Introducing Endless AI and Endless MCP. Operate from anywhere.

It is week 12 of a 16-week season, and the planning meeting has one item: a holiday gift set that sold well for four weeks and then stopped. There are 2,400 units left. Someone proposes 40% off. Someone else says wait until after the next email campaign. The meeting ends with a compromise of 25% off starting next week.

Markdown optimization is the practice of choosing when to mark down, how deep, and for which products, so that stock clears by a target date at the lowest total margin cost. The example below is a composite built from typical patterns, with round numbers so the arithmetic is easy to follow. It walks the season week by week, then replays it with three rules that would have cleared the same units for more margin.

The setup

ItemValue
Units bought6,000
Unit cost$14
DTC full price$40
Wholesale price$20
Season length16 weeks
ChannelsDTC site, two wholesale accounts
TargetClear by week 16; anything left goes to an off-price buyer at $10 a unit

The season as it happened

Weeks 1–4: strong start

The set sold 450 units a week on the site, and the wholesale accounts placed a 600-unit reorder. By the end of week 4, 2,400 units were gone: 1,800 DTC and 600 wholesale. Sell-through was 40% after a quarter of the season, ahead of plan.

Weeks 5–8: the stall

Weekly DTC sales dropped to 150, then 100. Nothing changed in price or marketing. The likely causes were ordinary ones: the gifting window was closing, and a competitor launched a similar set. By week 8, total sold was 2,900, and weeks of cover at the current rate had gone from about 8 to more than 30.

Nobody acted, because the season-level sell-through number still looked healthy. 48% sold at the halfway point was close to plan. The weekly rate, which had collapsed, was not on the report anyone read.

Weeks 9–12: waiting

The team held full price for a planned email feature in week 10. It produced one good week of 400 units, then sales went back to 100 a week. By week 12, 3,600 units had sold and 2,400 remained, which is where this piece started.

Weeks 13–16: the markdown

25% off ($30) in week 13 lifted sales to 250 a week for two weeks. 40% off ($24) in week 15 lifted them to 450 a week. That cleared 1,400 units. The last 1,000 went to the off-price buyer.

What the markdowns cost

DTC at full price:       3,000 x $40 = $120,000
Wholesale reorder:         600 x $20 =  $12,000
DTC at 25% off:            500 x $30 =  $15,000
DTC at 40% off:            900 x $24 =  $21,600
Off-price:               1,000 x $10 =  $10,000
Total revenue                          $178,600
Total cost               6,000 x $14 =  $84,000
Gross margin                           $94,600   (53.0%)

If every DTC unit had sold at $40 and the wholesale units at $20, revenue would have been $228,000 and gross margin $144,000. The gap is $49,400. Some of it was unavoidable once demand stalled. The question is how much.

Three decisions that cost the most

1. The trigger was sell-through to date instead of the weekly rate

Season sell-through blends the strong start with the stall. At the end of week 6, 2,700 units had sold. At 150 a week for the remaining 10 weeks, the set would finish with about 1,800 units unsold at full price, and the rate was still falling. A weekly projection of remaining units would have flagged the problem six weeks before the meeting that finally acted.

2. The first markdown came late, so it had to be deep

A markdown needs time to work. Starting at week 13 left four weeks to clear 2,400 units, which forced a second, deeper cut and an off-price sale for the rest. An earlier markdown can be shallower, because it has more weeks to sell through.

3. Excess went to off-price before anyone asked the other channels

The two wholesale accounts had reordered in week 3 and were still in their own selling window through about week 10. Nobody offered them more stock. Every unit they would have taken at the regular $20 wholesale price was a unit that instead went to the off-price buyer at $10.

The same season, replayed with rules

Same demand, same stall. Three rules applied:

  1. Trigger a review when projected end-of-season stock at the current weekly rate exceeds 15% of the original buy.
  2. Take a first markdown of 20%, with a second step only if the projection still misses after the first step has had time to work.
  3. Before planning any off-price sale, offer excess to channels still selling, at their regular price.
WeekActionUnits (illustrative)
6Projection shows about 1,800 units left at season end. Review triggered.—
7Last full-price week. Wholesale accounts offered more stock at $20 and take 500.500 wholesale at $20
8–14DTC marked down 20% ($32). Weekly sales rise to about 250, including the email week.1,750 at $32
15–16Final 30% step ($28) on what remains.600 at $28
16Remainder to off-price350 at $10
DTC at full price (weeks 1–7):   2,200 x $40 = $88,000
Wholesale reorder (week 3):        600 x $20 = $12,000
Wholesale top-up (week 7):         500 x $20 = $10,000
DTC at 20% off:                  1,750 x $32 = $56,000
DTC at 30% off:                    600 x $28 = $16,800
Off-price:                         350 x $10 =  $3,500
Total revenue                                 $186,300
Total cost                       6,000 x $14 = $84,000
Gross margin                                  $102,300   (54.9%)

The replay clears the same 6,000 units and produces $7,700 more gross margin, even though it starts discounting five weeks earlier. The unit figures are illustrative. In general, an early, shallow markdown, with excess offered to channels that are still selling, beats a late, deep one.

The rules, generalized

  1. Watch the weekly rate. Project units remaining at season end from the current rate, every week, for every seasonal SKU.
  2. Set a trigger in advance. Decide before the season what projected leftover share starts a review.
  3. Start shallow and early. A first step of 15–25% with time to work usually recovers more margin than a 40% cut late.
  4. Ask the other channels before you liquidate. A wholesale account or retail partner still in its selling window may take stock at its regular price.
  5. Set a floor. Know the price below which off-price or donation recovers more than another markdown step.
  6. Review the buy, not only the markdown. A season that ends in heavy markdowns is often an assortment or depth problem.

What the rules need in order to run

Each rule depends on current data: weekly sales by channel, on-hand stock, and open purchase orders that might still land. When those live in separate systems, the projection gets rebuilt from exports once a month, which is how a stall in week 5 gets discussed in week 12.

Endless keeps orders from DTC, wholesale, and retail channels, inventory, and purchase orders on one record, with forecasting on top of that data. A weekly projection of units remaining reads from live numbers instead of a reconciled spreadsheet.

The break-even question for every markdown step

Each markdown step trades margin per unit for speed. Whether it pays comes down to a comparison you can make before you cut the price.

Margin from marking down now:
  Units expected to sell at the new price x (New price - Unit cost)

Margin from holding:
  Units expected to sell at full price before the deadline x (Full price - Unit cost)
+ Units left at the deadline x (Liquidation price - Unit cost)

Take the gift set at week 8, with 3,100 units left and eight weeks to go. Holding at $40 and 100 units a week sells 800 units for $20,800 of margin, and leaves 2,300 units to liquidate at $10, a loss of $9,200 against cost. Net: $11,600.

Marking down 20% to $32, at the 250 a week the replay assumed, sells about 1,750 units over seven weeks for $31,500 of margin. The remaining 1,350 units, sold at a final step or to off-price, add or subtract a little. Even before those, the markdown is ahead by about $20,000.

The comparison turns on one estimate: how much the markdown lifts the weekly rate. That lift is the number to track from past markdowns, by category and by depth. A brand that knows a 20% markdown on seasonal gifting lifts sales about 2.5 times can make this decision in a minute. A brand without that history is guessing, and most guesses err toward waiting.

Questions buyers ask

What is markdown optimization?

Markdown optimization is choosing when to discount, how deep, and for which products, so excess stock clears by a target date at the lowest total cost to margin. It weighs the margin given up on each markdown against the cost of holding or liquidating the stock that does not sell.

When should you take a first markdown?

When the projection of units left at season end, at the current weekly sales rate, exceeds the leftover share you set before the season. Waiting for season-to-date sell-through to look bad usually means waiting too long.

Are deeper markdowns better for clearing stock?

Deeper markdowns clear stock faster, but an earlier, shallower markdown with more weeks to work often clears the same units at a smaller margin cost.

Commerce is chaos.

Tame your tech stack with one system that brings it all together—and actually works.

Get a Demo

Share this post