Warehouse metrics
A reading of the shelf,not a count of what went wrong.
Pick-face fill is a state: how much of a forward pick location's capacity carries stock at a given moment, typically read at the start of a wave. WareBee compares live occupancy at each location against its rated capacity continuously, so the reading is a point-in-time level rather than a count of failed picks or a measure of how long a restock took to arrive.
See what's keeping the pick face topped up
What it is
A reading taken at the moment that matters.
Pick-face fill is a state, not an event or a duration: how much of a forward pick location's capacity carries stock at a given moment, most usefully read at the start of a wave, before demand has had any chance to draw it down. WareBee reconciles live occupancy at each location against its rated capacity on an ongoing basis, so the reading at any point is a snapshot of what's actually there, not an estimate carried forward from the last time someone walked the aisle.
The level moves with how much a single replenishment round adds relative to what the wave ahead will draw, where the reorder point that triggers a top-up is set, and how fast the wave itself pulls stock back down once picking starts. A location topped up generously right before a light wave reads full for hours; the same top-up ahead of a heavy wave on a fast-moving SKU can read thin again within the same shift, and neither reading is wrong — they're describing genuinely different demand hitting the same starting level.
Two pick faces can carry the same reading at wave start and mean two different things: one sitting exactly where a thin, fast-turning zone always sits, the other actually running short for an operation with more room to carry stock. WareBee separates the two by checking the reading against a peer set matched on order profile, SKU count and footprint, so what counts as comfortably full gets decided by operations shaped like yours rather than a fixed level nobody set out to justify.
What moves this number
What keeps the shelf topped up — or doesn't.
Each one is measured on your digital twin from data you already generate, so a thin reading comes with a cause attached before it becomes a stalled pick.
How much a single round actually adds
A top-up sized for an average wave can leave a location thin ahead of a heavier one.
The twin checks the quantity a replenishment round adds against what the coming wave is actually expected to draw, rather than a fixed round size set once and reused regardless of demand, so the level after a top-up reflects the wave ahead of it, not an average wave that may not be this one.
Where the reorder point is set
A trigger set too close to empty leaves little buffer before the next wave arrives.
Reorder points calibrated against a location's actual demand curve rather than a single value applied everywhere give a wave more runway before a thin reading turns into a stalled pick, especially on fast-moving SKUs where the curve is steep.
How fast the wave draws the level back down
A high-velocity SKU pulls its location's fill down far faster than a slow mover in the same size slot.
The twin tracks draw-down rate per location against the wave's pick pace, so a location that looked comfortably full at release can be flagged as trending thin well before it actually empties, rather than only being noticed once a picker meets the gap.
Whether the slot was ever sized for the demand
A location too small for a SKU's velocity starts every cycle closer to empty than one sized correctly.
Slotting decisions set the ceiling a fill reading can ever reach at that location, so a persistently thin reading can be a capacity mismatch rather than a replenishment timing problem, and the twin flags which of the two is actually driving it.
Full at wave start is the only reading that matters.
A fill reading taken mid-shift tells you less than the same reading taken at wave release, because that's the moment a wave's demand is about to start drawing it down. WareBee checks fill against rated capacity at wave start for every forward location, then tracks the draw-down as the wave runs, so a location trending toward thin surfaces while there's still time to act rather than once a pick has already stalled.
From there it's a round-size and trigger question, not a hiring one. Replenishment quantity and reorder point are set against what the coming wave is actually expected to draw, and a location whose slot was never sized for its SKU's velocity gets flagged as a capacity fix rather than a timing one. Changes are proven on the digital twin against your own wave history before a revised trigger or round size reaches a live shift.
- Fill checked against rated capacity at wave start, then tracked as the wave draws it down
- Round size and reorder point set against the wave ahead, not a fixed value reused everywhere
- Slotting flagged separately from timing when a location was never sized for its SKU's velocity
Questions
Pick-face fill, read at the right moment.
What exactly does 'pick-face fill' measure?
How much of a forward pick location's capacity carries stock right now, checked against that location's rated capacity rather than an assumed full slot. It's a state, not a total accumulated over a shift — the reading describes the location as it stands at the moment it's taken, most usefully at wave start, before demand has drawn it down at all.
How is pick-face fill different from a pick-face stockout?
Fill is a level; a stockout is an event. Fill tells you how much stock a location was carrying at a given moment, most usefully wave start. A stockout is logged separately, the instant a pick is attempted against a location with nothing left to take. A location can show a healthy fill reading at release and still log a stockout later in the same wave once demand has drawn it down past the point anyone checked.
Does a low fill reading always mean a stockout is coming?
Not necessarily. A location can read thin at wave start and still clear the wave without a single failed pick, if the demand behind it is light enough that even a small buffer holds. The reading only becomes a genuine warning once it's checked against the pace of the wave about to draw on it — a thin reading ahead of a heavy, fast-moving wave is a different situation from the same reading ahead of a light one.
Couldn't a fill reading just be checked against one flat target?
A fill reading needs a comparison because 'enough' changes with what the location is about to feed — a level that's comfortable on a slow-turning SKU can be dangerously thin on a fast one occupying an identical slot size. WareBee places your location inside a peer set built on order profile, SKU count and footprint, so a thin-looking reading only counts as thin once it's set beside others feeding a genuinely similar pace of demand — the same slot size on a slower SKU would call for the opposite verdict.
Is pick-face fill measured continuously, or only at a point in time?
Both, in different ways. WareBee reconciles live occupancy against rated capacity on an ongoing basis, so a current reading is always available. But the reading that matters most for planning a wave is the one taken at release, before that wave's demand has touched it — everything the twin does with fill afterwards is tracking how that starting level draws down as the wave runs, not replacing the starting reading itself.
Does a rebalanced replenishment trigger reach the WMS?
Yes. Once a revised round size or reorder point is approved, it goes out as the replenishment tasks your WMS already runs, sized and timed against the wave the location is actually about to feed rather than an average wave. Only the locations whose fill pattern actually needed adjusting change — the rest keep their existing settings — and occupancy data flows back afterwards, so the next reading reflects what's genuinely on the shelf.