1. Which system owns each field?
Choose a source of truth for product identifiers, prices, descriptions, images and stock. Ownership can differ by field, but it should be explicit. Otherwise, two automatic updates may overwrite each other's changes.
2. What counts as sellable stock?
Separate boxed stock, display samples, reserved items and damaged products. Decide what quantity the website may sell and how adjustments reach the other system. Include an agreed approach to discrepancies and overselling.
3. Are supplier quantities a different kind of availability?
They should be. A manufacturer feed is not proof that an item is in your warehouse or that it can meet a delivery date. Confirm update frequency, order cut-offs, lead-time information and who checks exceptions.
4. How are products matched?
Check identifiers at variant level. A shared model name is not enough when finish, size or voltage differs. Test duplicate SKUs, discontinued products and manufacturer codes that change between data sources.
5. What happens after a sale or return?
Walk through an online order, an in-store sale, a cancellation, a partial refund and a return to a different location. Confirm when stock changes, which location changes and who investigates a failed update.
6. Which pickup and payment features are actually available?
Review the Shopify plan, POS subscription, country, hardware and app requirements for your workflow. For example, preparing pickup orders directly in Shopify POS requires POS Pro; the right setup depends on what the team needs to do and where.
7. How will you notice a failure?
An integration needs an owner, a visible exception queue or alert, and a safe way to recover. Agree what happens when a supplier file is late, a credential expires or an import contains unexpected data. Test those cases before relying on automation.
A good integration is an agreed set of rules with a connection behind it. Document the rules first, then test the everyday work and the exceptions.
Talk about your store