Your advisor's pet metric became the number you ship on

Your advisor says "weekly actives" on three calls in a row.

By the fourth, your standup opens with that number, your deck leads with it, and your roadmap is quietly sorted by how much each item moves it. You didn't decide this. You drifted into it, one nod at a time, because the person saying "weekly actives" is the person whose respect you most want to keep.

The dilemma

This is a composite — the kind of founder read we see often in this category, stitched together from many, not one real person.

Arjun runs a two-sided scheduling tool, solo, nine months in. His lead advisor — call him Prakash — has two exits behind him and a calendar most founders would sell a kidney for. Prakash gives Arjun forty-five minutes every other week, and in every session the conversation lands on the same place: weekly active users. Growth is the story. Growth is the number the next round will underwrite. Prakash is not wrong about any of it.

So Arjun ships to the number. He added a streak counter. He added a nudge that pings you if you haven't opened the app in three days. He added a "you'll lose your saved slots" banner. Weekly actives went up eleven percent in a month. Prakash was pleased.

Here's the detail Arjun hasn't said out loud, even to himself: the features that moved the number are the ones that make people feel slightly worse. The streak isn't a feature, it's a leash. The three-day ping lands on people who were happily done. The banner manufactures a loss that doesn't exist. Nobody is lying. Users really are opening the app more. And the product, measured by anything Arjun actually believes in, is getting worse.

He's not asking whether the number is true. He's asking whether it's his.

The read

Four of the eight lenses bite hard here. Watch two of them disagree.

Svadharma: whose company is this, really?

The duty lens asks a narrow question — not "what's the best metric" but "whose call is this to make." Arjun is the founder. The vision of what the product is for is his to hold. An advisor advises; he doesn't own the answer to "what matters here." What this lens sees: Arjun has quietly transferred the definition of success to the one voice he most wants to keep. That's not delegation. That's abdication wearing delegation's clothes. It pulls hard toward reclaiming the number — or at least owning that he, not Prakash, is the one who answers for it when it's up and the product is hollow.

Wider welfare: who the number acts on

The welfare lens asks who carries the weight of the action. The streak, the ping, the loss-banner — these land on users who were content. The lens sees a growth engine built out of small manufactured discomforts. People opening the app more isn't the same as people being better served by it. This lens pulls the same direction as svadharma: the features winning the number are losing the room.

Honesty: a true number is not the same as the truth

The honesty lens is tight here, and this is where it gets interesting. Arjun isn't lying. Weekly actives really did rise eleven percent. The deck isn't fraudulent. But a true number presented as the whole story is its own kind of falsehood. The lens sees a founder who could stand in front of an investor, say one accurate sentence, and leave everyone in the room believing something false about the health of the product. Technically clean. Substantively bent.

Detachment: the respect you can't let go of

Now the disagreement. The detachment lens asks what you're gripping so tightly it's steering your hand. And here it points not at the metric — but at Prakash's approval. Arjun is attached to being the founder a respected operator is pleased with. That's the engine. And this is where detachment and the welfare lens pull in genuinely opposite directions in Arjun's head: welfare says the honest move is to change course, but every fiber of the approval-grip says changing course means telling Prakash his north star is wrong, and risking the look that follows. The tension isn't between the right thing and the easy thing. It's between the right thing and the relationship Arjun has mistaken for the right thing.

Run discernment — viveka — across all four, and the knot untangles. Discernment separates "the advice is good" from "the advice is mine to obey." Prakash's point about growth can be entirely sound and the trap-features can be entirely Arjun's to answer for. Both true. Only one of them has Arjun's name on it.

The verdict: keep the metric, kill the mechanisms that trap, and tell Prakash plainly that growth bought with manufactured loss is a number he can't defend — so he's chasing it a different way. The leash goes. The authorship comes home.

This is the part a general AI chatbot can't do for you, and it's worth being precise about why. Ask one to pressure-test "should I optimize for my advisor's metric," and it will build you a tidy case — then if you push back, build the opposite case just as warmly, because agreeableness is a documented behavior of these systems: they lean toward the side you're leaning. KarmaLens asks for the deciding facts first — is the number manufactured or earned, who carries the cost, what exactly did the advisor say versus what you inferred — and only then makes one committed call, with the real options laid beside it and what each one costs. The call doesn't renegotiate itself the moment you flinch. That's what makes it a read you can act on instead of an echo.

The takeaway: the signature test

You don't need us for this one. Tonight, write your headline metric on a line. Then finish this sentence out loud, in the room, as if your toughest user is across the table:

"This number went up because we ___."

Listen to how the blank ends. If it names something the user wanted — faster booking, fewer no-shows, a job done — the metric is serving the product. Sign it. If it names something you did to them — a streak they're afraid to break, a ping they didn't ask for, a loss you invented — the number is yours to answer for, and right now you can't.

Then the harder follow-up, the one your advisor genuinely cannot answer for you: if this number went up and the product got worse, would you still call that a win? If the honest answer is no, the metric isn't the problem — the mechanism moving it is. Keep the number. Change what feeds it. Browse the gallery for more founder reads where a respected voice and an honest one pull apart, or read others in this category. An advisor can point at a destination. Only the founder signs for the road taken to get there.