You'd take the acqui-hire. Your team gets nothing.

The term sheet is on your second monitor. The acquirer wants you and your two staff engineers. The consideration is real money — life-changing, retention-package money — routed to you three as compensation, not as a purchase of the company.

Which means the cap table gets what a distressed cap table gets. Your first four hires, the ones who took 30% below market because you sat across a bad coffee and promised them the upside was coming, net roughly nothing.

The math is clean. That's the part that keeps you up.

The dilemma

This is a composite — the kind of founder decision we see often enough in this category to stitch into one honest example. No real company.

Call the founder Arjun. Three years in, a seed round, a product that found a shape but never found a market big enough. Runway is seven months and thinning. An enterprise buyer offers an acqui-hire: Arjun plus his two senior engineers join for a package that vests over two years and totals more than any of them have earned in their lives. The buyer does not want the product, the brand, or the four early employees who built the thing that made Arjun worth hiring.

The offer is structured as employment, not acquisition, precisely so the payout skips the preference stack and the option pool. That is not a loophole. It is the deal. The acquirer's counsel designed it that way, and Arjun's counsel confirmed there is nothing to negotiate on the employees' behalf that the buyer is obligated to hear.

The four early hires — one of whom left a stable role after a fifteen-minute call where Arjun said "I need you to bet on me" — will get a polite email and a final paycheck. Arjun can carve out something from his own package. It would be real to them and a rounding error to him. He hasn't decided whether to.

Walking away likely means a slow bleed to zero for all seven people, including the four. The ugly detail: part of Arjun wants the deal to close fast, before anyone can ask him a question he doesn't have a clean answer to.

The read

Four of the eight lenses actually bite here. We'll walk those.

Duty (−1)

Duty asks what your specific role and your specific promises obligate you to do — not what a generic good person would do. Arjun's role is not "employer." It is the person who converted four people's caution into a bet using his own word as collateral. The below-market pay was the premium they paid; the upside was the policy he sold. The deal pays out that upside to everyone except the people who bought the policy. That is not a market outcome that happened to him. It is a distribution he is choosing to sign. Duty doesn't demand he torch a life-changing exit. It demands he stop pretending the cap table made this call. He did. Score sits negative — not deeply, because the debt is discharge-able — but negative until he acts on it.

Detachment (+3)

Detachment separates what you owe from the outcome you can't control. And here it pulls hard the other way. Arjun cannot control that the market is thin, that the buyer values three heads and no product, or that the preference stack works the way it works. The slow bleed to zero is not a punishment he can convert into justice by refusing the deal. Declining out of guilt would harm the four and the three and him, and buy nobody anything but Arjun's clean conscience. Detachment scores high because the honest read is: the exit itself is not the wrong. The exit is a reasonable response to conditions he didn't author.

Here is the tension, stated plainly. Duty says you owe the four something you're arranging not to pay. Detachment says you can't fix their outcome by wrecking everyone's. Both are true. The lazy founder resolves this by picking one. "It's just business" is detachment eating duty. "I owe them everything" is duty eating detachment. Neither is the read.

Motive (−2)

Motive checks the engine under the reasonable-sounding sentence. Arjun's stated reason is "protecting my seniors and myself from a zero." Defensible. But he told us the quiet part: he wants the deal to close before anyone can put a question to him. That is not a reason to take the deal. It is a reason to take it a particular way — fast, quiet, no carve-out conversation. The speed isn't diligence. It's avoidance wearing efficiency's jacket. Motive doesn't sink the exit. It flags the manner. And the manner is where the actual choice lives.

Wider welfare (+1)

Wider welfare counts everyone the action lands on. Take the deal: three people are made whole, four get an email, and a carve-out — funded from Arjun's package, painful to him, meaningful to them — can move four people from "nothing" to "acknowledged." Refuse the deal: seven people ride the runway into the ground. The welfare column favors closing, but only if the carve-out is on the table. Without it, welfare barely clears zero. With it, the same exit becomes the version Arjun can say out loud.

The verdict: Take the acqui-hire — the exit isn't the wrong — but the deal is only clean if Arjun funds a real carve-out for the four from his own consideration and tells them himself before signing, not after.

Notice what just happened. Two lenses openly disagreed, and the tool did not average them into mush or hand you back your own lean. Ask a general AI chatbot instead and watch it lean where you lean: type "I have to protect my team" and it eulogizes the four; type "this is just how cap tables work" and it hands you absolution. Agreeableness is what these systems are tuned to do — the read you get is often a mirror of the prompt you paid it in. KarmaLens runs eight fixed lenses through a deterministic aggregation: the score for duty doesn't soften because you sound sad about it, and the verdict doesn't renegotiate itself to make closing fast feel better. When it quotes scripture it quotes it verbatim, cited inline, so the standard isn't something the model improvised to match your mood. You can dislike the call. You cannot talk it down.

The takeaway you can run tonight

When an outcome you can't control (the bleed to zero) tempts you to skip a duty you can discharge (the four), run the two-envelope debt test — no login required.

Envelope one, labeled the outcome: write only what conditions decided for you. The market, the preference stack, the buyer's appetite. Things you'd have faced no matter how good a person you are.

Envelope two, labeled the promise: write the exact sentence you said to the people affected, in your own voice. Arjun's was "I need you to bet on me." Not a paraphrase. The words.

Now ask one question: does my leaning move honor envelope one by quietly rewriting envelope two? If taking the deal requires the promise to have never been made, that's the tell. The fix is almost never "refuse the deal." It's "pay envelope two out of your own pocket, and say it to the people in it while you still can."

Envelope two has a delivery date, and it's before the ink dries. If your plan is to open it after the deal closes, you've already read your own verdict — the delay is the decision.

Want the committed call on your own version of this? Run your founder dilemma through the eight lenses, or read how other founders' calls landed in the gallery.

If the deal closed tomorrow and the four read about it online, which envelope would you wish you'd opened first?