Start with these five checks.
- Freeze the launch outcome and evaluation window.
- Match category, price, season, channel and exposure.
- Separate demand from stock and fulfillment limits.
- Compare content, placement, promotion and returns.
- Check cannibalization and portfolio role.
Ask for only what can change the answer.
- Product masterAttributes, category, supplier and launch
- ExposureImpressions, placement and store count
- AvailabilityIn-stock time and fulfillment promise
- CommercialsPrice, discount, cost and margin
- BehaviorViews, carts, purchases and returns
- ContextSeason, campaign and comparable launches
Test one reversible move.
Choose one supported execution gap on a current product group, such as missing stock or weak detail-page content. Change only that factor for a randomized or matched subset, then compare contribution margin, conversion and returns. Roll back immediately if the change damages margin, availability elsewhere or customer experience.
A decision your team can use.
- 01A fair peer comparison
- 02A launch-gap decomposition
- 03The first supported lever
- 04A reversible market test
Common questions.
How long should a launch be evaluated?
Use a horizon that fits the category and purchase cycle; state it before comparing products.
How do stockouts affect conversion?
They censor demand and can distort both traffic and purchase rates, so availability must be modeled explicitly.
Does low sales prove a bad product?
No. Low exposure, poor placement, stock, content or price can produce the same result.