First: establish product truth
Pick a representative slice of your catalog, including a simple item, a variant family and an out-of-stock item. Write down the authoritative values and compare them with the public storefront. Fix identity and offer disagreements before expanding distribution. Keep the evidence and the owner of each repair.
Next: choose one surface
Select a target based on your customers and documented eligibility. Confirm its data path and what your current platform already covers. Map any remaining capabilities using Stack Finder. Evaluate candidate tools only after the gap is specific enough to test. A broad vendor claim should become a scoped question in a proof-of-fit review.
Then: build a repeatable observation loop
Record feed delivery, public data checks and shopping observations separately. Set a manual recheck date. If repeated use proves valuable, define what an automated alert should detect and who will act on it. V1’s Watch follows market evidence; it does not silently schedule store monitoring.