Inventory Metrics — 5 calculators

Turnover (COGS basis) uses cost of goods sold divided by average inventory. This is the conservative figure finance uses — it shows how many times you sold through your stock at cost. Turnover (Sales basis) uses revenue instead of COGS and always looks higher because retail markup inflates revenue. Use COGS when comparing to benchmarks; use Sales when your own historical trend is what matters.

Days to sell = 365 / turnover. A grocery store turning 12× a year holds stock for about 30 days. A furniture retailer turning 2× holds it for ~182 days. Don't compare absolute turnover across industries without this normalisation.

What "good" looks like. Higher is generally better, but a turnover that's much higher than your industry benchmark can mean chronic stockouts. If your apparel business is doing 9× but peers are doing 4×, you may be leaving sales on the table because nobody can find a size.

Limitations. Turnover is blind to gross margin. A 4× turnover on 50% margin is a different business from a 4× turnover on 12% margin — that's why GMROI exists (see Financial Impact).

The curve is the answer. The shaded area to the right of your safety-stock cut-off is the probability you'll stock out during lead time. If the curve is fat (high σ), moving the cut-off right costs you a lot of stock for little extra protection — that's when lead-time reduction beats more safety stock.

The Z-score slider. 90% → Z≈1.28; 95% → 1.65; 97.5% → 1.96; 99% → 2.33; 99.9% → 3.09. Notice how a 4.9-percentage-point jump from 95% to 99.9% requires roughly double the safety stock for the same σ. That's the famous "last few percent is the expensive part."

When this math is wrong. If your demand is lumpy (B2B orders, not retail), σ stops being a good description and you should be using a service-level approach based on order history percentiles. If your lead time is highly variable (customs, port delays), use the Lead-Time Demand Distribution tool instead, which can include lead-time variance.

Who this is for. Operations and procurement analysts in retail, ecommerce, FMCG distribution. Less useful for project-based or make-to-order manufacturers.

The chart shows total annual cost as a function of order quantity. The minimum of the U-curve is your EOQ. Notice the curve is flat near the optimum — that's intentional, and it's why you don't have to nail the inputs perfectly.

Sensitivity band. If your real order cost is 25% lower or higher than the estimate, the EOQ shifts by roughly 12% in either direction. Compare that to the impact of getting demand wrong by 25% (about 12% EOQ shift in the other direction).

What EOQ doesn't model. Quantity discounts (if your supplier offers tiered pricing, the optimum often is the largest break), storage capacity limits, MOQ constraints, perishability, and supplier reliability. Use this as a starting point, not a procurement contract.

Who this is for. Steady, repeatable demand. If you sell ten of one SKU a day for years, EOQ is your friend. If you sell zero most weeks and 200 in one week, this number is a fiction.

Three reorder points, one decision. Using the best-case lead time as your basis is the most common cause of avoidable stockouts. Using worst case inflates working capital. The expected-case figure with a safety-stock buffer usually lands closest to reality.

If best = expected = worst, your supplier is unusually reliable. Consider negotiating a longer contract at the current lead time rather than building extra stock.

If worst is more than 2× expected, your supplier's variance is the real problem. Use the Lead-Time Variance tool and consider qualifying a backup supplier before adding more safety stock.

ABC thresholds. We use the standard cumulative-value breakpoints: A = top 80% of value, B = next 15%, C = last 5%. That typically lands as ~20% of SKUs being A, 30% B, 50% C. If your A list is bigger than that, you don't have an "A" problem — you have a "too many tail items" problem.

Cycle counting by class. A SKUs cycle-counted monthly, B quarterly, C annually is a defensible default. Anything tighter is theatre.

CCC interpretation. Cash Conversion Cycle = DSI + DSO − DPO. A positive number means you're funding your suppliers and customers; a negative number means your suppliers and customers are funding you. Most retail and grocery operations are negative. If yours is positive and growing, look at payables terms before you look at inventory.

Limitations. ABC by value is blind to volume. Two SKUs with identical annual value but one selling 1,000 units/year and one selling 5 will sit in the same class — yet the slow-mover is the obsolescence risk. Consider an ABC × XYZ cross (volume × variability) for the next iteration.