Expanding into a new category is genuinely useful for finding growth, and genuinely risky if it's done with the same buying volume as a category you already understand. A bounded test run separates the learning from the risk.
The bounded test-run method
- 1
Set a fixed spend cap before sourcing anything
Decide the maximum total you'll spend on this category before you've bought a single item — a number small enough that a total loss wouldn't meaningfully hurt the business.
- 2
Buy a small, varied sample within that cap
A handful of different items within the category, not one large purchase — variety gives you more signal about what sells than depth in one sub-type.
- 3
List everything and track real outcomes, not predictions
Actual sell-through time, actual realized price versus estimate, actual buyer questions and objections — this is the data the test is for.
- 4
Decide based on what actually happened, not on the two best-case items
Weight the decision by how the whole sample performed, not by the one item that happened to sell fast for a great price.
Reading the results
Interpreting a test batch
| Result | Read |
|---|---|
| Most items sold within a reasonable window near estimated price | Category looks viable — consider scaling gradually, not immediately to full volume |
| Mixed results, some sold well, some are stalling | Narrow to the sub-type that performed, rather than abandoning or fully adopting the whole category |
| Most items are stalling or underselling estimates | Category likely isn't a fit for your market — stop before the spend cap forces the decision |