Influencing Without Authority

P62.productive-disagreement.03 · Audience: guest, it-ml, language-pro · Prerequisites: Disagreeing with a Decision — Steelman First

Real LLM grading for this pageLLM grading (this page):

A product manager wants the flashy feature: an LLM assistant on the checkout page, demoed to the board next month. The engineer knows what the demo will not show — every request adding two seconds of latency to the one page where seconds are money. Here is the trap: the engineer is right, and rightness is not the game. She cannot veto the feature; she can only persuade someone who does not report to her, does not share her vocabulary, and is measured on things she has never had to care about. Her first instinct — explain the architecture, louder — has failed everywhere it has ever been tried, from engineering floors to school staff rooms to translation agencies negotiating deadlines with sales teams. What works is a different craft entirely: making your concern land inside the other person's world, arriving with evidence instead of opinion, and doing the real persuading before the meeting where the decision gets made.

Step 1 / 5 — Their incentives, not your reasons

The engineer's concern is real, but "the p95 latency will regress" is not a sentence the PM can act on — it is homework in a foreign language. Influence starts with a translation exercise: what does this person get measured on, worried about, congratulated for? The PM owns conversion, the board demo, the quarter. So the concern must arrive as: "two extra seconds on checkout historically costs us conversions — this feature could win the demo and lose the quarter it's supposed to justify." Same fact, different owner. The rule generalises to every field: the head cares about inspection results, not your pedagogy; the client cares about their launch date, not your revision process. Speak to the incentive, and you stop being an obstacle and become a person protecting what they already care about.

ⓘ Concept: Translate the concern into the listener's ledger
influencebeginswhereyourconcernintersectstheirincentives:findwhattheyaremeasuredon,restatetheriskasarisktothatinfluence begins where your concern intersects their incentives: find what they are measured on, restate the risk as a risk to that

Why it matters — People do not resist good arguments; they resist arguments that arrive as costs in someone else's currency. Until your concern touches something the stakeholder is measured on, hearing you is optional — you are asking them to care on your behalf, which is a favour, and favours are rationed. Translated into their ledger, the same concern becomes information about their own success, which no competent professional ignores. The translation also forces a discipline on you: to do it, you must actually learn what they care about, and that inquiry — often one honest conversation ('what does a great quarter look like for you?') — routinely reveals that your positions conflict less than your vocabularies did. This is the steelman's outward twin: fidelity to their world instead of your own.

🗣️ From a linguist's perspective: Skopos: the audience decides the translation
Translation theory has a name for this — skopos: a translation is governed by the purpose it must serve for its target audience, not by loyalty to the source text's surface. A translator localising a safety manual for site workers does not preserve the engineer's elegant subordinate clauses; they preserve what keeps fingers attached. Influencing without authority is skopos applied to your own argument: the source text is your concern, the target audience is the PM, and a "faithful" rendering that ignores the audience's purpose is simply a failed translation — accurate, elegant, and unread.

Nothing in the loop requires an org chart or a codebase. Teachers move heads, interpreters move agencies, and junior engineers move roadmaps with exactly these four moves — because influence without authority is not a workaround for lacking power; it is what most professional persuasion has been all along.

Ask the mentor about this module

Ask a question about this content. The mentor explains and grounds its answer in what you are studying; asking is recorded as a learning signal, not a grade.

Images, PDF or text. Kept on this device only.
Keeping your files on this device

Off by default. The mentor always gets your file; this only decides whether your own copy stays here. Copies live in this browser only - they do not follow you to another device, and clearing site data removes them.

Ctrl/Cmd + Enter to send

🎓 Practice ladder

2 graded rungs · ~22 min

Two mentor-graded craft rungs. Rung 1 gives you the flashy-feature scenario: a latency concern to rewrite for the PM's own incentives, with an evidence offer and a shrunk ask. Rung 2 is your own influencing-without-authority story with the evidence move named; a low-stakes story is always an accepted substitute. The mentor grades the telling, never your history.

Rung 1 — Speak to their ledger (explorer)

Loading exercise…

Rung 2 — The decision you moved without authority (practitioner)

Loading exercise…

Try it yourself

A scratch console for this page's ideas — ungraded, nothing you run here is recorded.

Scratch console

A scratch console with the scientific stack (pandas, numpy, scikit-learn). Runs on the server — no network, resource-limited and measured.

Output appears here.

My notes on this module

Loading your notes...

Influencing Without Authority — TransformerLab