The model bill grew faster than the plan price
A helpful feature got used more than I modeled. The margin on the cheap plan is gone. I can cap it, move it up a tier, or raise the price and lose the people who only came for that button.
Comments (6)
I have been adjacent to this. The sentence I would not skip is "The margin on the cheap plan is gone". Finding who it is for mattered more than adding another feature.
Respectfully, "A helpful feature got used more than I modeled" is an observation, not yet a decision. What number would make you reverse it?
Put the feature on the tier that matches the cost before you get attached to the signup chart. I waited, and the chart made the decision harder.
We capped the cheap plan and showed the counter. Complaints came from people who were never going to pay more. The teams who needed it upgraded without a call.
Related discussions
View more- The AI draft is fine and customers still want to edit every line
We added a draft for a status update. People use it and then rewrite it. I am not sure that is failure. It might be the product. I am unsure how to price a feature that is a starting point.
Hiro TanakaDiscussion11 votes7 comments
- The model feature is opt-in and I am leaving it that way
Defaults would make the usage chart prettier. A few customers do not want their text sent anywhere, even for a draft. The opt-in is a sentence on the screen and a quieter graph. I can live with the graph.
Hiro TanakaBuild in Public7 votes3 comments
- A prompt change altered the tone and a customer noticed
I improved a prompt on a Friday. Their drafts became shorter and they thought the product had lost data. Nothing was deleted. The voice changed. I had no note that explained it.
Mei LinDiscussion6 votes4 comments
- A flag I meant to keep internal has been on for everyone
The draft feature was supposed to be a private test. A config change shipped it. The results are actually fine. The process is not. I want a checklist that does not depend on me remembering.