Overview

Analysis Overview

Revenue Time Series Forecast

Analysis overview and configuration

Configuration

Analysis TypeTime Series Forecast
CompanyExecutive / Finance Team
ObjectiveForecast Sales for the next 6 months with confidence intervals, broken down by product Category, Region, and customer Segment. Board-ready projections showing trend, seasonality, and growth drivers.
Analysis^Order Date | Order Date | Order Date$2026-03-17
Processing Idtest_1773783919
Total Observations9800

Module Parameters

ParameterValue_row
forecast_horizon_months6forecast_horizon_months
confidence_level0.95confidence_level
seasonality_modemultiplicativeseasonality_mode
segment_forecastsTRUEsegment_forecasts
trend_changepointsTRUEtrend_changepoints
board_ready_outputTRUEboard_ready_output
Time Series Forecast analysis for Executive / Finance Team

Interpretation

Purpose

This analysis delivers a 6-month revenue forecast using exponential smoothing (ETS) on 48 months of historical sales data. The objective is to provide board-ready projections with confidence intervals segmented by product category, enabling executive planning and resource allocation decisions.

Key Findings

  • Total Historical Revenue: $2.26M across 9,800 transactions with 100% data retention, ensuring robust model training
  • 6-Month Forecast: $312,513 projected revenue with established confidence bounds (80% and 95% intervals)
  • Model Accuracy (MAPE): 21.92% error rate indicates moderate forecast reliability; RMSE of $9,750 reflects typical monthly deviation
  • Category Performance: Technology leads at 36.6% of revenue with 21.4% YoY growth; Office Supplies shows strongest growth at 31.8% YoY
  • Seasonal Pattern: ETS(A,N,A) model captures additive seasonality without trend, suggesting stable baseline with recurring monthly fluctuations

Interpretation

The forecast reflects consistent seasonal patterns in retail sales without underlying growth or decline trends. The 21.92% MAPE is acceptable for monthly retail forecasting but indicates meaningful uncertainty—confidence intervals widen substantially at the 95% level. Technology and Office Supplies categories drive growth momentum,

Data Preparation

Data Preprocessing

Monthly Aggregation Pipeline

Data preprocessing and column mapping

Data Quality

Initial Rows9800
Final Rows9800
Rows Removed0
Retention Rate100

Data Quality

MetricValue
Initial Rows9,800
Final Rows9,800
Rows Removed0
Retention Rate100%
Processed 9,800 observations, retained 9,800 (100.0%) after cleaning

Interpretation

Purpose

This section documents the data intake and cleaning phase for the revenue forecasting model. Perfect retention (100%) indicates no rows were excluded during preprocessing, which is critical for maintaining the full temporal signal needed for accurate time-series forecasting. This directly supports the board-ready 6-month revenue projections by ensuring the model trains on complete, unfiltered historical data.

Key Findings

  • Initial & Final Rows: 9,800 observations retained unchanged—no data loss during cleaning or validation steps
  • Retention Rate: 100% indicates no filtering for missing dates, non-positive sales, or outliers occurred at the preprocessing stage
  • Train/Test Split: Not explicitly documented, suggesting the full 48-month history (2015–2018) was used for model training without holdout validation
  • Transformations Applied: No explicit transformations noted; data appears to have passed quality gates as-is

Interpretation

The perfect retention rate reflects either exceptionally clean source data or minimal preprocessing rigor. While this preserves all available signal for the ETS(A,N,A) model, it raises questions about outlier handling and data validation. The absence of a formal train/test split means model performance metrics (MAPE: 21.92%, RMSE: 9,749.89) are in-sample estimates rather than true out-of-sample validation

Executive Summary

Executive Summary

Key Findings & Recommendations

Key Metrics

forecast_6m_total
312513.12
historical_total
2261536.78
model_mape
21.92
months_of_history
48
ets_model_used
ETS(A,N,A)
forecast_horizon_months
6

Key Findings

FindingValue
Forecast MethodETS(A,N,A)
History Used48 months
Forecast Horizon6 months
Model Accuracy (MAPE)21.92%
6-Month Revenue Forecast$312,513
Historical Total Revenue$2,261,537
Forecast vs Last Period-45%

Summary

Bottom Line: The ETS(A,N,A) model — trained on 48 months of historical data — projects $312,513 in Sales over the next 6 months, with MAPE of 21.92%.

Key Findings:
• Forecast projects decline of 45% vs the most recent historical period
• Seasonal decomposition reveals recurring annual patterns that should inform timing of promotions and staffing
• Category breakdown shows relative growth rates across product lines
• Model accuracy of 21.92% indicates moderate reliability — supplement with judgment adjustments for known events

Recommendation: Use the 80% confidence interval lower bound for conservative budgeting and the point forecast for target-setting. Refresh the model monthly as new actuals become available to keep confidence intervals narrow and the forecast relevant.

Interpretation

EXECUTIVE SUMMARY

Purpose

This analysis delivers a 6-month revenue forecast using 48 months of historical sales data to support board-level planning and resource allocation. The forecast quantifies expected sales with confidence intervals, enabling data-driven budgeting and risk assessment across the planning horizon.

Key Findings

  • 6-Month Forecast Total: $312,513 — represents expected cumulative revenue through June 2019
  • Historical Baseline: $2,261,537 — total revenue across 48-month training period (2015–2018)
  • Model Accuracy (MAPE): 21.92% — indicates moderate forecast reliability; typical for seasonal business data
  • Seasonal Pattern: Decomposition confirms recurring annual cycles that drive timing variability
  • Category Performance: Technology leads at 36.6% of revenue with 21.4% YoY growth; Office Supplies shows strongest growth at 31.8%

Interpretation

The forecast projects $312,513 in sales over six months, reflecting seasonal demand patterns embedded in four years of historical data. The 21.92% error rate is acceptable for strategic planning but signals that external factors (promotions, market conditions) not captured in the model should inform final budgets. The 80% and 95% confidence intervals provide upper and lower bounds for scenario planning.

Context

Figure 4

Revenue Forecast

Historical + 6-Month Projection with Confidence Intervals

Revenue forecast for the next 6 months using ETS(A,N,A) exponential smoothing. Shaded bands show 80% and 95% confidence intervals.

Interpretation

Purpose

This section presents a 6-month revenue forecast using exponential smoothing to support board-ready financial planning. The model quantifies expected sales with confidence intervals, enabling leadership to understand both the central projection and the range of plausible outcomes for budgeting and resource allocation decisions.

Key Findings

  • 6-Month Forecast Total: $312,513 — represents expected cumulative revenue across the forecast horizon
  • Model Accuracy (MAPE): 21.92% — indicates moderate predictive fit; typical for monthly revenue data with seasonal variation
  • Forecast Trend: 45% decline versus the last historical period, reflecting the model's capture of seasonal patterns and recent momentum
  • Confidence Bands: 80% and 95% intervals widen progressively, with upper bounds reaching ~$89K and lower bounds ~$13K by month 6, reflecting compounding uncertainty over time

Interpretation

The ETS(A,N,A) model identifies additive seasonality without trend in the 48-month historical window. The projected $312,513 total reflects an average monthly forecast of ~$52,086, consistent with the historical mean of $47,668. The widening confidence intervals demonstrate that while near-term forecasts (months 1–2) are relatively precise, longer-horizon projections (months 5–6) carry substantially

Figure 5

Seasonal Decomposition

Trend, Seasonal & Remainder Components

STL seasonal decomposition separating the Sales time series into trend, seasonal, and remainder components.

Interpretation

Purpose

This section decomposes the 48-month revenue series into trend, seasonal, and remainder components to isolate structural growth patterns from predictable cyclical behavior. Understanding these components is essential for the 6-month forecast, as the ETS(A,N,A) model relies on additive seasonality to project board-ready revenue projections with confidence intervals.

Key Findings

  • Observed Sales Range: $4,520–$117,938 monthly (mean $47,668), showing substantial volatility across the 4-year history
  • Remainder Component Volatility: Ranges from –$28,929 to +$11,451, indicating moderate unexplained variation after trend and seasonality removal
  • Decomposition Structure: Four components (Observed, Trend, Seasonal, Remainder) each with 48 monthly observations, enabling clean separation of signal from noise
  • Model Alignment: ETS(A,N,A) specification confirms additive seasonality with no trend component, meaning seasonal patterns are constant in magnitude year-over-year

Interpretation

The decomposition reveals that revenue fluctuations are primarily driven by recurring seasonal patterns rather than a persistent upward or downward trend. The relatively modest remainder values (±$28,929 vs. mean $47,668) suggest that historical seasonality explains most variation, supporting the model

Figure 6

Category Revenue Trends

Monthly Revenue by Product Category

Monthly Sales trends by product Category showing relative growth rates across 3 categories.

Interpretation

Purpose

This section tracks monthly revenue performance across three product categories (Furniture, Office Supplies, Technology) over a 48-month period to identify which lines are driving overall growth and how seasonality patterns differ. Understanding category-level trends is essential for validating the aggregate forecast and informing resource allocation decisions during the 6-month projection period.

Key Findings

  • Total Revenue: $2.26M across 3 categories with mean monthly category revenue of $15,705 (sd: $9,322)
  • Revenue Range: Category monthly sales span $1,072 to $49,409, indicating high volatility and unequal contribution levels
  • Distribution Pattern: Positive skew (0.91) suggests occasional high-performing months, with Technology showing peak performance ($49,409 in Nov 2018)
  • Category Parity: Each category represented equally in the 48-month history (48 observations each), enabling fair trend comparison

Interpretation

The wide revenue spread ($1,072–$49,409) and positive skew reveal that category performance is episodic rather than steady-state. Technology's peak ($49,409) substantially exceeds typical monthly means, suggesting seasonal spikes or promotional events drive category-level volatility. This heterogeneity across categories means the aggregate ETS(A,N,A) forecast captures overall season

Table 7

Forecast Table

Monthly Projections with Confidence Intervals

6-month Sales forecast with 80% and 95% confidence intervals.

periodforecastlower_80upper_80lower_95upper_95
2019-014.57e+043.085e+046.055e+042.3e+046.841e+04
2019-023.664e+042.134e+045.194e+041.325e+046.003e+04
2019-036.526e+044.952e+048.099e+044.119e+048.932e+04
2019-045.322e+043.706e+046.938e+042.851e+047.794e+04
2019-055.938e+044.281e+047.596e+043.403e+048.473e+04
2019-065.231e+043.533e+046.929e+042.634e+047.828e+04

Interpretation

Purpose

This section delivers the core deliverable of the revenue forecasting analysis: a month-by-month projection of sales for the next 6 months with quantified uncertainty bounds. It translates the ETS(A,N,A) model into actionable planning ranges for the Executive/Finance Team, enabling confident budget allocation and capacity decisions across the forecast horizon.

Key Findings

  • 6-Month Forecast Total: $312,513.12 — Represents expected cumulative revenue across Jan–Jun 2019, averaging ~$52,086/month
  • 80% Confidence Interval: Defines the high-probability planning band where actual sales have an 80% likelihood of falling; tighter than the 95% range for operational planning
  • 95% Confidence Interval: Provides conservative downside and upside bounds for risk and opportunity scenarios; wider margins reflect model uncertainty over the 6-month horizon
  • Seasonal Pattern: Forecast incorporates the additive seasonal component identified in decomposition, with March and May showing elevated projections relative to February and June

Interpretation

The forecast reflects 48 months of historical data (2015–2018) with a MAPE of 21.92%, indicating moderate accuracy typical of seasonal business data. The widening confidence intervals across months reflect increasing forecast uncertainty as the horizon extends. The point estimates balance the observed trend (

Table 8

Model Performance

ETS Accuracy Metrics

ETS model accuracy metrics: MAPE, RMSE, MAE, MASE, and AIC for ETS(A,N,A).

metricvalue
MethodETS(A,N,A)
AIC1097.58
MAPE (%)21.92
RMSE9749.89
MAE7799.26
MASE0.64
Periods Used48
Forecast Horizon6 months

Interpretation

Purpose

This section evaluates the statistical accuracy and reliability of the ETS(A,N,A) forecasting model used to project 6-month revenue. Understanding model performance is critical for assessing confidence in the board-ready projections and identifying whether forecast uncertainty bands are appropriately calibrated for executive decision-making.

Key Findings

  • MAPE (21.92%): Moderate forecast accuracy indicating the model's predictions deviate from actuals by roughly one-fifth on average. This reflects inherent sales volatility rather than model failure.
  • RMSE (9,749.89): Absolute prediction error of ~$9,750 per month—meaningful relative to mean sales of $47,668, suggesting forecast intervals should be interpreted conservatively.
  • MAE (7,799.26): Average absolute error slightly lower than RMSE, confirming no extreme outliers dominating model error.
  • AIC (1097.58): Model selection metric; useful only when comparing alternative ETS specifications.
  • ETS(A,N,A) Structure: Additive error and seasonality with no trend component—appropriate for stable revenue with recurring seasonal patterns but no consistent growth direction.

Interpretation

The model captures monthly seasonality effectively but struggles with irregular sales fluctuations, reflected in the 22% MAPE. The absence of

Table 9

Category Summary

Revenue Share & Year-Over-Year Growth

Product Category Sales summary with total Sales, share of total, and year-over-year growth.

Categorytotal_revenuepct_of_totalyoy_growth_pct
Furniture7.287e+0532.28.4
Office Supplies7.054e+0531.231.8
Technology8.275e+0536.621.4

Interpretation

Purpose

This section provides a portfolio-level view of revenue distribution and growth dynamics across the three product categories that comprise the business. Understanding category-level performance is essential for validating the overall 6-month forecast, as the aggregate projection ($312,513 over 6 months) is a blended result of distinct category trajectories with different growth rates and market maturity profiles.

Key Findings

  • Total Historical Revenue: $2,261,537 across 48 months (2015–2018) establishes the baseline for forecast calibration
  • Category Count: 3 distinct product lines (Furniture, Office Supplies, Technology) with varying revenue shares and growth profiles
  • Growth Heterogeneity: YoY growth ranges from 8.4% to 31.8%, indicating divergent category momentum—some mature, others accelerating
  • Revenue Distribution: Relatively balanced portfolio (31–37% per category) reduces single-category dependency risk

Interpretation

The forecast's 21.92% MAPE reflects aggregated category performance; variance in category-level growth rates means some categories will outperform or underperform the blended forecast. Technology's 21.4% YoY growth and Office Supplies' 31.8% growth suggest upside potential, while Furniture's 8.4% growth indicates maturation. This heterog

Want to run this analysis on your own data? Upload CSV — Free Analysis See Pricing