Why Your Stock Is in the Wrong Stores

Every modern retailer with more than a handful of stores knows the feeling. You check the sales report and size 7 is sold out in Sandton. You check inventory and there are fourteen units sitting in Pretoria, untouched. By the time someone notices, calls the store, arranges a transfer, and gets the stock moved, the customer has walked across the mall to your competitor.

This is not an edge case. It is the default state of multi-branch modern retail.

The scale of the problem

Stock misallocation is the single largest preventable margin leak in South African modern retail. Conservative estimates put the cost at 3-8% of revenue for a typical mid-market chain with 50+ stores. For a retailer doing R500 million in annual revenue, that is R15-40 million in lost margin every year.

The problem compounds across three dimensions:

• Lost full-price sales. When a customer wants size 7 and it is not on the shelf, that sale is gone. Not deferred. Gone.
• Unnecessary markdowns. The size 7 sitting in Pretoria will eventually get marked down at 30-50% off because it has been sitting too long. That same unit could have sold at full price in Sandton.
• Stockout bias in future planning. The Sandton store recorded zero sales for size 7 during the stockout period. Your planning system reads this as low demand and allocates fewer next season. The cycle deepens.

That third point deserves its own discussion -- we cover it in depth in The Stockout Bias Problem.

Why it happens

Stock misallocation is not a failure of effort. Merchandise planners work hard. The problem is structural.

Allocation is a one-shot decision

In most South African retail operations, initial allocation happens once at the start of a season. A planner looks at last season's sales, applies a growth factor, and pushes stock out to stores. This decision is made weeks or months before the season starts, based on historical patterns that may no longer hold.

For importers, the disconnect is even worse. The buying decision was made four to six months before the stock arrives. By the time those containers clear customs and hit the warehouse, market conditions, competitor positioning, and consumer preferences may have all shifted. Yet the allocation formula has not changed.

The fundamental issue: demand is dynamic, but allocation is static.

Manual rebalancing does not scale

When a planner spots a mismatch, the typical process is:

1. Export inventory data to Excel
2. Compare stock levels across stores
3. Identify potential transfers
4. Call or email store managers to arrange moves
5. Track the transfer manually

For a 50-store chain carrying 2,000 SKUs, the number of possible transfers is in the millions. No human team can evaluate even a fraction of these opportunities. The result is that only the most obvious mismatches get caught, usually too late.

ERP systems were not built for this

South African mid-market retailers typically run on ERPs like Posibolt, which handle transactions and inventory tracking well. But these systems were not designed to answer the question: "Where should this stock be right now?"

They tell you where stock is. They do not tell you where stock should be. They can generate a stock report that shows size 7 units at every store. They cannot tell you which stores have genuine demand for size 7, which stores have surplus, and whether moving 12 units from Pretoria to Sandton will generate more gross profit than the freight cost of getting them there.

This is not a criticism of ERPs. They are systems of record, not systems of intelligence. The gap between "where is the stock" and "where should the stock be" is exactly where a merchandise intelligence platform adds value.

Distance and logistics add real cost

South Africa's geography makes transfers expensive. Moving a box from Johannesburg to Cape Town costs R200-500 in freight. A naive rebalancing approach that ignores distance will recommend moves that cost more than they earn. This is why spreadsheet-based rebalancing often fails -- it cannot factor in the true cost-benefit of every potential move.

The size run problem compounds it

Stock misallocation is not just about individual units being in the wrong place. It is about size runs breaking down.

A shoe style with sizes 4, 6, 10, and 11 on the shelf is not a sellable offering. Customers browse the rack, see their size is missing, and leave. The remaining stock is effectively dead until it gets marked down. Meanwhile, those missing sizes -- 5, 7, 8, 9 -- might be sitting in another store where the style is not performing.

The pilot data showed that broken size runs have a sell-through velocity 60-75% lower than complete runs. That is not a gradual decline. It is a cliff. Once the size run breaks, the remaining stock is on a fast track to markdown.

This means that a single well-chosen transfer -- one that reconstitutes a complete size run at a store with genuine demand -- can be worth more than ten random unit movements. But identifying those opportunities across a 50-store network with 2,000 SKUs requires computational power that no spreadsheet can deliver.

What good looks like

Effective stock rebalancing requires three things that manual processes cannot deliver:

Speed. The ability to evaluate every possible transfer across the entire network in seconds, not days. Demand patterns shift weekly; your rebalancing cadence needs to keep pace. A weekly planning cycle in Excel takes days. By the time decisions are made, the data is already stale.

Precision. Every recommended move must account for demand signals, distance costs, minimum quantities, broken size runs, and dozens of other constraints. A move that reconstitutes a sellable size run in a high-velocity store is worth far more than randomly redistributing units. The engine needs to understand not just where stock is, but where it should be -- and whether the cost of getting it there is justified by the margin it will recover.

Accountability. Every move should be measurable. Not "we think this helped" but "this specific transfer generated R1,247 in gross profit uplift." Closed-loop measurement is the difference between faith-based merchandising and evidence-based merchandising. Without it, you are making expensive logistics decisions based on hope.

This is what Replenify's rebalancing engine was built to deliver. A 3-pass architecture that evaluates warehouse-to-retail pushes, inter-store transfers with distance-aware routing, and sweep consolidation -- all in 265 milliseconds.

The cost of waiting

Every day that stock sits in the wrong store, two things happen: you lose a potential full-price sale at the store that needs it, and the clock ticks closer to markdown season at the store that has it. The margin impact is not linear. It accelerates as the season progresses.

Consider the timeline for a typical winter season launch in South African footwear:

• Weeks 1-4: Stock arrives, allocation is fresh. Mismatches are small. Opportunities for correction are large.
• Weeks 5-8: Demand patterns diverge from forecasts. Size runs start breaking. The window for full-price recovery narrows.
• Weeks 9-12: Fast sellers are stocked out at high-demand stores. Slow stores are sitting on surplus. Markdown pressure builds.
• Weeks 13+: Markdown season. Everything that was in the wrong place gets discounted. The margin that could have been saved through earlier rebalancing is gone.

The retailers who rebalance in weeks 1-4 capture the most value. The ones who wait until weeks 9-12 are already too late for many of their best-selling items.

In our pilot with a 28-store South African footwear retailer, Replenify processed 28,348 inventory snapshots and generated 1,572 intelligent moves that produced R745,898 in measured gross profit uplift. These were not projections. This was actual, closed-loop measured impact. Read the full story in The R745K Question.

The question is not whether your stock is in the wrong stores. It almost certainly is. The question is how much margin you are willing to leave on the table before you do something about it.

Where to start

If you recognise these patterns in your business, three resources will help:

1. How 3-Pass Rebalancing Works -- a technical deep dive into the engine architecture that makes this possible.
2. Inventory Optimization: The Complete Guide -- a comprehensive look at the strategies and technologies that drive modern inventory management for modern retail.
3. The Stockout Bias Problem -- understanding the feedback loop that makes misallocation worse every season.

Stock misallocation is a solvable problem. The technology exists. The question is whether you are ready to stop managing it with spreadsheets and phone calls.