Back to Project Library
Completed
Public-safe proof

Multivariable regression

TJIX Net Sales Drivers

Linked advertising, e-commerce growth, and market trends to a business-readable net-sales growth story for TJX.

RegressionCorrelation AnalysisExcelDecision TreesScenario Modeling

Exploratory Advertising Coefficient

12.737

Exploratory association, not statistically proven lift: ~$12.7M in estimated net-sales association per +$1M in advertising spend. Based on 10 annual observations; p = .103 and a 95% coefficient interval of −3.47 to 28.94, including zero.

TJIX Net Sales Drivers

Multivariable regression · commercial analytics

What this proves

  • Problem: needed to explain which levers most credibly drive TJX net sales growth and where e-commerce and advertising investment mattered.
  • Method: used regression, correlation analysis, scenario modeling, Excel, and competitor comparison to tie advertising and e-commerce growth to net-sales outcomes.
  • Outcome: final materials attribute roughly $12.7M in estimated net-sales association per additional $1M in advertising spend.

Rechecked model and interpretation

This academic study used regression and benchmarking to explore sales associations. It did not demonstrate a causal advertising return or realized revenue gain.

The workbook's advertising regression has 10 annual observations and controls for year and US retail sales. Advertising's coefficient is 12.737, with both sales and advertising measured in millions of dollars; p = .103, and the 95% coefficient interval is −3.47 to 28.94, including zero. Thus the estimate is not statistically significant at the 10% level. The worksheet also labels a narrower interval as 89%, not 90%.

The original report is retained as a historical academic artifact. Its causal language and 90%-confidence interpretation are superseded by this note. Some historical e-commerce inputs were estimated, adding uncertainty. A stronger follow-up would audit those inputs, assess collinearity/time trends, and validate on additional out-of-sample observations.

Live proof bundle

This project has the strongest commercial proof bundle in the current public portfolio set:

  • the report explains the analytical narrative and recommendation framing
  • the workbook shows the supporting model structure and scenario logic
Open final report Open workbook