Altfield was not a product idea. It was a job we could not get done on a Shopify store with about 50,000 SKUs, and nothing we tried did it the way we wanted.
On a catalogue that size the SEO fields are never done. There is always a batch that arrived from a supplier with no meta description, no alt text and no category, and there is never a free afternoon to sit and type them in.
We went looking for something to do it for us. What we found split two ways. Some apps wrote straight into the store and showed you the result afterwards, which is fine until the run is wrong across nine hundred products. Others metered the work in credits priced per field, which makes a catalogue-wide pass something you have to do arithmetic on before you can start.
What we actually wanted was narrow. Read the products we already have. Propose the missing values. Let us read them beside the product they belong to. Write only the ones we tick. Keep the previous value of everything, so a run we regret is one click to put back. Nothing else.
So we built it, and ran it on our own 50,000 products before it was sold to anybody. That is also why the free plan has every feature in it and why reverting costs nothing and needs no AI call — those are the parts we would have wanted to test ourselves before trusting an app with a catalogue the business depends on.
There is no mode where it writes without showing you. We would not have switched that on either.
So you can price a catalogue-wide pass before you start it, rather than after.
On every plan, including after you stop paying. An undo you have to pay for is not an undo.
Run it on your own catalogue and judge it on the output.