I moved PopSmart Clerk from Live to the Graveyard.
The reason is not complicated: clear demand did not show up, and the original Clerk has fully pivoted.
That is enough to close the old version.
Clerk was a real Shopify app. It shipped. It reached the App Store. It tried to be an AI store clerk that read shopper behavior and wrote contextual product nudges instead of showing everyone the same popup.
I still like parts of that idea.
But liking the idea is not demand.
Why the paused state was no longer honest
For a while, Clerk was private while I worked through the pivot. Keeping it on the Live shelf made sense during that window. It showed that the product had shipped, but the current version was not installable.
That state is now too soft.
The original Clerk is no longer the active bet. It has not earned the right to stay beside live products as if it only needs one more pass. The product direction has changed enough that keeping the old entry live would make the record less useful.
So it goes to the Graveyard.
The rule is simple: if the shipped version no longer represents the current product, and if demand was not clear enough to protect it, write the decision down and move it.
The problem was demand, not effort
Clerk had real work inside it: catalog sync, behavior triggers, generated copy, cooldowns, theme controls, and a Shopify app surface.
That matters, but it does not settle the product question.
The product question is whether merchants repeatedly pull for the job. Not whether the idea is clever. Not whether the app is technically real. Not whether the pitch makes sense in a vacuum.
Repeated merchant behavior is the signal.
For Clerk, that signal was not strong enough.
What comes next
I am preparing to announce two Shopify apps at the same time: one new app, and one pivoted app that came out of the Clerk work.
That is the cleaner public step.
The old Clerk record now belongs in the Graveyard, where the decision is visible. The new apps should stand on their own jobs and their own demand, not inherit a vague paused state from a product that did not prove enough pull.
This is the part of building I want the record to show clearly.
Shipping is not validation. A private listing is not momentum. A coherent idea is not demand.
Demand is repeated behavior from the right people.
When it does not show up, the right move is not to keep the story alive. The right move is to pivot, archive the old bet, and make the next one easier to judge.