ATTENTION, PEOPLE HOLDING TWO GUIDELINES THAT BOTH SOUND EXTREMELY CONFIDENT!
PEOPLE WHO HAVE RECEIVED THREE PDFS, FOUR LOGOS, AND ONE EMAIL SAYING “THIS IS NON-NEGOTIABLE” FROM A PERSON WHOSE JOB TITLE CONTAINS THE WORD SYNERGY!
Are you tired of not having a committee decide your life for you?
Have you stared at a recommendation, a policy, a code, a standard, and a manager’s deeply meaningful eyebrow — wondering whether any of them has just made the decision on your behalf?
Then congratulations.
You qualify for the Five Number One Rules of When the Experts Disagree About What You Should Do — five Number Ones for the moment official language meets an actual human consequence.
“Wait,” you’re yelling at the screen, “they can’t all be Number One!”
That’s exactly what Big Numbering wants: one glowing rule to use as a coupon for ignoring the other four.
Disagreement fails in specialist ways.
Knowing the issuer won’t reveal the evidence grade.
A strong evidence grade won’t create a mandate.
A mandate won’t show you who bears the cost of its default.
And a beautiful stakeholder map won’t govern a conflicted exception.
One more thing before we start, because it belongs at the top rather than in a footnote: formal consensus can be strong and still subject to revision, ethical duties vary by profession and jurisdiction, and none of this substitutes for professional, legal or clinical advice.
So here they are.
Five rules.
Five Number Ones.
No substitutions. No “but the PDF had a crest.” No pretending a footnote is a tiny judge with a gavel.
RULE #1: IDENTIFY WHO ACTUALLY ISSUED THE POSITION
Introducing LOGO-LOCATOR 9000™, the only device asking the radical question: who, precisely, is making this claim — and what can that institution actually do?
Start with the named issuer.
A clinical guideline commits its issuing society to a recommendation, an evidence grade, a date, and a planned review. Trustworthy development standards require that review be planned rather than left indefinite.
That makes a guideline a meaningful reference point. It does not adjudicate every individual case, establish a legal standard of care, or create a licensing consequence for a documented deviation.
Now watch the same word arrive from a completely different machine.
The USPSTF is an independent volunteer panel of sixteen primary-care clinicians, appointed to four-year terms, grading preventive services A through D plus I for insufficient evidence.
And under the Affordable Care Act, an A or B grade requires most private insurance plans to cover that service without patient cost-sharing.
That isn’t a society asking for professional agreement. It’s a statutory trigger bolted to a letter grade.
Same word — recommendation. Utterly different apparatus behind it.
So the practical move is almost embarrassingly small: write down the full issuer name before accepting the position’s force.
Is this a society guideline? A task-force grade? An ASTM specification? An Accounting Standards Update? An IPCC Summary for Policymakers?
A citation cascade is one body lending another a vocabulary. A coverage mandate is one body’s output operating a payment rule.
Their typography looks equally stern. Their authority is not the same species.
A name on the cover tells who has made a commitment. It does not, by itself, tell what that commitment commands.
An official seal is the beginning of due diligence. Not a substitute for it.
RULE #1: READ THE EVIDENCE GRADE AND THE DATE
NOW AVAILABLE: FOREVER-CERTAIN DELUXE™ — the recommendation arriving with no date, no grade, and a reassuring promise that evidence stopped moving the moment somebody exported the PDF.
Read the grade beside the sentence people keep quoting.
GRADE separates certainty — high, moderate, low, very low — from strength — strong or conditional. Those are two independent dimensions, and they get merged constantly.
A shared vocabulary across more than 120 organisations makes cross-specialty reading less like translating a menu in a wind tunnel. It does not turn a conditional recommendation into a mandate.
Then read the date as an operational fact rather than decorative metadata.
The 2017 ACC/AHA hypertension guideline lowered the diagnostic threshold from 140/90 to 130/80 mmHg.
Applied to the same U.S. adult data, estimated prevalence went from 31.9% to 45.6%.
No artery received a software update. The line moved.
Fourteen million people became hypertensive overnight, in the only sense that a definition can do that — which is worth sitting with before quoting any threshold as though it were a fact about bodies.
And the new line wasn’t universally accepted. The 2018 ESC/ESH guideline, followed by later guidelines in Canada, Japan and Latin America, retained 140/90 for the general population — reserving the tighter target for people already at elevated risk.
One evidence base. Two live consensus positions. Different committees, different institutional contexts, different choices.
Which is why “experts say” is unfinished grammar. Which experts? When? With what certainty? About which population?
Evidence has a grade. Guidance has a date. Neither arrives wearing a lifetime warranty.
The map may well be current. Check before using it as a floor plan.
RULE #1: SEPARATE A RECOMMENDATION FROM A MANDATE
FROM THE MAKERS OF “INDUSTRY STANDARD” COMES MANDATE-MATIC™: press one button and every confident sentence becomes legally binding, contractually enforceable, and somehow the fault of accounting.
Don’t buy it.
An ASTM specification is a testable contract — a material either meets its stated test method and tolerance or it doesn’t.
The process behind it is genuinely rigorous. Producers must make up less than half of any standards-developing committee. Under ANSI accreditation, every comment and negative vote gets a documented response, with an appeal available when the process fails.
Those rules say something real about how the consensus was built.
They do not make the published document mandatory.
The mandate appears only when a regulator, a building code, or a purchasing contract cites the standard by number. Publication and adoption are separate events, occurring at different times, sometimes by decades.
Saying “the standard requires” without naming what adopted it mistakes a specification for its enforcement mechanism.
Accounting gives the opposite extreme, and it’s instructive. Once a final standard applies, compliance is not optional for a public filer, and an auditor approving an undisclosed deviation faces real professional liability.
But even there — an accounting standard is a chosen convention for representing a transaction. Not a laboratory discovery about the world.
So the question is never “is this official?”
It’s: what external mechanism turns this document into force here? A coverage rule? A code reference? An audit requirement? A contract?
Or none of them?
A recommendation tells what an institution favours. A mandate tells what mechanism will act if it is not followed.
Confusing the two is how a footnote gets promoted to sheriff.
RULE #1: MAP WHO BEARS THE COST OF THE DEFAULT
BEHOLD STAKEHOLDER-O-VISION™, now with two axes, three attributes, and absolutely no permission to treat the loudest person in the room as the whole room.
Once you know the document’s force, map what applying it by default actually does to people.
The power-interest matrix crosses somebody’s power to affect the decision against their interest in the outcome. It gives you a triage order — manage the high-power, high-interest people closely, keep the rest informed.
Quick. Legible. Useful.
And it quietly risks treating low power as low consequence, which is a different claim entirely.
For a decision that has to survive scrutiny, use the salience model instead: power, legitimacy and urgency as three independent attributes, producing eight named types depending on which a party holds.
That legitimacy question is what earns its keep. It’s how you notice somebody bearing a serious cost while lacking any leverage to make a meeting uncomfortable.
Merck’s Mectizan decision shows the gap starkly. In 1987 Merck registered ivermectin for human use against river blindness. The affected populations couldn’t pay a commercial price, and no buyer had committed at scale.
A shareholder-only screen sees no purchaser and stops. It isn’t malfunctioning. It has genuinely finished the analysis.
In 1988 Merck committed to donate what was needed across more than 30 countries. Between 1995 and 2010 the programme is estimated to have prevented around seven million years of disability.
That case doesn’t prove stakeholder mapping reliably produces donations. It shows the question a narrower screen cannot even ask: who is exposed, whose claim is legitimate, and what happens to them precisely because they can’t pay or threaten anybody?
Power tells who can stop a decision. Stakeholder mapping asks who must live with it.
The absence of a complaint is not evidence of the absence of a cost.
RULE #1: ESCALATE THE EXCEPTION WITH REASONS ON THE RECORD
AND NOW, EXCEPTION-EXPRESS™ — an executive escape hatch with no paperwork, no witnesses, and one button labelled trust me, this one is different.
Available wherever governance becomes interpretive dance.
An exception can be entirely justified. A private exception, with no reason, no route and no record, is a different object.
So run the sequence. Get the facts. Ask whether the request appears improper. Clarify who shares responsibility. Take it to a supervisor. And where the normal channel is compromised, use the external reporting route that exists for exactly that.
The record needs to say what conflicts, what evidence supports the departure, who bears the exposure, who had authority, and what review follows.
Enron supplies the warning label, and the sequence matters.
A written Code of Ethics was circulating internally as late as July 2000.
But in 1999, the CFO had received a formal exemption from its conflict-of-interest provision — so he could run partnerships transacting with the company he served.
The code spotted the conflict. It worked exactly as designed.
Then the body with authority over the code suspended it, for the person who benefited most from the suspension.
Bankruptcy followed in December 2001.
The lesson isn’t that ethics codes are decorative, or that every waiver is corrupt. It’s that a waiver is a separate governance fact from the existence of a code — and it deserves more scrutiny than the code did, not less.
An exception earns trust by leaving a trail strong enough for someone else to challenge.
If a rule disappears precisely where the stakes rise, it wasn’t governing. It was decorating.
BUT WAIT, THERE’S MORE!
“What if the disagreement is between a doctor and an insurer?”
Same rules. Name the issuer, the grade, the date, and the actual coverage mechanism. Map the patient’s exposure and use the available professional escalation path.
“What if it’s about a product specification?”
Same rules. Identify the process behind it, then find the regulator, code, or contract that adopts it. A specification nothing cites remains voluntary, however many people call it industry standard.
“What if the official answer hurts the people with the least bargaining power?”
Same rules. Use the quick matrix for triage and the salience model when legitimacy and urgency need saying out loud. The default may still apply. The unequal burden must not become invisible.
“What if an executive says the exception is obvious?”
Same rules. Facts, authority, conflict, exposure, written reason, review. “Obvious” is not an escalation process. It’s the sample copy on the Exception-Express box.
“What if the consensus changes next year?”
Same rules. It’s dated, and it’s allowed to move. After the 2002 Women’s Health Initiative results, U.S. hormone-therapy prescribing fell roughly 40% within about a year — and later re-analysis added nuance to that.
The architecture isn’t certainty. It’s accountable movement.
The details change.
The architecture doesn’t.
THE FIVE, WITHOUT THE LETTERHEAD
Identify who issued the position and what that institution actually commits itself to.
Read the evidence grade, the date, the review status, and the population it covers.
Separate recommendation from mandate by finding the mechanism — statute, code, audit, contract — that converts a document into force.
Map power, interest, legitimacy and exposure. A default is not neutral merely because it is official.
Escalate a genuine exception with facts, named authority, written reasons and a review trail.
DO THIS BEFORE THE NEXT PDF ARRIVES
Tonight, pick one recommendation currently making a decision feel settled.
Open the actual document. Not the summary, not the email quoting it. The document.
Write down the issuing body, the date, the evidence grade, the review status, the population, and the exact sentence being relied on.
Then write what turns it into force here. A coverage trigger. A code reference. A purchasing contract. Or possibly nothing beyond the authority of a professional body.
“Only a professional body” isn’t an insult. It’s a category, and knowing which one you’re in changes what you owe the document.
Next, list the people who absorb the cost if the default applies. Put them on the matrix. If the explanation has to survive a sceptical room, add legitimacy and urgency.
Then look specifically for the person with the most exposure and the least ability to make noise. They will not appear on their own.
And if the default genuinely doesn’t fit — don’t improvise an invisible exception. Gather the facts, state the conflict, write the reason, take it to the designated authority, and set the review date.
That isn’t indecision dressed as procedure. It’s how a person respects evidence without letting institutional confidence conceal institutional limits.
For the low, low price of reading the issuer line before the conclusion, the complete Five Number One system is yours.
No chrome binder. No emergency stamp. No committee of twelve people in identical lanyards arriving at your door.
Just five questions that keep an official answer from becoming an unexamined one.
Operators are no longer standing by.
The operator is the person who has to decide, record, and live with what happens next.