60 Days Before Peak Season: Pre-Slotting Seasonal SKUs and Re-running ABC

Here's the conclusion up front: 60 days before peak season is the one window of the year when you can move slotting around at scale and get away with it. After that — once you're inside 30 days — anyone who proposes "re-running ABC" should be told to wait until next year. Three things need to happen: pre-slot seasonal winners into the golden pick zone, re-run ABC on forecast data (not history), and lock the slotting two weeks out so the only moves allowed are replenishments. Do all three and a 20%+ peak-day productivity gain is a conservative estimate. Skip them, and you'll watch your pickers run a marathon on Black Friday.

Why exactly 60 days? Run the two numbers yourself. First, the workload math: in a mid-size 30k–50k sq ft facility, seasonal SKUs usually make up 15%–25% of the total SKU count — several hundred locations to move. Moving a slot isn't just carrying product; it's system transactions, verification, exception handling. A solid crew relocates 80–120 slots per day cleanly — and I mean cleanly, not "dumped it there." That's 2–3 weeks of execution, plus data pulling and planning up front. Sixty days is exactly right, not a day extra. Second, the risk math: while slots are mid-move, system records and physical stock are most likely to disagree. You need at least two weeks of locked-down verification. Still moving slots 30 days out is like changing your answers the night before the exam. I've seen this exact failure three times, and it ends the same way every time.

The counterintuitive part: never re-run ABC on historical sales before peak

This is the single most important judgment in this article, and the most common place to crash.

ABC classification exists to answer one question: "who will be busiest in the coming period?" Normally, the last 90 days of sales predict the next 90 days just fine. Peak season destroys that assumption. Holiday string lights sell 200 units all year and 5,000 in one week. Desktop heaters sit dead until October, then stock out daily. Run ABC on trailing 90-day sales and these SKUs stay parked in the C zone forever — and on peak day your pickers sprint from the back of the building to the pack station just to grab them.

The fix: feed ABC with forecast demand for the peak period. The forecast can be rough — last year's same-period volume times this year's buy plan, or just ask sales for a Top 50 list. Even a gut-feel forecast beats precise historical data before peak. Remember this line: before peak, ABC ranks the future, not the past.

Two projects: one heaven, one hell

In September 2023, an e-commerce seller of holiday decor and small appliances near Los Angeles — about a 40k sq ft owned warehouse — brought us in for pre-peak slotting. We pulled two years of October–December outbound data, layered on that year's purchase plan, and re-ran ABC on a forecast basis. Ugly result: 37 SKUs — string lights, desktop heaters, Christmas wrapping paper — sat in the C zone by history but belonged in A by forecast. We pre-slotted all 37 into the two flow-rack rows closest to the pack stations, and pushed peak-season laggards (summer fan accessories and the like) back to B. Execution took 11 working days, then 3 days of full-zone verification after lockdown. Peak-week actuals: average pick travel per order dropped from roughly 120 meters to roughly 80, and pick UPH rose about a quarter. The warehouse manager told me afterward it was the first Black Friday in his 8 years when he "dared to leave the office on time."

The cautionary tale is a 3PL client from 2024, fulfilling for several e-commerce brands. They remembered ABC two weeks before peak and fired off 400+ move tasks in a panic. On Black Friday morning, a third of those moves were still open: the system showed SKUs in A locations while the physical product still sat in C. Pickers scanned empty slots, exceptions piled up, and in the first two hours, 60+ orders missed carrier cutoff — the escalation calls went straight to the owner's phone. Recovery took a full week: freeze all moves, roll back to the pre-move location snapshot, and reconcile system versus physical, slot by slot. I've seen this exact failure three times. The script never changes — "we still have time," then the move collides with peak day, and no consultant on earth can save you.

The 60-day countdown checklist, week by week

Use this table as-is. Every step names its deliverable — the "night-before-go-live checklist" mindset starts right here.

Week 1 (T-60 to T-53): Pull data, set scope

1. Export 2–3 years of peak-season (Oct–Dec) outbound data by SKU from the WMS — at minimum SKU, description, units shipped, and order lines.
2. Get this year's buy plan and hero-product list from sales/purchasing. Flag "new seasonal SKUs launched this year" — they don't exist in history and are the easiest to miss.
3. Freeze routine moves: no non-urgent relocations until pre-slotting is done. Get written confirmation from the ops lead; a verbal promise doesn't count.
4. Deliverable: draft seasonal SKU candidate list.

Week 2 (T-53 to T-46): Re-run ABC on forecast

1. ABC input = last year's same-period volume × this year's buy factor (use 1.0 if you don't have one); new products use sales estimates. Don't obsess over precision — direction matters more than decimals before peak.
2. Produce three lists: SKUs that should enter A, SKUs that should leave A (peak-season laggards), and SKUs that stay put.
3. Walk the floor with the list: is there enough A-zone capacity? Any bulky SKUs that won't fit the flow racks? A floor walk catches half the problems.
4. Deliverable: ABC change list, each line with SKU, current slot, target slot, and reason.

Week 3 (T-46 to T-39): Build the pre-slot plan, size the work

1. Prioritize seasonal SKUs on three dimensions: forecast peak volume × cube/weight × affinity (what gets ordered together with the winners goes nearby).
2. Define the golden pick zone with a number, not a feeling: slots within a 2-minute walk of the pack stations, or the first row of the pick path. Measure it with a tape.
3. Size the workload: total slots ÷ slots per day (conservative 80–120) = working days, plus a 30% buffer. Over 15 working days means your scope is too big — cut it.
4. Deliverable: execution plan down to which slots move each day, who moves them, who verifies.

Weeks 4–5 (T-39 to T-25): Execute the moves

1. Move "out of A" first, then "into A" — clear the space before the new tenants arrive.
2. End-of-day rule: every move task from that day must be 100% transacted in the system plus a 10% cycle-count sample. Anything unposted gets fixed that night, never carried to tomorrow.
3. Schedule bulky and odd-shaped SKUs in separate shifts from smalls — mixing them halves throughput.
4. Deliverable: daily move report (completed / exceptions / remaining).

Week 6 (T-25 to T-18): Verify and top up

1. Full-zone count of the pre-slot area: 100% system-to-physical match, and chase every single discrepancy to the root.
2. Raise replenishment max levels for A-zone SKUs temporarily — running dry on an A SKU during peak wipes out everything you just gained.
3. Don't forget consumables: stage cartons, bubble wrap, and tape near the pack stations at 1.5× average daily peak volume. Every year somebody perfects the slotting and then discovers the cartons are stacked at the far end of the building on peak day.
4. Deliverable: pre-lockdown verification report, signed by the ops lead.

Weeks 7–8 (T-18 to T-0): Lock it down

1. Issue the lockdown order: no non-replenishment moves from today. Change the move-order approver in the WMS to the warehouse manager personally — want to move something? Come convince me first.
2. One job only: watch A-zone inventory levels and replenish the moment anything dips below its trigger.
3. Three days before peak: print the paper "hot SKU location cheat sheet" and tape it at the pick-zone entrance. If the system goes down, that sheet buys you two hours.

Five failure modes I've actually watched happen

1. Running ABC on historical sales. Covered above. I've seen this exact failure three times — seasonal SKUs wiped out in the C zone every time.
2. Moves half-done when peak starts. The root cause is always the same: drift early, panic late. Countermeasure: hard-wire end-of-day closeout into the plan; escalate same-day, never accumulate.
3. Pre-slotting the SKUs but not the consumables or the labor. Perfect slots mean nothing with no cartons at the pack station and a short night shift. The plan needs consumables and staffing sections, period.
4. Moving the slots but not the wave strategy. SKUs land in A, but wave templates still cluster by the old groupings, so pick paths stay tangled. Review wave templates while you re-run ABC.
5. A lockdown order nobody enforces. During lockdown, only the warehouse manager can approve a move in the WMS — 99% of the "just a few slots" requests die at that step.

One last piece of advice: pre-peak slotting is won on the calendar, not on muscle. Get the calendar right and you watch the dashboards from your office on peak day. Get it wrong and you're running the aisles with the pickers. The nice thing about leading with the conclusion is there's only ever one of them — go pull the data now, not next week.

By Lin Xiao, WMS Implementation Consultant