Promotions and markdown strategy without margin erosion

# Promotions and markdown strategy without margin erosion

A buyer orders 10,000 linen blazers for spring. By late June, 4,000 are still on the floor and the fall set needs the space. Most teams reachreachThe number of unique people exposed to your message in a given period. Unlike impressions, reach counts each person once, no matter how often they see it.View full definition → for 50 percent off across the board. The disciplined ones already know, from a plan written in February, how deep each step goes and which sell-through number triggers it.

The distance between those two behaviours is worth several points of full-year gross margingross marginGross margin is the share of revenue left after subtracting the direct cost of producing goods or services, expressed as a percentage of revenue.View full definition →. Points of margin are what this lesson computes.

Why markdowns are a strategy, not a failure

A markdown is a permanent reduction in selling price, usually to clear stock. A promotion is a temporary discount to stimulate demand, after which the price goes back. Confusing them is where erosion begins: a promotion you never reverse is an unplanned markdown, and a markdown you dress up as a limited offer becomes a "was 120, now 60" claim that has to survive the substantiation rules the retail advertising lesson owns.

Markdowns are not evidence that buying went wrong. In apparel, perishability is built into the product. The goal is not zero markdowns. The goal is to harvest the highest willingness to pay from each layer of demand before the price drops.

The markdown cascade, layer by layer

Back to the blazers. A well-run end-of-season cascade looks like this.

Full price (weeks 1 to 8). You sell to the customers who want the item most. Highest-margin window, and the one most often given away for free.

First markdown, 20 to 30 percent. Sales have slowed. You reactivate shoppers who liked the blazer and hesitated. Usually the most profitable step: real volume, margin still intact.

Second markdown, 40 to 50 percent. Remaining stock now competes with incoming fall product for floor space and attention.

Final clearance, 60 percent and beyond. Whatever is left has to go. At this point the comparison is no longer full price versus discount, it is your net recovery versus what a liquidator would pay for the lot, minus the cost of the space it occupies until then.

Zara is the counter-example that makes the cascade look expensive. Small initial buys, replenishment in weeks rather than months, and sale periods confined to two short windows a year mean far less inventory ever reaches step three. Zara discounts well below the fast-fashion norm, and its customers have learned that hesitating means the item is gone, not cheaper. The cascade is a tool for the inventory you committed to too early.

What a discount has to earn back

Before arguing about depth, do the arithmetic. To hold gross profit flat, the unit lift you need is the discount divided by what is left of the margin after it:

required_lift = discount_pct / (gross_margin_pct - discount_pct)

On an item carrying 50 percent gross margingross marginGross margin is the share of revenue left after subtracting the direct cost of producing goods or services, expressed as a percentage of revenue.View full definition →:

• 20 percent off needs 67 percent more units just to stand still
• 30 percent off needs 150 percent more
• 40 percent off needs 400 percent more

On a 35 percent margin item, 20 percent off already needs a 133 percent lift. Very few promotions move volume by that much, which is why depth is the decision, not the afterthought. A 15 percent offer that works beats a 40 percent offer that "worked" on units.

One exception matters. Once stock is bought and sitting, the cost is sunk, and the choice is between money now and less money later. The cascade steps are judged on recovery against decay, not against original margin. Confusing those two frames is how buyers defend clearance depth with logic that should never have been applied to a full-price seller.

Price elasticityPrice elasticityHow sensitive demand is to a price change. High elasticity means customers react strongly to price increases.View full definition →: your core diagnostic

Elasticity measures how quantity responds to price. A 10 percent cut that lifts units 30 percent is elastic. The same cut lifting units 5 percent is inelastic, and on the arithmetic above that discount destroys profit.

Measure it per category, not store-wide. Denim behaves nothing like occasionwear. A blanket 30 percent off treats them as identical and subsidises the customers who were already at the till. For a non-technical primer, the Khan Academy lesson on price elasticity is free.

Forecasting: when to trigger the next step

Timing is driven by sell-through, not by the calendar or by nerves. Sell-through is the share of received inventory sold in a period: 6,000 of 10,000 blazers is 60 percent. Set the target curve before the season, then compare pace against remaining time.

weeks_of_supply = units_on_hand / avg_weekly_sales

if weeks_of_supply > season_weeks_remaining:
    trigger_markdown()   # you will not clear at current pace

Four thousand blazers left, 300 a week, 8 weeks before fall takes the floor: you clear 2,400 and miss by 1,600. That gap, not a hunch, sets both the trigger and the depth.

🎬 [VIDEO: "How Retailers Set Prices and Markdowns" - youtube.com - a short explainer on retail pricing and markdown logic for non-specialists]

Cannibalisation, subsidy and pull-forward

The number on the promo report is gross units, and most of them were coming anyway.

Take a 25 percent offer on a 45 percent margin item. Promo week sells 1,200 units. A matched set of non-promoted stores says 800 would have sold at full price. Baseline profit: 800 units at 45 of margin, 36,000. Promo profit: 1,200 units at 20 of margin, 24,000. You paid 12,000 for 400 incremental units, and breakeven needed 1,800.

The subsidy rate, baseline units divided by promoted units, is the fastest sanity check you have. Above roughly two thirds, you are mostly discounting demand you already owned.

Two further leaks. Cannibalisation sideways: the promoted SKU steals from its full-price neighbour in the same wardrobe slot, so category units barely move while category margin falls. And pull-forward: the two weeks after the event run below trend because buyers stocked up. Judge a promotion over six weeks, never over the week it ran.

You need clean identity to see any of this, which is why the exposed-versus-unexposed comparison depends on the persistent cross-channel profile described in the loyalty lesson. Tools such as Segment, a customer data platformcustomer data platformA Customer Data Platform unifies customer data from all sources into persistent, actionable profiles that other systems can use.View full definition → selling exactly this plumbing, exist because stitching web, app and till events into one shopper record is harder than the pitch decks suggest.

Demand-generating deals versus margin-destroying reflexes

A demand-generating deal brings volume you would not otherwise have captured, or clears stock heading for write-off. A margin-destroying reflex is a discount triggered by anxiety, habit or competitor mimicry. Signs you are in reflex territory:

• You match a competitor without checking whether your customers comparison-shop that item.
• You run the same holiday event annually because you always have.
• You discount inelastic product that is still selling, "to be safe".
• You promote full-price sellers, teaching customers to wait.

That last one compounds. J.C. Penney learned the reverse lesson in 2012: Ron Johnson removed the coupons and constant sales in favour of "fair and square" everyday pricing, and annual revenue fell roughly a quarter, from about 17 billion dollars to around 13 billion. Johnson was gone by April 2013 and the promotions came back. The cautionary reading is not that everyday pricing fails, it is that a customer base trained on coupons for a decade prices your goods off the discount, not off the ticket. Retraining them costs years, and you pay the bill up front.

A quick test before any discount

1. Incrementality. Sales that would not have happened, or a rebate on sales I already had?

2. Elasticity. Does the lift clear the required-lift number for this depth?

3. Signal. What does this teach customers about my future pricing?

Fail all three and it is a reflex.

The margin bridge: knowing which lever moved

"Gross marginGross marginGross margin is the share of revenue left after subtracting the direct cost of producing goods or services, expressed as a percentage of revenue.View full definition → came in 6 points light" is not a diagnosis. Decompose it. An illustrative season:

planned gross margin              54.0%
  markdown depth vs plan          -3.1
  markdown volume vs plan         -2.0
  category mix shift              -1.6
  supplier markdown funding       +1.0
  shrink and returns              -0.7
actual gross margin               47.6%

Depth and volume demand opposite fixes. Deeper than planned means your cascade steps were mistimed. More units marked down than planned means the buy was wrong, and no pricing discipline saves it. Mix means the promotions worked and pulled the wrong basket. Supplier markdown funding is the line most teams forget to negotiate before the season, when it is still negotiable.

Putting it together: the season plan

Pre-season. Target sell-through curves per category, planned cascade depth, the thresholds that fire each step, and the required-lift figure for every discount level you might approve.

In-season. Weekly sell-through against plan. Trigger on the pace gap, not on competitor noise.

Protect the full-price window. If an item is selling to plan, leave it alone. Early access for members, on the mechanics the loyalty lesson owns, moves committed buyers forward without a public price cut.

Differentiate by item. Deep clearance on true laggards, little or nothing on inelastic pieces still moving. One blanket percentage leaves money on the table twice.

Post-season. Build the bridge, attribute the misses to depth, volume or mix, and feed it back into next season's buy.

Key Takeaways

• A markdown is permanent price relief to clear stock; a promotion is temporary demand stimulation. Blur them and you train customers to wait.
• Run the required-lift number before approving any depth. At 50 percent margin, 30 percent off needs 150 percent more units to break even.
• Judge promotions on incremental units over six weeks, net of baseline and pull-forward, not on gross units in the promo week.
• Trigger markdowns on the gap between selling pace and remaining season time, not on the calendar or a rival's poster.
• Bridge the margin variance into depth, volume, mix and supplier funding. Each one points at a different team and a different fix.