How to detect trend and seasonality in sales data?

Trend is detected by computing the Pearson correlation between time index and sales values. Correlation above 0.3 indicates upward trend, below -0.3 indicates downward trend. Seasonality is detected by computing autocorrelation at different lag periods and identifying the lag with the highest autocorrelation above 0.3 threshold. Common seasonal periods are 4 (quarterly), 7 (weekly in daily data), 12 (monthly), and 52 (weekly in yearly data).

Which forecasting model should I use for my data?

Model selection depends on the time series patterns present. Data with no trend and no seasonality suits Simple Moving Average or Simple Exponential Smoothing. Data with trend but no seasonality suits Holt Linear Trend method. Data with both trend and seasonality suits Holt-Winters method. Data with only seasonality suits Seasonal Naive. The Mathnal tool auto-races 7+ models against your data and selects the best by lowest MAPE on a holdout set.

What is a good MAPE for demand forecasting?

MAPE benchmarks vary by industry: below 10% is excellent, 10–20% is good, 20–30% is fair for volatile demand, and above 30% typically needs intervention. Stable FMCG products often achieve 5–15%, while fashion or new products may see 30–50%. However, MAPE should be assessed alongside forecast bias — a forecast with 15% MAPE but consistent 10% over-forecast bias is more damaging than 20% MAPE with zero bias.

How to calculate forecast bias?

Forecast Bias = Average of (Actual - Forecast). Positive bias means chronic under-forecasting (stockout risk), negative bias means chronic over-forecasting (excess inventory). Bias should ideally be within plus or minus 5% of average demand. Track bias alongside the tracking signal (CFE/MAD) — when tracking signal exceeds plus or minus 4, the forecast is systematically biased.

What is a tracking signal in demand forecasting?

A tracking signal is the cumulative forecast error divided by the mean absolute deviation (MAD). It measures whether forecast errors are random or systematic. Values between –4 and +4 indicate a well-calibrated forecast. Beyond ±4, the model has statistically significant bias and should be recalibrated or replaced.

Demand Forecasting Diagnostic Tool

Upload sales data or enter manually. Detect trend, seasonality, and variability. Auto-select the best forecasting model. Get forecast volumes, MAPE, and bias — instantly.

What this tool does

Mathnal's Forecasting Diagnostic Tool auto-detects demand patterns (trend, seasonality, intermittence, level shifts), recommends an appropriate model (ARIMA / ETS / XGBoost / LSTM / Prophet) and computes MAPE, wMAPE, RMSE, MAE, MASE, bias and tracking signal — instantly, browser-based, no signup.

📂 Upload Excel / CSV

✍️ Enter Manually

Results loaded. Clear all to start fresh.

Upload Sales Data Excel / CSV

Upload a file with columns: SKU Name, Category, Sub-Category, Warehouse, Period 1, Period 2, ... Period N (12–30 periods). Max 10 SKUs per file.

Period Type *

Periods to Forecast *

📁

Click to upload or drag & drop

.xlsx, .xls, .csv · Max 10 SKUs · 12–30 periods of sales data

→ Or load sample data to try it out

Enter Sales Data Manually Up to 10 SKUs

Period Type *

Number of Periods *

Number of SKUs *

Periods to Forecast *

📰 Latest from Mathnal Insights

Supply Chain Intelligence You Can't Afford to Miss

Issue #10 · Crisis Analysis

Demand Forecasting Diagnostic Tool

Upload Sales Data Excel / CSV

Enter Sales Data Manually Up to 10 SKUs

Related Mathnal Resources

Supply Chain Intelligence You Can't Afford to Miss

Iran-Israel-USA War: Quantified Supply Chain Impact & 12 Mitigation Strategies

ESG Compliance & Scope 3: Genuinely Compliant or Exposed to Greenwashing Risk?

Supply Chain Risk & Resilience Simulator (SCRRS)