The raise you'd earn by threatening to leave

You know the number is wrong. Two years, no bump, and the guy you onboarded now out-earns you. A recruiter pinged you last week about a role you have zero intention of taking. And a thought lands, clean and cold: what if you just got the offer — walked it into your manager's office — and let the fork do the talking?

Half your colleagues swear this is the only move that ever works. That's exactly why the question is worth slowing down on.

The dilemma

Here's a composite — stitched from the kind of career dilemma we see in this category, not one real person.

Nikhil has been a backend engineer at a mid-size firm for three years. He's good; he's the person paged when billing breaks at 2am. His pay has drifted about 20% below what the same title clears elsewhere, and he knows it because two recruiters told him so this quarter. He likes his team. He likes his manager, Reema, who once fought to keep him when a reorg nearly cut his role. He does not want to leave.

What he wants is the raise.

So he's built a plan. Interview at a competitor purely to extract a written offer — no intent to accept — then bring it to Reema as leverage. He's rehearsed the sentence: "I've got a competing offer at [X], and I'd rather stay, but I need you to match it." Every colleague he's asked says the same thing: this is how the game is played, stop being naive, get the paper.

The part he doesn't say out loud: he'll have to keep doing this. The raise he wins by threat is a raise Reema now watches for the next threat. He'll have manufactured a fork he has to keep manufacturing. And the manager who once went to bat for him becomes someone he negotiates with at knifepoint.

The question he's actually asking isn't "will it work." It's whether the version of him that wins this way is one he wants to keep being.

The read

Four of the eight lenses bite hard here, and two of them are openly at war.

Duty — is the underlying claim legitimate?

Duty asks what the role and the relationship actually owe. Nikhil's core claim is true: he is underpaid for the work he does, and asking to be paid fairly is not a betrayal of anyone. A firm that lets a key engineer drift 20% under market has already broken a small promise. So the ask scores well — a person naming their worth to the person who sets their pay is duty in plain clothes. Duty: +3. The problem isn't the request. It's never been the request.

Honesty — what does the leverage require you to say?

Honesty asks whether the words match the world. And this is where the plan buckles. The sentence "I'd rather stay, but I need you to match it" contains a fabricated fact: an offer he will never take, presented as a live threat to leave. He isn't overstating a real fork. He's building a fake one and handing it to someone who trusts him. The competing offer is real paper wrapped around a lie about his intent. Honesty: −4. Notice the gap: the claim underneath is honest, the instrument he's chosen to press it is not.

Motive — what is the engine actually running on?

Motive asks what's driving the specific move, not the noble summary. Strip the story and Nikhil's engine is fair: he wants correct pay. But the method's engine is different — it's a bet that Reema will only move under threat, so threat becomes his vocabulary. Once fear is the tool that reliably gets you what you want, it stops being a tool and becomes a habit. He isn't just extracting a number. He's teaching himself that leverage is how he talks to people who trust him. Motive: 0 — clean want, corrupted delivery.

Detachment — can you survive the honest 'no'?

Here's the war. Detachment asks whether you can act rightly and release the outcome. The whole appeal of the fake offer is that it removes the risk of hearing "no." A plain ask — "I'm 20% under market, here's my evidence, I'd like you to fix it" — can be declined. The manufactured threat is engineered precisely so it can't be, because now walking is on the table. Detachment sees a man so unwilling to risk the real answer that he'll forge a gun to avoid asking the question. Detachment: −3. And this is where it collides with duty: duty says your claim is strong enough to make plainly, while the plan exists only because Nikhil doesn't believe that. The stronger the honest case, the less the bluff should be necessary — and the more its presence exposes what he's actually afraid of.

Two lenses, one contradiction: duty says you have the standing to ask straight. The fake offer says I don't trust my standing to survive daylight. Both can't be true.

The verdict: name the number plainly with real evidence and risk the no — the fabricated offer wins a raise you'll have to defend, by threat, every single year.

Here's what a general AI chatbot does with this dilemma, and why it's a different machine. Lean into it defeated and it consoles you: everyone bluffs, get the paper, protect yourself. Lean in guilty and it moralizes: honesty above all, never deceive. It argues whichever side your prompt is already leaning — sycophancy is a documented behavior, not a bug you can prompt away. KarmaLens doesn't renegotiate to match your mood. The same four lenses fire, the scores aggregate the same way, and the verdict is committed before it knows which answer you were hoping for. When honesty lands at −4, it stays −4 whether you argue with it or not — and the scripture it cites is attached verbatim, not paraphrased to flatter the read you wanted.

The takeaway you can run tonight

You don't need us to test this one. Run the two-sentence fork test.

Write your ask as two separate sentences. Sentence one, the claim: what you're worth and why, in numbers you can defend — "I'm 20% under market for this title, here's the data." Sentence two, the lever: the thing you're threatening to make it happen — "...and I have an offer I'll take if you don't match."

Now check two things. First: is sentence one strong enough to stand alone? If your claim genuinely holds, sentence two is decoration — and if it's decoration, ask why you need it. Second, and this is the tell: read sentence two aloud and ask, is every word in it true right now? Not "could become true." True tonight. A real offer you'd actually take is a fact. An offer you engineered purely to brandish is a forgery wearing a fact's clothes.

If sentence one is weak, you don't have a leverage problem — you have a case you haven't built yet. If sentence one is strong and sentence two is false, you're not negotiating your worth. You're rehearsing a threat you'll have to keep re-pulling on someone who used to trust your word.

The plain ask can lose. The manufactured one can only be won the same way twice.

Want the full eight-lens read on your actual number — the one with your manager's name in it? Walk your raise dilemma through the door. Or browse how other career forks scored, and see the gallery of committed verdicts.

So which sentence is your plan actually leaning on — the claim you could say to Reema's face, or the fork you built so you'd never have to?