Nobody sets out to leave a warehouse arranged for a range it no longer sells. It happens because go-live is a snapshot. The pick faces were sized, the fixed bins were assigned and the replenishment quantities were entered against the products, order profiles and pack sizes of that year, and then the team moved on to the next project. This is one of the places where optimising a live EWM system pays back, because the fix is mostly settings and moves, not a rebuild.
What it looks like on the floor.
You do not need a report to see this. Walk the pick area at the busiest hour of the day and look for four things.
Empty pick faces at peak.
A picker arrives at a face and it is bare. The stock exists, it is on a pallet two levels up or in the next aisle, but it has not come down. The picker either waits, skips the line and returns later, or reports it to a supervisor. Each of those is a small delay, and at peak they stack.
Overflow locations that never go away.
When a product has outgrown its home, the surplus lands somewhere else: the end of an aisle, a spare bay, a floor position with a handwritten label. Overflow is a polite word for stock sitting in more than one place, and it means the system bin and the real picking habit have separated.
Pickers waiting on forklifts.
Count how often a picker stands idle while a driver is somewhere else. If the forklift is mostly doing emergency replenishment, chasing faces that have already run dry, then it is not doing planned work, and put-away and loading start to queue behind it.
Fast movers in the far aisles.
Look at where the products people order every day actually live. A product that moved slowly three years ago may now be one of the top sellers and still sit at the back of the building, while a former fast mover occupies the best bays near the dispatch end. Pickers walk the difference every shift.
Any one of these is normal for a week. All four in the same zone, month after month, means the layout and the settings have stopped matching the demand. It also belongs in the same family as the slow regression we describe in labour drift: nothing broke, the building just aged.
How replenishment works in EWM.
In plain terms, replenishment is the job of moving stock from reserve storage (pallet racking, bulk) down to the pick face before the face runs out. EWM supports several ways of triggering that, and most sites use more than one.
Planned replenishment runs against minimum and maximum quantities held for each product at its pick location. When the planning run finds the face below its minimum, it creates the work to top it up towards the maximum. Order-related replenishment looks at what is on the open orders instead of a fixed level, and brings down what the coming work needs. Automatic replenishment fires as stock drops during picking, so a face that crosses its threshold generates the move without waiting for a scheduled run. Direct replenishment is the last-resort version: a picker finds the face empty, and the system creates the move there and then so the line can still be completed.
The first three are the ones you want doing the work. A site where direct replenishment is the main way faces get filled is a site where the planned settings have stopped catching demand. Direct replenishment is a safety net, not a plan.
The settings that drift.
Three things go stale, and they go stale together.
Minimum and maximum quantities come first. They were set from the order profile and the pack sizes at go-live. If a product now ships in a different carton, or the typical order line has doubled, a minimum that once covered a couple of hours of picking may now cover forty minutes. The face runs dry before the replenishment task can be completed, even though nothing is misconfigured in a technical sense.
Fixed bin assignments are next. Many sites give each pick product a dedicated bin so pickers learn where things are. That is a good habit, until the bin is too small for today's volume, or sits in a poor position for a product that has become a top mover. A fixed bin is a promise that this product belongs here, and promises should be reviewed when the range changes.
Timing against wave release is the one people overlook. If waves are released in blocks and replenishment is planned on a different rhythm, the faces can be full at the start of a wave and bare by the middle of it. The question is not only how much sits at the face, but whether the top-up happens ahead of the work that will consume it. If you run waves, the replenishment run needs to sit before the release, not alongside it.
None of this needs a custom program. It is the standard behaviour of the product doing what it was told, long after the instruction stopped being right. If somebody proposes a custom workaround for a stale setting, read the Z-code arrogance tax before you approve it.
Slotting and rearrangement.
Slotting is the discipline of putting each product in the best place for how it moves: fast movers close to dispatch and at a comfortable pick height, slow movers further out, heavy items low, and products that are ordered together near each other. Rearrangement is the act of moving stock into those better locations once you know where they are, and it is driven by movement data rather than by someone's memory of what sells.
A licensing note, because it catches people out. SAP places slotting in the advanced EWM scope, and the exact boundary varies by release and by contract. Check yours before you plan around the feature, and our note on basic or advanced EWM explains how to read that line. Luminar does not sell licences, so we have no reason to push you either way.
The good news is that you do not need the feature to capture most of the value. Take the pick history from the system (the warehouse monitor, /SCWM/MON, is a starting point for what is open and moving, and your reporting covers the rest), rank the products by how often they are picked, and look at the top group. Work out by hand where those products sit today, where they should sit, and what their faces need in minimum and maximum terms. A manual re-slot of the top movers is a spreadsheet, a walk and a handful of bin changes. On many sites that is where most of the gain lives, because a small number of products account for a large share of the picking.
A pick face is a bet on what the range will look like. Review the bet when the range changes, not when the face runs dry.
Re-slotting a live warehouse.
This is where good ideas go wrong. A warehouse that is shipping every day cannot be rearranged in one big weekend without a real risk of lost stock, mismatched labels and a Monday morning full of pickers hunting for products that were moved on Saturday.
Move in small steps instead. Pick a handful of products, and change their faces and their minimum and maximum quantities together, so the system and the shelf agree. Do it on a quiet shift, with the supervisor briefed and the physical move confirmed in the system before the next picking window opens. Check the result for a few days: are the faces holding, has the walking dropped, has direct replenishment fallen? Then take the next group. If a move turns out wrong, it is a small one to reverse.
Keep a short log of what moved and why. Six months from now, somebody will ask why a product lives where it does, and the answer should not depend on who happens to remember. The same applies to the settings: write down the reasoning behind the new minimum and maximum quantities so the next review starts from evidence, not guesswork. Reviewing them after the first busy period of a year, and after any significant range change, is a sensible rhythm. If hypercare was handled well, as we describe in our hypercare note, you will already have a baseline to measure against.
How Luminar approaches stale pick faces.
We are independent and we do not sell the build or the licence. A senior consultant spends a day on your floor at the hours that hurt, watching which faces run dry, where overflow collects and how far your fastest products travel, then sits with your inventory planner and shift leads to compare it with what the system holds. You get a plain-language memo in five working days: which faces and settings to change first, what can be done in standard EWM without custom code, whether the advanced slotting feature would add anything on your site, and what a sensible sequence of small moves looks like. No proposal attached. If the answer is a spreadsheet and a few quiet shifts, that is what the memo will say.