Numbering wants — because a ranked list lets a writer spend all morning polishing the opening sentence while the reader wonders whether somebody should be escalating.
Writing failures are specialists.
A readable sentence won’t assign an owner.
A named owner won’t supply missing context.
Full background won’t tell a reader when to bring in more people.
And an escalation path won’t rescue a decision buried three paragraphs beneath the company’s founding mythology.
Here’s the split that matters. Writing rules make a sentence legible. Communication rules test whether that sentence survives a handoff — whether a different person, in another time zone or on another shift, can identify the status, the next action, and the condition that changes the plan.
That’s the proof obligation. Not “this sounds professional.” Not “there was a meeting.”
The reader knows what happens next.
So here they are.
Five rules.
Five Number Ones.
No substitutions. No “per my last fog machine.” No putting a semicolon in a message and calling the evacuation complete.
RULE #1: PUT THE DECISION BEFORE THE BACKGROUND
Introducing PREAMBLE-O-MATIC™, the document appliance beginning with the weather conditions under which a request was first conceived, and reaching the request itself shortly before the reader’s retirement party.
Put the decision, the request, or the current condition first. Then the background needed to understand it.
The Federal Plain Language Guidelines say it plainly: write for the reader rather than the institution, main point first, use “you,” one main idea per sentence, active voice, common words where they carry the same meaning.
That is not an invitation to send a note reading DO THING and then vanish into the bushes. It’s an order of operations. The action arrives before its family tree.
And for an agency writer, that guidance has teeth — the Plain Writing Act of 2010 makes it compliance guidance rather than a decorative preference.
Which is worth distinguishing from house style. AP or Chicago or an in-house guide is enforced through the edit. A statutory requirement is enforced elsewhere.
The sentence-level move can look identical. The authority behind it isn’t. A house style can remove a comma; a statute can create a compliance problem.
Readability formulas help here, as indicators rather than a small digital oracle in a waistcoat.
Flesch Reading Ease runs roughly 0 to 100, with higher meaning easier — 90 to 100 is “very easy,” below 30 “very difficult.” Flesch-Kincaid converts the same inputs into a grade level, where 8.0 approximates eighth grade.
But look at what those formulas actually count: words per sentence, and syllables per word.
Surface features. Neither one has any way to detect whether a handoff succeeded.
Put the decision where the reader can find it before the reader needs to find it.
Background is support staff. Don’t make it open the show.
RULE #1: USE VERBS THAT NAME THE ACTOR
BEHOLD MYSTERY-DOER DELUXE™: every action arrives swaddled in a noun, nobody performs it, and the sentence stands nobly beside a broken copier while refusing to identify the person holding the screwdriver.
Use a verb for the action, and name the actor when the actor matters.
“Make a decision” becomes decide. “Conduct an assessment of” becomes assess. “Have an understanding of” becomes understand.
Nominalisation is the institutional habit of turning an action into a formal-looking object, then hiring “make” or “conduct” to carry it around. It adds words and hides the moving part.
Now — don’t start arresting every “was” and “is” on suspicion of passive voice. This is where enthusiasm outruns grammar.
A true passive has a form of be or get immediately followed by a past participle. Then it survives the by-agent test: insert “by [someone]” after the verb.
“The report was written by the analyst” describes an action. Passive.
“The door is closed, so no one enters” describes a state — “closed” is a predicate adjective and “is” is a copula, not an evasive passive lurking in a trench coat.
The question worth asking isn’t grammatical anyway. It’s whether the sentence withholds a responsible actor when that actor determines the next move.
“The incident was reviewed” hands the reader a ceremonial plaque.
“Mina reviews the incident and posts the decision” hands them an owner and a next action.
Expletive constructions make the same little cloud of administrative weather. “There is a need to review the budget” spends four words moving the subject away from its job. “The finance lead reviews the budget” doesn’t.
A verb tells the reader what moves. A named actor tells the reader who moves it.
If a sentence can’t name the doer, it had better name the reason.
RULE #1: CUT WORDS, NOT INFORMATION
NOW AVAILABLE: THE VERBOSITY VACUUM™. It removes every useful condition, exception, deadline and technical term — leaving a glossy one-sentence slogan incapable of surviving being forwarded.
Don’t buy it.
The rule is omit needless words. Not make every message smaller until it becomes a hostage note.
Cut a phrase only when the deletion preserves the information. “In order to” becomes “to.” “Due to the fact that” becomes “because.” “At this point in time” becomes “now.” Kill the redundant pairs — each and every, first and foremost, past history, future plans.
Strunk’s rule against overwriting takes aim at stacked qualifiers: somewhat, rather, quite, very, really. A claim stated plainly is usually stronger.
But that rule does not authorise deleting a real hedge.
“Response norms depend on team and incident severity” is information. An invented universal reply deadline is not information — it’s confidence, and the vacuum will happily trade you the second for the first.
This is also where the readability formulas reach their limit. They count two proxies. Their constants were fitted against measured comprehension on graded passages, and institutions do rely on them — the US Army adopted Flesch-based requirements for technical manuals, and Florida requires personal-lines insurance language to score at least 45.
Those are institutional uses of a measure. Not a completeness certificate.
A good cut preserves the name, the condition, the owner, the action, and the consequence.
“Send the report” is shorter than “please facilitate the transmission of the report.”
But “send the report to Priya by the agreed response window; escalate if the status changes” is longer — and better, because it carries the handoff.
That extra clause isn’t clutter. It’s the load-bearing wall the vacuum would most like to remove.
Brevity removes the wrapping. Clarity keeps the contents.
The shortest message isn’t the one with the fewest words. It’s the one that doesn’t require a second message explaining the first.
RULE #1: STATE THE OWNER AND THE RESPONSE WINDOW
FROM THE MAKERS OF REPLY-ALL RESPONSIBILITY™ comes SOMEONE-WILL-HANDLE-IT PRO™ — a magnificent group inbox in which every recipient confidently assumes a different recipient has developed a plan.
Name one owner. State the current status. State the next action. State the response window.
A Directly Responsible Individual model puts one named person on each area, accountable for it. That’s the difference between sending a request to a group and assigning work to somebody who can accept, decline, or visibly transfer it.
Then the response window, which is the part people skip and the part async work runs on.
Doist documents an expectation that async messages get answered within 24 hours rather than in real time, alongside a mix of roughly 70% asynchronous, 25% scheduled synchronous, 5% in person.
Those are one company’s norms, not a stopwatch to bolt onto every team. But the structural lesson generalises completely:
A reader needs to know whether silence means working on it, haven’t seen it, or time to escalate.
Without that, silence means all three simultaneously, and everybody picks the reading that suits them.
Handbook-first practice supplies the other half — decisions, policies and context written into a searchable public record by default, so an async message can point at stable information instead of asking somebody to reconstruct private chat history.
Healthcare has the most heavily measured version of this. The I-PASS bundle — illness severity, patient summary, action list, contingencies, and synthesis by the receiver — was studied across nine hospitals and more than 10,000 admissions, reporting a 23% relative reduction in medical errors and a 30% reduction in preventable adverse events.
That’s evidence about a bundle, in that setting, with training and a written document attached. Not proof that a mnemonic alone makes an email work.
A handoff is complete when the next person can act, not when the sender has pressed Send.
Ownership without a window is a waiting room. A window without an owner is a smoke alarm in an empty building.
RULE #1: WRITE THE ESCALATION TRIGGER IN ADVANCE
AND NOW, PANIC-BASED ROUTING™ — the revolutionary system where every incident begins with everybody asking whether it’s “bad enough,” until it becomes bad enough to require a meeting about why nobody called anyone.
Write the condition that requires escalation, and the response it triggers, before the condition occurs.
PagerDuty classifies incidents from SEV-1 to SEV-5 against written qualifying conditions — a SEV-1 pages an incident commander, a SEV-5 becomes a routine ticket.
And when classification is ambiguous, the rule is to default to the higher severity rather than pause to deliberate.
The point isn’t the number of tiers. The point is that anxiety doesn’t get to be the decision-maker.
There’s no universal severity ladder hiding in the supply closet, either. Google’s SRE guidance escalates on judgment: when an incident is complex enough to involve multiple teams, or when responders can’t estimate an upper bound on duration.
Clear paths and cross-team rotations do the work a numbered classification does elsewhere. Name the threshold, the route, and the next person.
This is also why a beautifully written status update can fail completely.
“We are monitoring the situation” is prose. Lovely prose, if the writer is unusually gifted with fog.
“Jordan owns monitoring; next update within the response window; escalation if the written condition is met” is communication — it tells a reader who wasn’t there how the message becomes an action.
And copying a severity label without its governing condition is just buying a fire extinguisher that says probably.
Escalate on a condition, not on a feeling that has opened too many tabs.
The worst time to negotiate the alarm is while it is ringing.
BUT WAIT, THERE’S MORE!
“What about a short email?”
Same rules. Ask first, actor named, condition kept, response window stated, and a line about when it goes past the recipient.
“What about a policy?”
Same rules. It still needs an owner, a searchable location, an action, and an escalation condition for when ordinary handling fails.
“What about a shift handover?”
Same rules. Status, open work, discrepancies, and what happens next — rather than a warm feeling that the previous shift was extremely articulate.
“What about a remote team?”
Same rules. Handbook-first for context, a named individual for ownership, an explicit window so silence becomes interpretable.
“What about an incident?”
Same rules. Numbered tiers where external classification matters, judgment-based escalation where it doesn’t. Either way, write the trigger before the incident forces improvisation.
The details change.
The architecture doesn’t.
THE FIVE, WITHOUT THE CIRCLING BACK
Put the decision before the background, so the reader meets the action before the explanation.
Use verbs that name the actor — and run the by-agent test before treating every “is” as a crime.
Cut throat-clearing, nominalisations and stacked qualifiers, while preserving the information that makes the instruction safe.
Name one owner, the status, the next action, and the response window.
Write the escalation threshold and route in advance, calibrated to your team rather than borrowed as a sacred number.
DO THIS TONIGHT, BEFORE THE THREAD BECOMES A DIG SITE
Take the last message that produced a reply chain nobody enjoyed.
Move its decision into the first sentence. Just that, first — you’ll often find the decision was in paragraph four, wearing a disguise.
Then find every “there is,” “it is,” “make a decision,” “in order to,” and “at this point in time.” Cut only what carries no information. Run the by-agent test before flagging anything as passive.
Now do the part that style advice can’t do for you.
Add the named owner. The current status. The next action. The response window.
Put the durable context somewhere a later reader can search, rather than somewhere a later reader has to excavate.
Finally, write the condition that leaves the ordinary route — a severity tier, a multi-team threshold, or an inability to estimate how long this runs.
Then ask the only question that matters:
Could a different reader, with no meeting and no telepathy, know what happens next?
If not, the message is still being written.
For the low, low price of refusing to make the recipient solve a riddle, the complete Five Number One system is yours.
No subscription. No ceremonial “just bumping this.” No meeting convened to establish that nobody owned the response.
And if you act now, we’ll include the most valuable document feature ever invented, at no additional charge:
a next step.
Operators are no longer standing by.
The reader is the operator. Give them something to operate.