Better prompts,
better results.

Operator hands your request to a specialist that starts fresh each time. The more explicit you are up front, the less context gets lost. Put the marketplace, date range, ASIN, and filters all in your first message rather than drip-feeding them.

Filters worth including up front:

Keep separate jobs in separate chats — sourcing in one, repricing in another, label creation in a third. Each chat starts fresh with no memory of past sessions, so mixing workflows muddies results. When you want to adjust something, restate the full request rather than saying "do it again."

If you don't specify, Operator defaults to US — a common cause of empty or surprising results. Say "on Amazon UK" or "on Amazon DE" in your message when you mean a different one.

For anything beyond a handful of rows, ask for an export. Operator produces a downloadable file you can open in Excel or Google Sheets.

Some jobs take time — reverse searches, large sourcing runs, batch creation, big exports. Start them, keep working, and come back. If something looks stuck, switch to another chat and return later; it may still be processing.

When a prompt gives strong results, save it. Swap the retailer, category, or date and run it again. Good prompts become repeatable systems for you and your team.

Operator can explain its own features and recommend a starting point. The suggested prompts are designed to kick off useful workflows quickly.

Operator's output is only as good as your account data. Before running anything analytical, make sure your inputs are in place — missing data produces misleading results.

Sourcing data moves fast and isn't always perfectly current. Treat Operator's finds as strong leads, not final decisions.