We want to be data-driven in demand and supply planning. This helps us navigate the questions thrown at us during the S&OP (sales and operations planning) meetings.
I share the Spartan credo:
“Sweat more in practice, bleed less in war.”
However, I still get hard questions, and it is with the data that I answer them.
I put together in an infographic the 12 Fundamental Calculations Every Demand and Supply Planner Must Master.
It received great positive feedback.
Here I am expanding more on each metric, including the potential challenges you will encounter.
1. MAPE (Mean Absolute Percentage Error)
Concept: measures the average percentage error between forecasted and actual demand.
Formula: Sum of (Forecast error for time period / Actual sales for that period) / Total number of forecast errors x 100
Key Challenges:
- Very sensitive to low-volume items; a small miss on small demand skews MAPE.
- Ignores direction: it doesn’t show if the team consistently over- or under-forecast. For this, we measure bias.
2. WMAPE (Weighted Mean Absolute Percentage Error)
Concept: A weighted version of MAPE, giving more importance to high-volume items. This makes it more representative for businesses where some products are much bigger than others (a common situation in supply chain and demand planning).
My suggestion is using WMAPE instead of MAPE.
Formula: Sum of (Actual sales for a time period – Forecast sales for same period) / Actual sales for same period
Key Challenges:
- Hides small product issues: small but strategic SKUs might be masked.
If the company has a small but strategic product (like a high-margin or flagship item), it might have big forecasting errors but not move WMAPE much.
This could represent an important miss.
- Demand spikes can still distort WMAPE if weighting isn’t right. Even though WMAPE reduces distortion compared to MAPE, wild demand spikes or anomalies can still mess up the weighted results, especially if not filtered or smoothed properly
3. Forecast Bias
Concept: measures whether forecasts consistently over- or under-predict.
Bias tells you if you’re systematically wrong in one direction.
The focus is not on size of error.
Formula:
Sum of observed forecast errors over multiple periods / Total
number of observed periods
Key Challenges:
- Positive bias (over-forecasting) inflates inventory.
- Negative bias (under-forecasting) causes stockouts.
- Hard to correct bias without deep root cause analysis.
Bias is critical because even a “good” forecast accuracy (like a low WMAPE) can hide a systematic error. If the business consistently over-forecast, it wastes cash on excess inventory.
If the business consistently under-forecast, it loses sales and hurt customer service.
4. FVA (Forecast Value Added)
Concept: Measures improvement or degradation in forecast accuracy after changes.
Formula: Baseline Forecast Error−New Forecast Error
Positive FVA = Improvement. Negative FVA = Made it worse.
Key Challenges:
- Difficult to benchmark baseline forecasts.
- Process noise; too many manual overrides hide real FVA results.
FVA forces you to challenge if your forecasting process is adding real value.
It’s one of the best metrics to justify investments in automation or training.
5. Demand Variability (Coefficient of Variation, CV)
Concept: Assesses how much demand fluctuates relative to the average.
Formula:
Demand Variability = (σ / μ) × 100
σ: Standard deviation of demand
μ: Average demand
Key Challenges:
- High variability products are harder to forecast and plan inventory for.
- External shocks (e.g., promotions, seasonality) can artificially inflate CV.
6. Promotional Lift Factor
Concept: Measures demand increase due to promotions.
Formula:
Lift Factor = Promotional Demand / Baseline Demand
Key Challenges:
- Baseline demand can be hard to establish.
- Cannibalization: promotions for one SKU often steal demand from another.
7. OTIF (On Time In Full)
Concept: Tracks the percentage of orders delivered complete and on time.
Formula:
Shows how many orders are delivered on time in full
OTIF = (Number of On Time In Full Deliveries / Total Number of Deliveries) x 100
Key Challenges:
- Different OTIF definitions across customers make comparison hard.
- Upstream delays (like suppliers) can ruin OTIF even if your team executes well.
8. Inventory Turnover
Concept: How many times inventory is sold or used in a period.
Formula:
(Cost of Goods / Average inventory) X 100
Key Challenges:
- Too high turnover = frequent stockouts.
- Too low turnover = excess and obsolescence risk.
9. Safety Stock
Concept: Extra stock to protect against demand and supply variability.
Formula (basic form):
Safety Stock = Z × σ × √Lead Time
Where
Z = Z-score (based on the desired service level)
Σ = Standard deviation of demand or lead-time demand
Key Challenges:
- Unreliable lead times or demand data cause wrong safety stock settings.
- Too much safety stock ties up cash; too little causes stockouts.
It is important to regularly review and update safety stock settings as demand patterns, lead times, and business conditions change
10. Obsolescence Ratio
Concept: measures percentage of inventory at risk of becoming obsolete.
Formula:
Tracks the portion of inventory that is no longer usable
Obsolescence Ratio = (Obsolete Inventory / Total Inventory) x 100
Key Challenges:
- Poor forecasting or product lifecycle management fuels obsolescence.
- Inconsistent definitions of “obsolete” across the business.
11. Production Yield
Concept: measures how efficiently raw materials are converted into finished goods without defects or waste.
High yield indicates efficient production processes; low yield signals problems like quality issues, machine errors, or process inefficiencies.
Formula (basic form):
Yield = (Good Units Produced / Total Units Produced) x 100
Where:
- Good Units Produced = Units that meet quality standards
- Total Units Produced = Total units that entered the production process
Key Challenges:
- Hidden Defects and Rework: If reworked products are not properly tracked, reported yields may look better than reality.
- Material Waste: Poor yield often leads to higher material usage, driving up costs and reducing margins.
Bottom of Form
12. Capacity Utilization
Concept: measures how much of a plant’s or equipment’s maximum potential output is actually being used.
Formula (basic form):
Utilization = (Actual Output / Maximum Capacity) x 100
Where:
- Actual Output = What is actually produced in a given period
- Maximum Capacity = Theoretical maximum production if running at full capacity without any stoppages (like a dream)
Key Challenges:
- Overloading can cause burnout, equipment wear, and quality issues
- Underutilization means wasted fixed costs and underperformance against investment expectations
What are your favorite calculations? Why?
