What to check before you trust a sync app with your catalogue
Every sync app looks the same on the listing page: fast, automatic, reliable. The differences show up in the situations nobody demos - a cancelled plan, a revoked permission, a half-finished write. Here are the questions that separate the two, and they are all answerable in a ten-minute trial.
What happens when you press stop?
Ask it plainly: if a write is already in flight and you disconnect, does that write still land? A lot of apps check permission once, before a batch, and then run the batch to the end. The screen says syncing stopped, the other store says otherwise, and both are telling the truth about different moments.
Test it: start a sync on a large-ish catalogue, press disconnect while it runs, then look at the receiving store. Anything that appears after your click is a promise the app could not keep.
What happens when your plan ends?
Plans are usually enforced at the moment you pick products - and then never again. Downgrade, and the selection you made on the expensive plan often keeps syncing on the cheap one. It sounds like a gift; it is an app that does not know what it is doing, and the same blind spot decides other things too.
Test it: pick more products than the free tier allows while you are on a paid trial, then let the trial end. An honest app draws the line and tells you where it drew it. It should not silently keep going, and it should not silently throw your selection away either.
Does it tell you when it did less than you asked?
Add a whole collection of a hundred products to a plan that covers twenty-five. The reply should say both numbers. «Added 25 products» on its own is the dangerous kind of true: you will believe the collection is synced, and seventy-five products will quietly live in one store only.
The general rule: an app that caps, truncates or skips anything owes you a sentence about it in the moment, not a support reply three weeks later.
What does it do when it is unsure?
Two products with the same SKU. Two variants with the same option values. A product that matches by handle but not by barcode. The honest answer to all three is «stop and ask», and it is also the least impressive one in a demo.
Guessing is what creates duplicates, and duplicates are the most expensive way a sync fails: they split inventory, reporting and search results, and cleaning them up is manual work nobody planned.
Can it prove the stores still match?
Copying is easy to demo and easy to trust. Staying correct three months later is the actual job, and it needs a scheduled comparison you did not have to remember to run. Ask what the report names: a number is not a report. Product, field, and what each store says - that is a report.
Where we stand on these
We have failed several of these ourselves and fixed them: a plan limit that survived a downgrade, a bulk add that did not mention the remainder, a stop button that raced a write in flight. We list them because an app that has never found a bug like this in itself has not looked.
What sync apps can and cannot do · Why duplicates appear · What a sync audit is