inventory-reorder-calculator（库存重新订购计算器）

Inventory Reorder Calculator

Estimate when to reorder and how much to buy before stock risk turns into lost revenue or excess inventory.

This skill goes beyond plugging numbers into a formula. It applies a structured inventory-planning workflow — demand analysis, lead-time modeling, safety stock calibration, and cash-vs-stockout tradeoff framing — to produce reorder recommendations operators can actually act on.

Quick Reference

Solves

Most ecommerce teams get reorder planning wrong not because they lack data, but because:

• Gut-feel ordering — buying "about the same as last time" without modeling demand changes
• Ignoring lead-time variability — treating supplier lead time as fixed when it fluctuates 20–50%
• No safety stock logic — either zero buffer (stockouts) or massive buffer (cash drag)
• Formula without context — calculating ROP without explaining what drives it or when it breaks
• Missing constraints — ignoring MOQs, carton multiples, storage limits, or cash flow
• No risk framing — presenting a single number without showing the stockout vs overstock tradeoff
• Static calculations — one-time number with no guidance on when to recalculate

Goal: Produce a reorder recommendation that an ops lead, buyer, or founder can act on today — with the math shown, assumptions visible, and risks framed.

Use when

• You need a practical reorder point for a SKU or product group
• Demand is growing, volatile, or seasonal
• Lead time is long or unreliable
• You want to reduce stockouts without overbuying cash-intensive inventory
• A team needs to explain reorder logic to a buyer, founder, or ops lead
• You're setting up initial reorder rules for a new product or supplier
• Transitioning from gut-feel ordering to data-informed replenishment

Do not use when

• You need a full supply-chain planning system or ERP implementation
• Historical demand is too weak to support even rough assumptions
• Supplier constraints are unknown and nobody can estimate them
• The task is warehouse slotting or operations design rather than reorder planning
• You need multi-echelon or multi-warehouse optimization

Inputs

Gather these inputs — mark any gaps explicitly:

Demand data:

• Average daily or weekly unit sales (last 30/60/90 days)
• Demand trend direction (growing / stable / declining)
• Demand variability (standard deviation of daily sales, or coefficient of variation)
• Known seasonality, promotions, or launches upcoming
• Historical stockout periods (to adjust demand estimates)

Supply data:

• Supplier average lead time (order-to-receipt, in days)
• Lead-time variability (best case / worst case / std dev)
• Minimum order quantity (MOQ)
• Carton multiples or packaging constraints
• Supplier reliability notes (late shipment frequency, quality issues)

Inventory data:

• Current on-hand stock (units)
• Current in-transit stock (units, ETA)
• Storage capacity constraints
• Current inventory carrying cost (% of COGS per year, or $/unit/month)

Business context:

• Target service level (e.g., 95%, 98%, 99%)
• Stockout cost estimate (lost margin + customer impact)
• Cash flow constraints or budget limits
• Review cycle / reorder cadence (daily / weekly / monthly)
• Product lifecycle stage (launch / growth / mature / clearance)

See references/safety-stock-guide.md for service level and z-score tables. See references/demand-analysis-guide.md for demand estimation methods.

Workflow

1. Analyze demand pattern

Before calculating anything, understand the demand signal:

Average daily demand: [X] units/day
Demand std deviation: [σd] units/day
Trend: [growing / stable / declining at Y% per period]
Seasonality: [none / seasonal with peak in Z months]
Data quality: [strong (90+ days) / moderate (30–90 days) / weak (<30 days)]

If demand data is weak, flag this prominently — the entire calculation depends on this input.

See references/demand-analysis-guide.md for methods to handle trend, seasonality, and sparse data.

2. Model lead time

Supplier lead time is rarely constant. Model both average and variability:

Average lead time: [LT] days
Lead time std deviation: [σLT] days
Best case: [X] days
Worst case: [Y] days
Data source: [supplier quote / historical POs / assumption]

Rule: If lead time is based on a supplier quote alone (not historical data), add 20–30% buffer. Suppliers are optimistic.

3. Calculate safety stock

Safety stock bridges the gap between average expectations and real-world variability:

Method 1: Service-level approach (preferred when data exists)

SS = z × √(LT × σd² + d² × σLT²)

Where:
z = service level z-score (1.65 for 95%, 1.96 for 97.5%, 2.33 for 99%)
LT = average lead time in days
σd = standard deviation of daily demand
d = average daily demand
σLT = standard deviation of lead time in days

Method 2: Days-of-cover heuristic (when data is limited)

SS = average daily demand × safety days

Where safety days = typically 5–14 days depending on:
- Lead time length (longer LT → more safety days)
- Demand variability (higher variability → more safety days)
- Stockout cost (higher cost → more safety days)

See references/safety-stock-guide.md for z-score tables and method selection guidance.

4. Calculate reorder point

ROP = (average daily demand × average lead time) + safety stock
ROP = (d × LT) + SS

Interpret the result: "When on-hand inventory drops to [ROP] units, place a new order."

If in-transit stock exists, use effective inventory position:

Inventory position = on-hand + in-transit - backorders
Trigger reorder when: inventory position ≤ ROP

5. Determine reorder quantity

Basic approach:

Reorder quantity = average daily demand × days of coverage target

Constraint-adjusted approach:

Raw quantity = demand × coverage days
Adjusted for MOQ: max(raw quantity, MOQ)
Adjusted for carton multiple: round up to nearest carton multiple
Adjusted for cash: min(adjusted quantity, budget ÷ unit cost)
Adjusted for storage: min(adjusted quantity, available storage)

EOQ approach (when holding and ordering costs are known):

EOQ = √(2 × annual demand × order cost / holding cost per unit per year)

See references/output-template.md for the complete output format.

6. Frame the risk tradeoffs

Every reorder decision involves tradeoffs. Make them visible:

Quantify when possible:

• "Stockout of [X] days costs ~$[Y] in lost margin"
• "Extra [X] units ties up $[Y] in cash for [Z] weeks"

7. Quality-check the recommendation

Before delivering, verify with assets/reorder-checklist.md:

• Is the demand estimate based on data (not just a guess)?
• Is lead-time variability accounted for?
• Is safety stock calibrated to a service level or risk tolerance?
• Does the reorder quantity respect MOQ and packaging constraints?
• Are cash flow implications visible?
• Are assumptions explicitly stated?
• Is there guidance on when to recalculate?

Output

Return a structured package (see references/output-template.md):

1. Assumptions table

- Every input value with source and confidence level

1. Demand and lead-time model

- Demand stats, trend, variability - Lead time stats and variability

1. Reorder point calculation

- Safety stock with method shown - ROP with formula and plain-English interpretation

1. Reorder quantity recommendation

- Raw quantity and constraint-adjusted quantity - Cash impact estimate

1. Risk scenario table

- Conservative / recommended / aggressive options - Stockout risk and cash tradeoff for each

1. Action items and review triggers

- When to place the next order - When to recalculate (demand shift, supplier change, etc.) - Sensitivity warnings

Quality bar

Strong output should:

• Show the math AND explain it in plain English
• Separate reorder point (when to order) from reorder quantity (how much)
• Account for demand variability, not just averages
• Account for lead-time variability, not just supplier quotes
• Respect real-world constraints (MOQ, cash, storage)
• Frame the stockout vs overstock tradeoff explicitly
• Flag fragile assumptions that could change the recommendation

What "better" looks like

Better output helps the operator act with confidence:

• Knows when to reorder (and understands why that number, not another)
• Knows roughly how much to buy (and sees the tradeoff in buying more or less)
• Sees the cash vs stockout tradeoff in concrete terms
• Understands where lead-time risk changes the answer
• Can explain the decision to a buyer, founder, or ops lead
• Has clear triggers for when to recalculate

Examples

Example 1: Growing DTC skincare brand

Inputs:

• Product: Vitamin C Serum 30ml
• Average daily sales: 42 units/day (last 90 days, growing ~8%/month)
• Daily demand std dev: 12 units
• Unit cost: $8.50, selling price: $29.99
• Supplier lead time: 21 days avg (std dev: 4 days)
• MOQ: 500 units, carton multiple: 50
• Current stock: 890 units, none in transit
• Target service level: 95%
• Review cadence: weekly

Output excerpt:

DEMAND MODEL
Avg daily demand: 42 units → adjusted for growth: 46 units/day (8%/mo trend)
Demand σ: 12 units/day
Data quality: Strong (90 days, consistent)

SAFETY STOCK (Service-level method, z=1.65 for 95%)
SS = 1.65 × √(21 × 12² + 46² × 4²)
SS = 1.65 × √(3024 + 33856)
SS = 1.65 × 192
SS = 317 units → round to 320

REORDER POINT
ROP = (46 × 21) + 320 = 966 + 320 = 1,286 units

⚠️ Current stock (890) is BELOW reorder point. Order now.

REORDER QUANTITY
Target coverage: 30 days post-receipt
Raw qty: 46 × 30 = 1,380 units
Adjusted for MOQ: 1,380 (above 500 MOQ ✓)
Adjusted for carton: 1,400 (nearest 50 multiple)
Cash required: 1,400 × $8.50 = $11,900

RISK SCENARIOS
| Scenario | Order Qty | Stockout Risk | Cash | Coverage |
|---|---|---|---|---|
| Conservative | 1,700 | <2% | $14,450 | 37 days |
| Recommended | 1,400 | ~5% | $11,900 | 30 days |
| Aggressive | 1,100 | ~12% | $9,350 | 24 days |

Example 2: Seasonal product with unreliable supplier

Inputs:

• Product: Insulated water bottle
• Average daily sales: 18 units/day (but seasonal: 30/day in summer, 8/day in winter)
• Current month: April (ramping up)
• Supplier lead time: 35 days avg, range: 28–50 days
• MOQ: 200, unit cost: $6.20
• Current stock: 520, 300 in transit (ETA 2 weeks)

Output excerpt:

DEMAND MODEL
Current avg: 18 units/day
Seasonal forecast (next 60 days): ramping to ~25 units/day
Using forward estimate: 25 units/day
Demand σ: 7 units/day (higher variability due to seasonal transition)

⚠️ LEAD TIME WARNING
Avg LT: 35 days, but range is 28–50 days (σLT ≈ 6 days)
This supplier has high variability — safety stock must account for this.

SAFETY STOCK (z=1.65 for 95%)
SS = 1.65 × √(35 × 49 + 625 × 36) = 1.65 × √(1715 + 22500) = 1.65 × 156 = 257 units

REORDER POINT
ROP = (25 × 35) + 257 = 875 + 257 = 1,132 units

INVENTORY POSITION
On-hand: 520 + in-transit: 300 = 820
820 < 1,132 → ⚠️ Below ROP. Order immediately.

Days until stockout (no reorder): 520 ÷ 25 = 20.8 days
In-transit arrives in ~14 days → post-arrival: (520 - 350) + 300 = 470 units
470 ÷ 25 = 18.8 more days → ~33 days total before stockout

ACTION: Order now. Lead time of 35 days means new stock arrives just as
current + in-transit runs out. Any delay = stockout during peak season.

Common mistakes

1. Using averages without variability — "We sell 20/day" ignores that some days are 8 and others are 35
2. Trusting supplier lead times — Quoted lead times are best-case; actual delivery is often 20–50% longer
3. Forgetting in-transit inventory — Reordering when stock is low but 1,000 units are already shipping
4. Ignoring MOQ and carton constraints — Calculating a perfect 347-unit order when MOQ is 500
5. No cash flow context — Recommending a $50K order to a business with $30K available
6. Static one-time calculation — Giving a number without saying when it should be recalculated
7. Safety stock = gut feel — Using "2 weeks of safety stock" without connecting it to demand variability
8. Not adjusting for trend — Using historical averages for a product that's growing 15%/month

Resources

• references/output-template.md — Complete structured output template
• references/safety-stock-guide.md — Service levels, z-scores, and safety stock methods
• references/demand-analysis-guide.md — Demand estimation, trend adjustment, and seasonality handling
• assets/reorder-checklist.md — Pre-delivery quality checklist