Many factories feel the same way: “we have stock, but can’t produce,” or “the file/system says we have it, but I walk to the shelf and it’s not there.” This kind of problem isn’t caused by one person’s mistake alone — it comes from “leaks” that accumulate across the process, from receiving, production, and issuing through to stock counting.

The “90%” in the title suggests “this happens to most factories,” because in the real world inventory accuracy often falls short of what management expects, and many organisations still have gaps in data and work discipline — so they decide on figures that aren’t real, with a domino effect that reaches profit.

This article drills into “where factories commonly go wrong” and “how ERP systematically plugs the leaks.”

1) The painful truth: inaccurate stock numbers are the start of every problem

Inventory accuracy is the closeness between the “system figure” and the “actually counted figure.” If accuracy is low, the impact shows up as shortages (line stoppages / late delivery) or excess (tied-up capital / full space / degradation), plus an inaccurate close.

Many sources indicate that average stock accuracy for many businesses still isn’t high — for example, a NetSuite article cites research finding an average accuracy of around 83% (2024), meaning “a sizeable gap” against the standard of work that demands high confidence in production and procurement planning.

2) Seven leaks that cause factories to mismanage stock (even with skilled, diligent people)

Leak 1: Late recording (Delay) makes data “stale” from the first second

Goods enter the warehouse and get keyed later / materials are issued to the line and posted retroactively. Result: the system shows goods still present, but they’ve already been used → immediate planning error.

Leak 2: Human error from manual keying + many files, many systems

Using paper/LINE chat in parallel leads to mis-keying, duplicates, or skipped steps, and creates “multiple versions of data” (no source of truth) — a classic cause of stock mismatch.

Leak 3: Non-standard item codes / units of measure

For example, the same material has 2 codes, or the purchase unit is “box” but the usage unit is “piece” with no clear conversion defined. Result: balances skew from the master-data level.

Leak 4: No locking of lot / expiry / storage location

When “where is this item” isn’t tied to receive–issue transactions, you get can’t-find-it problems, picking the wrong lot, or FIFO/FEFO that can’t really be done.

Leak 5: Over-issuing / incomplete returns / defects not deducted from the system

Production defects, scrap, or line returns — if there’s no easy, enforceable recording step, stock errors accumulate.

Leak 6: No Cycle Count, or counting once a year and hoping it’s accurate all year

A full annual physical count (closing the business) usually only fixes things “at year end,” while errors keep occurring daily through the year. The cycle-counting concept is continuous counting in rounds, to reduce the accumulation of errors and steadily improve accuracy.

Leak 7: Planning production–procurement on disconnected data

Sales confirms an order, but stock isn’t updated / MRP works off an unreal balance. Result: over-purchasing, or unable to produce on schedule.

3) How ERP plugs stock leaks (not just “going digital,” but “becoming one system”)

The core ERP concept is bringing key organisational processes onto one database, connected from purchasing, warehouse, production, sales, and accounting — reducing rework and reducing multiple versions of truth.

Here are the concrete “mechanisms” by which ERP plugs the leaks:

3.1 Create “one system” for stock transactions (Single Source of Truth)

Receive → stock increases
Issue/sell → stock decreases
Transfer between warehouses/locations → trackable

When every department uses one data set, communication and decisions rest on the same figures — no checking across many files.

3.2 Enforce workflow, reducing skipped steps and retroactive keying

An ERP can mandate “no issue without an issue slip” or “no receipt without a PO/reference document,” so data occurs with the real transaction, reducing the delay that skews stock at the source.

3.3 Support barcode/RFID and real-time shop-floor work

When the point of data capture is on the floor (receive–pick–issue–count), errors from “remember then key later” drop sharply — a commonly recommended approach to raise inventory accuracy.

3.4 Connect the warehouse with planning (MRP/Production) to reduce “can’t produce because of shortages”

If stock is accurate and tied to the BOM/production order, the system helps:

Reserve stock for jobs to be produced
Calculate material requirements
Flag shortage risks in advance

This reduces shop-floor firefighting and expensive emergency purchasing.

3.5 Run targeted Cycle Counts + make improvement routine

An ERP helps schedule cycle counts, record variances, analyse causes, and track whether accuracy is improving — consistent with the practices that explain the benefits and methods of cycle counting.

4) KPIs to track after adopting ERP so it isn’t “install and done”

For ERP to truly solve stock problems, measure clearly, aiming from “symptom” toward “cause,” such as:

Inventory accuracy (%)
Stockout rate / Line stoppage due to material
Overstock / Slow-moving value
Cycle count variance by SKU / by location
Receiving-to-putaway time
Pick accuracy / Issue accuracy

5) Conclusion: a factory missing stock isn’t unusual — letting it become “chronic” is the invisible cost

If stock numbers aren’t accurate, the factory loses money both directly (excess, defects, space) and indirectly (line stoppages, late delivery, lost customer confidence). ERP isn’t a magic pill, but it’s the “structure” that connects data and process, reduces leaks from people–files–multiple systems, and makes work discipline real.

Why "90%" of factories mismanage stock — and how ERP plugs the leaks