Analysis overview and configuration
| Parameter | Value | _row |
|---|---|---|
| significance_level | 0.05 | significance_level |
| min_promo_orders | 10 | min_promo_orders |
| discount_bins | 0,0.001,0.1,0.2,0.3,0.5,0.8,1.0 | discount_bins |
| use_profit_column | TRUE | use_profit_column |
| time_granularity | month | time_granularity |
| cohens_d_small | 0.2 | cohens_d_small |
| cohens_d_medium | 0.5 | cohens_d_medium |
| cohens_d_large | 0.8 | cohens_d_large |
| alpha | 0.05 | alpha |
This analysis evaluates how promotional discounts impact revenue, profit, and customer behavior across a superstore's product categories and customer segments. The study examines 9,994 orders over 48 months to determine optimal discount strategies that balance revenue growth with profitability. Understanding these trade-offs is critical for pricing strategy and promotional planning.
Data preprocessing and column mapping
| Metric | Value |
|---|---|
| Initial Rows | 9,994 |
| Final Rows | 9,994 |
| Rows Removed | 0 |
| Retention Rate | 100% |
This section documents the data preprocessing pipeline for a promotional discount analysis spanning 9,994 orders across three product categories and customer segments. Perfect data retention (100%) indicates no rows were removed during cleaning, suggesting either exceptionally clean source data or minimal validation criteria applied. Understanding preprocessing decisions is critical because data quality directly impacts the reliability of the profitability conclusions, particularly given the analysis reveals significant profit erosion from discounting.
The perfect retention rate suggests the dataset arrived in usable condition with no missing values, duplicates, or outliers requiring removal. However, the absence of documented transformations and train/test methodology raises questions about data validation rigor. Given that the analysis identifies 1,871 loss-making orders (18.7%) and a -110% profit impact from promotions, the lack of explicit data quality checks or anomaly detection could mask underlying data issues that might explain extreme profit swings in certain discount buckets.
The undocumented preprocessing approach limits transparency
| finding | metric | value | recommendation |
|---|---|---|---|
| Promotional Activity | Promo Rate | 52% | Balanced promotional strategy |
| Revenue Impact | Revenue Lift | +2.6% | Promotions increase average order value |
| Profit Impact | Profit Change | -110.0% | Deep discounts eroding margins significantly |
| Breakeven Threshold | First Negative Profit Bucket | 30-50% | Avoid discounts at or above 30-50% |
| Statistical Significance | Revenue t-test p-value | 0.6322 | Revenue difference is not statistically significant |
| Discount-Profit Correlation | Pearson r | -0.220 | Moderate negative relationship - discounts reduce profit |
| Loss-Making Orders | Loss Rate | 18.7% (1871 orders) | Warning: Significant portion of orders are unprofitable |
This analysis evaluates the effectiveness of a promotional discount strategy across 9,994 orders to determine whether discounting drives profitable business outcomes. The assessment examines revenue lift, profit impact, and statistical significance to inform discount policy optimization.
How do discounted orders compare to non-discounted orders across key metrics?
This section evaluates whether promotional discounts successfully drive incremental revenue and volume to justify the margin erosion they cause. By comparing promoted versus non-promoted orders across revenue, quantity, and profit metrics, it reveals the fundamental trade-off: whether the 2.6% revenue lift and 8.3% order volume increase offset the catastrophic 110% profit decline.
Promotions generate higher absolute revenue per order and attract more customers, but this comes at an unsustainable cost. The 30% average discount erodes margins so severely that promoted orders lose money while non-promoted orders remain solidly profitable. The revenue lift is insufficient to compensate for the margin sacrifice, indicating that current promotional strategy is destroying profitability
How are discounts distributed across orders? Histogram of discount percentages
This section reveals how discount percentages are structured across your order portfolio. Understanding discount distribution is critical because it shows whether your promotional strategy follows standardized tiers (suggesting controlled pricing) or exhibits scattered patterns (indicating ad-hoc discounting). This directly impacts profitability analysis, as the overall analysis reveals deep discounts are eroding margins significantly.
Your discount structure follows a deliberate two-tier strategy rather than random pricing. The dominance of no-discount and 15–20% tiers suggests intentional promotional segmentation. However, the presence of extreme discounts (65–80%) alongside the earlier finding that discounts above 30–50% generate
How does discount depth affect revenue and profit? Price sensitivity analysis across discount buckets
This section isolates how discount depth directly impacts profitability by segmenting orders into six discount brackets. It reveals the critical threshold where deeper discounts cease to generate value, helping identify the optimal discount range that balances revenue growth with margin preservation—a core objective for evaluating promotional effectiveness.
The data reveals a fundamental pricing paradox: while moderate discounts (10-20%) maintain profitability ($71.56 average profit), aggressive discounting beyond 30% systematically converts revenue into losses. The
How does discount depth relate to profit? Scatter plot reveals the discount-profit relationship
This scatter plot visualizes the fundamental relationship between discount depth and order-level profitability across 500 sampled transactions. The trend line reveals how average profit systematically declines as discounts increase, providing empirical evidence of whether promotional strategies are destroying or preserving margin. This is critical for evaluating the overall promotional effectiveness question: are discounts generating sufficient volume lift to offset margin erosion?
The scatter plot confirms that while discounts do not eliminate all profitable orders, they systematically shift the profit distribution downward. The -0.22 correlation is moderate but consistent—deeper discounts rarely produce offsetting volume
How are key variables related? Pairwise correlation heatmap
This section quantifies the linear relationships between four critical business variables: Sales, Discount, Profit, and Quantity. Understanding these correlations reveals whether variables move together or in opposition, which is essential for diagnosing why promotional discounts are undermining profitability despite generating modest revenue gains.
The correlation matrix exposes a fundamental disconnect: discounts reduce profit without meaningfully boosting sales or quantity. The weak Sales-Profit correlation (0.48 vs. theoretical 1.0) reflects margin compression from promotional activity. The negligible Sales-Discount and Quantity-Discount correlations suggest discounts are not achieving their intended volume-driving effect, making the 18.
Which orders are losing money? Breakdown of loss-making transactions
| metric | value | detail |
|---|---|---|
| Loss-Making Orders | 1871 | 18.7% of all orders |
| Total Losses | $156,131.29 | Aggregate profit from loss-making orders |
| Avg Discount (Loss Orders) | 48.1% | vs 8.1% for profitable orders |
| Avg Revenue (Loss Orders) | $250.51 | vs $225.1 for profitable orders |
| Worst Loss Category | Furniture | $60,936.11 in losses |
This section identifies orders generating negative profit and quantifies their financial impact. Loss-making orders reveal where discount depth has exceeded product margins, making this critical for understanding the -110% profit impact observed in the promotional analysis and establishing sustainable discount guardrails.
Loss-making orders demonstrate that promotional activity, while driving revenue lift (+2.6%), systematically destroys profitability when discounts exceed 30%. The 48.1% average discount on loss orders directly correlates with the negative profit correlation (-0.22) observed across the dataset. This pattern explains why promoted orders show -$6.
Which product categories tolerate discounts best? Category x discount heatmap reveals discount-resilient categories
This section identifies which product categories tolerate discounts most effectively by examining profit performance across discount levels. Understanding category-specific discount sensitivity is critical because the overall analysis reveals a -110% profit impact from promotions; however, this aggregate masks important variation—some categories may sustain margins better than others, enabling targeted discount strategies rather than blanket policies.
The data reveals heterogeneous category responses to discounting. Technology shows pockets of profitability at moderate discounts (10-20%), while Furniture deteriorates immediately. However, all categories collapse into
Which sub-categories are most and least profitable? Granular profitability ranking
This section identifies which of 17 product sub-categories generate or destroy profitability, revealing granular performance variation masked by category-level aggregates. Understanding sub-category performance is critical for targeted discount policy reform, as the overall analysis shows promotions are eroding profits—but the damage varies significantly by product line.
How do promotional patterns change over time? Monthly trends in promotion rate and effectiveness
This section tracks promotional activity patterns across 48 months (2014–2017) to identify seasonal fluctuations in discount deployment and revenue performance. Understanding how promotion rates and order values vary over time reveals whether discounting strategies are consistent, reactive to demand cycles, or seasonally driven—critical context for evaluating the overall profitability impact of promotions.
The data reveals that while promotional intensity remains relatively constant year-round, the revenue lift from promotions is inconsistent. Early periods (2014) show extreme variance in promoted revenue, while later periods show convergence, indicating that promotional effectiveness may have declined or
How do different customer segments respond to discounts? Segment-level promotional effectiveness
This section examines how three customer segments—Consumer, Corporate, and Home Office—respond differently to promotional discounts. Understanding segment-level promotional effectiveness is critical because uniform discount strategies may destroy profitability in some segments while underperforming in others. These insights reveal whether promotions should be tailored by customer type or applied uniformly.
Across all segments, promotions consistently erode profitability despite modest revenue gains. The
Are the observed differences statistically significant? Hypothesis tests with effect sizes
| test_name | statistic | p_value | effect_size | interpretation |
|---|---|---|---|---|
| Welch's t-test (Revenue) | 0.4786 | 0.6322 | 0.0096 | Not significant (Negligible effect) |
| Welch's t-test (Profit) | -15.74 | 0 | -0.3179 | Significant (Small effect) |
| Chi-square (Category x Discount) | 106 | 0 | Discount distribution differs by category |
This section determines whether observed differences between promoted and non-promoted orders reflect genuine business patterns or random variation. Statistical significance testing combined with effect size measurement reveals both whether differences exist and whether they matter practically—critical for validating promotional strategy effectiveness.
Revenue gains from promotions lack statistical backing—the modest 2.6% increase could easily occur by chance. Conversely, profit losses are both statistically confirmed and practically consistent. This asymmetry reveals the core problem: promotions drive volume and revenue appearance without protecting margins.