Start with a reproducible failure
A long catalog needs a coherent route through its items. Visitors should be able to share a later page and return without depending only on earlier button presses.
Use a small controlled test to separate a symptom from its cause. Record the starting state, affected URL and observed outcome so that another person can repeat the check.
Collect evidence before changing anything
Record item IDs on two adjacent pages. Repeat against the same data version and check gaps and duplicates. Add an item, then distinguish expected shifts from sorting without a stable tie-breaker.
Fictional example: Two products share a publication date. The database changes their order between requests, showing one on both pages while another disappears.
| Check | Expected outcome |
|---|---|
| Starting condition | The same stable dataset produces different page boundaries. |
| Implementation | Record item IDs on two adjacent pages. Repeat against the same data version and check gaps and duplicates. Add an item, then distinguish expected shifts from sorting without a stable tie-breaker. |
| Verification | Adjacent pages cover the intended items and a shared page link works in a new session. |
Fix the cause and repeat the test
Add a unique secondary sort key. Rapidly changing lists may need cursor-based retrieval; decide shareable URLs and indexing separately.
Repeat the same test after the change and add a nearby case that already worked. One successful case establishes that task in those conditions; it does not certify the whole website.
What the test establishes
Adjacent pages cover the intended items and a shared page link works in a new session.