Do Customers Prefer Different Products on Different Days? A Data-Driven Look at Buying Patterns

This report presents a data-driven analysis of purchasing behavior at the category level, focused on whether product preferences vary significantly by day of the week. Using a Chi-Square Test of Independence, sales data across seven days and nine product categories were analyzed to determine if there is a statistically significant relationship between product demand and day of purchase.

The objective is to uncover patterns in consumer behavior that may inform inventory allocation, staffing schedules, and promotion planning.

I. Observed vs. Expected Sales Distributions

To perform the chi-square test, actual category sales by day (observed values) were compared against the expected values under the assumption of independence (no relationship between category and weekday).

A. Observed (Actual) Sales Volume Table

Day	Bakery	Branded	Coffee	Coffee Beans	Drinking Chocolate	Flavours	Loose Tea	Packaged Chocolate	Tea
Monday	3,385	92	8,468	244	1,710	923	161	61	6,599
Tuesday	3,222	114	8,304	260	1,607	1,058	178	70	6,389
Wednesday	3,263	122	8,315	261	1,621	963	177	71	6,517
Thursday	3,275	90	8,488	230	1,725	877	175	77	6,717
Friday	3,308	109	8,567	258	1,593	1,032	182	66	6,586
Saturday	3,136	105	8,013	244	1,593	989	160	62	6,208
Sunday	3,207	115	8,261	256	1,619	948	177	80	6,433

B. Expected Sales Volume Table (Assuming Independence)

Day	Bakery	Branded	Coffee	Coffee Beans	Drinking Chocolate	Flavours	Loose Tea	Packaged Chocolate	Tea
Monday	3,308.66	108.42	8,478.62	254.43	1,664.49	985.51	175.62	70.68	6,596.56
Tuesday	3,241.24	106.21	8,305.86	249.25	1,630.57	965.43	172.04	69.24	6,462.15
Wednesday	3,257.75	106.75	8,348.16	250.52	1,638.88	970.35	172.92	69.60	6,495.07
Thursday	3,310.34	108.48	8,482.93	254.56	1,665.33	986.02	175.71	70.72	6,599.91
Friday	3,317.52	108.71	8,501.34	255.12	1,668.95	988.16	176.09	70.87	6,614.24
Saturday	3,135.45	102.75	8,034.77	241.11	1,577.35	933.92	166.43	66.98	6,251.23
Sunday	3,225.04	105.68	8,264.33	248.00	1,622.42	960.61	171.18	68.90	6,429.84

Expected values are derived using the formula:

Expected = (Row Total × Column Total) / Grand Total

II. Hypotheses Tested

To evaluate the relationship between day of the week and product preference, the following hypotheses were tested:

• Null Hypothesis (H₀): Product preference is independent of the day of the week.

• Alternative Hypothesis (H₁): Product preference depends on the day of the week.

  
  

III. Statistical Findings

Metric	Value	Interpretation
Test Type	Chi-Square	Evaluates relationships between categorical variables
Degrees of Freedom	48	Reflects (7 rows − 1) × (9 columns − 1)
p-value	0.0588	Slightly above 0.05; insufficient evidence to reject the null hypothesis
Significance Level	0.05	Standard benchmark for determining statistical significance
Conclusion	Independence	No statistically significant association between product category and day of week

IV. Interpretation and Insights

Despite some visible daily variations in category sales, the test results suggest no statistically significant pattern linking product category to a specific weekday.

• Coffee remains the top-performing category across all days, with daily volumes ranging from 8,013 to 8,567 units.

• Bakery, Tea, and Packaged Chocolate also show consistent sales throughout the week.

• Slightly higher total sales are observed on Fridays and Mondays, though not enough to establish a weekday preference at the category level.

This indicates a stable pattern of customer behavior across the week.

V. Operational Implications

1. Inventory and Workforce Planning

Balanced product demand supports uniform scheduling and stocking throughout the week. Slight increases in total transactions on peak days (e.g., Friday, Monday) may warrant minor resource adjustments.

2. Marketing Strategy

Since no strong weekday trends were detected:

• Promotions should focus on best-selling bundles such as Coffee + Bakery.

• Loyalty programs may target high-volume categories rather than day-specific behavior.

• Emphasis should shift to seasonal or event-driven campaigns over day-based targeting.

3. Opportunities for Further Study

More detailed transaction-level analysis can uncover:

• Product pairing behavior

• Time-of-day demand shifts

• Differences in behavior by customer segments or store format

VI. Conclusion

The chi-square analysis confirms that product preferences are statistically independent of the day of the week. Customer buying behavior is consistent at the category level, with Coffee leading performance, followed closely by Tea and Bakery.

This consistency allows for efficient operational planning across the week. While minor daily fluctuations exist, they do not require differentiated strategies by weekday. Deeper behavioral patterns may emerge through advanced analysis techniques such as basket analysis, time-series modeling, or customer segmentation.

By grounding retail decisions in objective statistical findings, businesses can improve accuracy, efficiency, and strategic focus.