In home and furnishings, the planning window closes long before the selling window opens. Promotional events want submissions weeks or months ahead; forward inventory wants placing before demand, not during it; and the specialists you will need in November cannot be trained in November. Sellers who treat peak as a Q4 activity are competing against sellers who treated it as a Q3 project.
Count backwards from the event
Every marketplace publishes its promotional calendar and submission deadlines. The operational calendar is those dates, worked backwards: funded discounts decided before the submission window, creative and imagery finalised before the discount decision, catalog data verified before the creative — because a promoted listing with broken attributes is paid traffic to a page that cannot convert.
- Deadlines gathered per channel, in one place, owned by one person
- Promotion decisions made against margin, not against the fear of missing the event
- Imagery and A+ content queued a sprint ahead of every submission
- A pre-event catalog check on everything entering a promotion: attributes, stock, lead times, price integrity
Stock allocation carries the year
For seasonal home categories, the allocation decision — how much, of what, placed where — carries a disproportionate share of annual margin. Under-allocate and the event sells out on day two; over-allocate and January is a clearance project. The inputs are unglamorous: honest sell-through history, honest lead times from suppliers, and a decision made on the calendar rather than when the forecast finally feels certain, because it never does.
Staff the surge without owning it
Peak roughly doubles the operational load — more orders, more buyer messages, more returns, more listing fixes at exactly the moment they cost the most. Hiring permanent headcount for a two-month surge is how Q1 layoffs happen. The alternatives are pre-trained overflow capacity, cross-training the existing team on the surge tasks, or accepting degraded service in writing rather than by surprise. Any of the three beats deciding in November.
During the event: monitor, don't renovate
Peak weeks are for a tightened monitoring cadence — stock positions, suppressions, buyer message response times, delivery promise integrity — and for resisting every other change. Catalog restructures, repricing experiments and platform migrations all have a season, and this is not it.
Clearance is a project, not an aftermath
End-of-season clearance has its own economics: markdown laddering, bundle decisions, and the call between carrying cost and margin sacrifice. Planned in advance it recovers real money; improvised in January it recovers whatever is left.