Blunt version first: doing ABC classification once and trusting it for a year is self-deception. SKU velocity is alive — seasons change, promotions change, customer mix changes, and half of last year's A-items may have slid to C by now. If the slots don't move, pickers walk further. Every extra step is labor cost.
Two years ago I walked a 120,000-square-foot e-commerce warehouse in Ontario. Their ABC dated from opening day and hadn't been touched in three years. I had them pull a heat map: in the longest-travel pick aisle, 40% of the slots held C-items, while the golden zone nearest packing sat full of "former A-items" that hadn't moved in six months. The ops manager looked at it and said he didn't even want to calculate how much money three years of that had cost.
The decay curve: how long does your ABC stay fresh
ABC shelf life tracks directly with your business type. My rules of thumb: fast-moving e-commerce, review quarterly. General 3PL with multiple clients and categories, every six months. Industrial parts and spares with stable velocity, twelve months is fine. Past the expiry date, the gap between the classification and actual velocity keeps widening — the golden zone fills with zombie A-items while new bestsellers languish deep in the aisles.
There's a rough field test for drift: monthly, pull the slot locations of your top 100 SKUs by pick count. If more than 20% of them sit outside the golden zone, your ABC exists in name only — time to reslot. The number doesn't need to be precise; the trend is what matters. Two consecutive months of deterioration, don't hesitate.
Four triggers: any one of them starts a reslot
Beyond the calendar review, four events trigger a reslot. Any one, don't wait.
Sixty days before peak or a major promo. Once the promo SKU list is locked, move promo items forward immediately. Don't wait for the volume spike to start moving freight — by then the whole building is picking and there's no labor or aisle space for relocations.
Losing or landing a major client. Customer mix changed, so SKU velocity mix will follow. After losing a big client, the A-zone built for them sits empty — leave it and you're wasting the golden zone.
New automation goes live. Goods-to-person, AMRs, AS/RS — the whole slotting logic changes. It used to be "minimize human travel"; now it's "minimize robot travel." This is the most overlooked one: the equipment is new, but the slotting strategy is still living in the manual era.
Per-picker UPH slides. After ruling out labor and system factors, a 10%+ drop in UPH means checking slots first. A lot of "productivity decline" isn't a people problem — it's freight in the wrong place.
How to actually execute: don't go for the big bang
The biggest reslotting mistake is the "great leap" — shutting down for three days and flipping the whole building. A live operation can't afford that. My approach is divide and conquer.
First, prioritize: only move the top 20% of SKUs. Eighty percent of reslotting gains come from twenty percent of SKUs — get those right and the rest can wait. Second, hide moves inside natural replenishment: SKUs that need to move forward don't get a dedicated move; they just get replenished into the new slot on the next cycle. Zero move labor. Third, run a transition window: old and new slots both stay valid in the system for two weeks, pick tasks point at the new slot first, and the old slot drops off naturally once it's picked empty.
One last thing: after the reslot, update the slotting policy document and tell every shift lead why things moved. Otherwise, in three months, veteran staff replenishing "by feel" will put everything back where it used to be "handy," and your reslot evaporates. Slotting management was never about freight. It's about people.
