You grandfathered the price. The new cohort pays double.

You wrote the word forever in a changelog at 1 a.m. because you needed the signups and it felt generous.

Now the new pricing page says double, your runway is a number you check too often, and there are forty people paying a rate that no longer covers their support tickets.

And you've started drafting an email with a subject line like "Legacy pricing is being retired." You know it isn't, exactly. That's the part you haven't said out loud.

The dilemma

This is a composite — the kind of knot we see over and over in this category, assembled from many founders, not one real person.

Maya runs a two-year-old scheduling tool, solo except for a part-time contractor. Early on, to get the first hundred users off a competitor, she promised: "$19/mo, locked for life, no matter what we charge later." She meant it. It was also the only lever a nobody had.

The product got heavier. Infrastructure, a support load, a feature set those early users lean on hardest. New signups now pay $34, and even that's tight. The forty remaining lifers at $19 are, by her own spreadsheet, the cohort she loses money on every month.

Three doors.

Door one: honor the lock. They paid in trust when trust was all she had. Eat the margin, raise new cohorts openly, carry the forty forever.

Door two: raise it openly. Tell the forty the truth — "I can't sustain $19, here's why, here's a fair legacy rate, and here's a clean way to leave with your data if you'd rather." A real choice, made in daylight.

Door three: engineer the nudge. Send the "legacy pricing is ending" email that implies a policy, a deadline, an industry inevitability — none of which is strictly true — and let most of them migrate without noticing they were moved.

The question Maya won't type into the box: do I resent the people who trusted me first?

The read

Four lenses actually bite here. Watch two of them go to war.

Honesty: do the words match the world?

The honesty lens asks one flat question of door three: is "legacy pricing is ending" true? It isn't. There is no policy ending it except Maya's wish that it would. The email dresses a decision as an event — "the market changed," "this is just how it is now" — so the person reading it blames the weather instead of the founder. That's not a hard conversation avoided. That's a false sentence sent to the exact people who believed her last true one. Honesty doesn't flinch: door three is a lie with a UI.

Duty: what did the role actually promise?

Here's where it gets interesting, because duty doesn't automatically side against the raise. The duty lens asks what Maya owes and to whom. She made a specific promise — a word given when a word was all she could give. Walking away from an obligation that is genuinely yours isn't freedom; it's a different kind of debt, paid later and with interest. So duty leans hard toward door one: the lock was a word given, honor it. But duty also sees the forty aren't the only people in the room. There's a contractor paid from this revenue, and a hundred-plus newer users whose product survives only if the business does. A duty to early trust and a duty to ongoing solvency are both real. This is the tension: honesty says the nudge is a lie, and duty says the lock may genuinely be unpayable. Both can be true. Neither licenses the other.

Motive: what engine is the email running on?

The motive lens looks under the hood of door three and finds the question Maya buried. The "legacy pricing is ending" framing isn't primarily about margin — margin is fixed by door two just as well, and more cleanly. The nudge's actual payload is making the forty complicit in their own re-pricing so Maya never has to face them. That's not survival. That's resentment wearing a product update's clothes. When the real engine is "I'm tired of the people who got the cheap deal," dressing it as policy doesn't launder it — it just hides the bill from the one person who'll pay it later, which is her.

Non-harm: who can't undo their move?

The non-harm lens asks who absorbs the cost and whether they can reverse it. In door two, the forty get a true sentence and a real exit — they can stay at a fair rate, move to the new plan, or leave with their data. Harm is named and bounded. In door three, the harm is that someone reorganized their billing around a fabricated deadline; when they find out — and in a small market, one of them always finds out — what breaks isn't a price. It's the thing that made the early promise worth anything in the first place.

The call: raise the price in daylight — door two. Say the old rate is no longer sustainable, offer a fair legacy tier, and give a clean way out. Kill the "ending" email entirely; there is no policy, only a decision, and she signs it.

This is the part worth watching, because a general AI chatbot would likely have handed Maya whichever door she leaned toward — framing door three as "strategic communication" if she arrived sounding tired, or door one as "integrity" if she arrived sounding guilty. Agreeableness is a known failure mode of these assistants; they tend to optimize for your assent, not your situation. KarmaLens asks for the facts that move the answer first — is the lock actually unpayable, or just annoying? — then makes one committed call and lays the other doors beside it with their real cost, so you can argue with the verdict without it quietly re-pricing itself to flatter you. And where a verse genuinely fits, it quotes the exact text; it won't paraphrase scripture to win the point.

What you can run tonight

Free, no account, before you touch your billing settings. Call it the event-or-decision test.

Write the one sentence you're planning to send the people you promised. Then circle every noun in it that implies something happened to you — "pricing is ending," "the market shifted," "we're aligning with industry standards."

For each circled phrase, finish this: "The actual decision underneath this is: I decided to ___."

If the two sentences say the same thing, your email is honest — send it. If the circled version makes you sound like a passenger and the "I decided" version makes you sound like the driver, you've written a decision disguised as an event. The gap between them is exactly how much of your promise you're trying to launder.

Then one last line: "I'd prefer the event version because facing them directly would cost me ___." If the blank is "revenue," that's a real constraint — price it openly. If the blank is "having to admit I resent the people who got in cheap," the nudge was never about the money.

Browse more founder dilemmas read this way, or see how committed calls look across different knots. And tonight: is the email you've drafted describing what happened, or deciding what you wish had?