IK Imran Khan
← All projects

Retail Promotion & Discount Effectiveness Study

Measuring how discounts, promotions and seasonal campaigns moved sales volume, revenue and profit margin, and finding the discount depth that lifts revenue without giving away the margin.

PromotionsMargin analysisIncrementalityExcelPython
+18%Incremental promo revenue
vs baseline
15%Optimal discount depth
best revenue/margin trade
−6.2 ptsMargin lost at 30%+ off
discounting too deeply
3.4×Volume lift on best campaign
seasonal peak

Baseline vs promotional sales · grouped bar

Daily units in normal weeks vs promotion weeks.

Revenue & margin by discount depth · dual axis combo

Revenue keeps rising, margin turns over, the crossover point is the sweet spot.

Discount depth vs incrementality · scatter / bubble

Each bubble is a campaign; x = discount, y = incremental lift, size = revenue.

Incremental revenue by campaign · diverging bar

Revenue above the modeled baseline, red campaigns barely broke even.

Price split: cost, margin & discount · 100% stacked bar

How each discount depth carves up the selling price.

Margin distribution by discount tier · box plot

Spread of product level margin in each tier, the box is the middle 50%, whiskers the range.

The question

A retailer ran frequent promotions but didn't know which ones actually paid off. Were discounts driving real incremental sales, or handing margin to people who would have bought anyway? And how deep can a discount go before it stops being worth it?

How I did it

  • Built a baseline of normal sales in Python from the weeks with no promotion, adjusted for season and day of week, so I had something honest to compare against.
  • Measured the real lift as promotion sales minus that baseline, which stops me crediting a discount for sales that would have happened anyway.
  • Compared revenue and margin across discount depths with Excel pivot tables, and used a 100% stacked view to show how the price splits into cost, margin and discount.
  • Dropped every campaign onto a bubble chart by discount depth, lift and revenue to see the winners and the duds at a glance.

What the analysis found

  • Discounts beyond ~15% kept lifting volume but started destroying margin faster than they added revenue.
  • Some recurring promotions had almost no incrementality, they mostly subsidised existing demand.
  • Seasonal campaigns timed to demand peaks delivered the strongest 3.4× volume lift at healthy margin.
Recommendation: Cap everyday discounts around 15%, retire the two weak recurring promos, and concentrate budget on seasonal peaks. Modeled effect: similar top line with materially better blended margin.

Tools

Excel · Python · Pivot tables · Visualization dashboard

▸Built on a simulated retail sales dataset so I can walk through the promotion and lift method openly.