Markdown Optimisation

Markdown optimisation is the discipline of deciding the timing and depth of permanent price reductions used to clear seasonal or slow-moving inventory while protecting as much margin as possible. Sources frame it as the end-of-lifecycle counterpart to Promotional Uplift Modelling: Toolio draws the line that "a price adjustment to manage in-season velocity is fundamentally different from a markdown to clear end-of-season excess — the first protects margin, the second recovers what's left of it" (Toolio, vendor, updated 2026-05-29). In fashion/apparel it is a structurally large lever because a high share of units now move at a discount, and each reduction erodes planned margin disproportionately. It sits inside the merchandise-planning stack alongside Merchandise Financial Planning (MFP) (which budgets the markdown reserve) and Open-to-Buy (OTB) (whose formula includes planned markdowns).

Sourcing note (2026-08-08 update): The 2026-06-26 run was web-only. This run adds registry (EU primary source, McKinsey PDF, IDC primary) and YouTube search (Apify unavailable — videos identified, no transcripts). Reddit remains unavailable in the Cowork cloud environment. The EU ESPR section has been upgraded from unverified to fully primary-sourced. McKinsey figures have been attributed to their correct 2019 and 2018 source documents with stale-risk flags added. The "unverified" callout on the 400–800 bps figure has been replaced by a contradiction callout. RELEX IDC MarketScape details updated with the correct 2025–2026 edition identifier.

Why it matters: the margin math

Eightx illustrates the asymmetry that makes markdowns expensive: a 15% markdown does not cost 15% of margin. On a $100 unit at 55% gross margin ($45 cost), a 15% cut drops gross profit from $55 to $40 — "more than a quarter of your gross profit dollars gone"; a 30% cut drops it to $25; and a ~55% clearance sells at cost for zero profit (Eightx 2026-06-08, labelled an illustrative model, not a measured stat). Toolio makes the same point on the velocity side: "a 70% [sell-through rate] with zero markdowns is worth more than a 90% STR after a 30% promotional discount" (Toolio, vendor, updated 2026-05-29).

The structural backdrop sources report (all vendor-attributed unless otherwise noted, treat as directional):

• Markdowns consume an estimated 20–50% of net sales at many fashion retailers (Markmi and Eightx, both citing McKinsey) (as-of 2026-06-26).

• Full-price sell-through has reportedly fallen from an old 70–75% norm to ~50% at many retailers — i.e. about half of all units now move at a discount (Markmi 2026; independently repeated by Eightx 2026-06-08) (as-of 2026-06-26). This is independently corroborated by McKinsey/BoF State of Fashion 2026 (see Sell-Through Rate for the primary-backed figure).

A worked example of the downstream P&L sensitivity is independently grounded: per its 10-K filed 2026-03-30, American Eagle's gross margin slipped to 36.5% in FY2025 from 39.2% the prior year, and operating income fell 47%, from $427.3M to $226.2M — a 2.7-point gross-margin slip nearly halving operating income (Eightx 2026-06-08, citing the SEC filing) (as-of 2026-06-26).

Markdown vs promotion (the definitional line)

Centric Software enumerates markdown types — clearance (at or below cost for slow sellers), seasonal, loss leaders, end-of-life, dynamic (real-time supply/demand timing), competitive, and promotional — and explicitly frames promotional cuts as "temporary price cuts, circular promotions, coupons, and other limited time price reductions," separate from inventory-clearing markdowns (Centric Software, vendor, 2024-06-25). This is the boundary that separates this page from Promotional Uplift Modelling: markdowns are permanent and inventory-driven; promotions are temporary and demand-driven.

Markdown cadence and depth (fashion)

Sources converge on a shallow, few-step, trigger-based approach rather than deep calendar-driven cuts:

• Eightx: best practice is a shallow first markdown of 10–25%, with only 2–3 steps across a season, differentiated by how price-sensitive the style is, on the principle "you can always discount deeper, you can never put the price back up" (attributed to Centric Software guidance plus BCG fashion-markdown research, Eightx 2026-06-08).
• Toolio's structured cadence: Phase 1 (weeks 1–3 post-peak) 10–15% on slow movers; Phase 2 (weeks 4–6) 20–30% if Phase 1 didn't hit velocity; Phase 3 (end-of-season) deeper clearance with a floor at landed cost plus minimum acceptable margin (Toolio, vendor, updated 2026-05-29).
• Eightx: the first markdown should be triggered by a sell-through threshold, not a calendar date — at the midpoint of a style's selling window, styles materially below their sell-through curve get a small automatic cut (Eightx 2026-06-08).
• Centric: markdown ladders (the discount percentages shoppers see) typically begin at 10% and deepen; cuts under ~5% "do not move consumers"; and blanket category-wide markdowns waste margin on healthy SKUs versus precise SKU/colour/size-level selection (Centric Software, vendor, 2024-06-25).

Sell-through benchmarks driving the trigger

Toolio reports healthy apparel/fashion Sell-Through Rate at 65–85% (fast fashion 85%+, basics 65–70%), with end-of-season STR below 60% triggering markdown escalation for most buying teams (as-of 2026-06-26). By vertical it cites Health & Beauty 75–90%, Sporting Goods 70–85%, General Retail 70–80%, Consumer Electronics 60–75%, Home & Furniture 55–75%, Luxury & Jewellery 50–65% (low STR partly intentional for scarcity) (Toolio, vendor, updated 2026-05-29).

AI / automated vs manual / rules-based

Sources describe a shift from calendar/rules-based markdowns to condition-based, SKU-level optimisation:

• Trail-ML: dynamic pricing models "combine demand forecasting, Price Elasticity estimation, and constrained optimization to find revenue or margin-maximizing prices subject to business rules," replacing both reactive run-to-failure pricing and fixed calendar schedules (trail-ml, vendor/use-case page, date unknown). Elasticity per product-segment is estimated via causal-inference methods — difference-in-differences, instrumental variables, or randomised price experiments.
• Beebi: unlike one-size-fits-all rules, products in the same category respond differently — "one SKU may benefit from an earlier, modest intervention. Another may need a deeper price reduction later... A third should remain untouched" (Beebi, snippet only, date unknown).
• Centric: AI markdown optimisation uses "optimization engines" over sales history, real-time tracking, customer behaviour, market trends and competitor analysis to deliver SKU-level suggestions per channel/location, with adjustable automation and human override (Centric Software, vendor, 2024-06-25).
• Toolio: moving from spreadsheet-based to purpose-built planning "typically" yields a 5–15% improvement in end-of-season STR (Toolio, vendor) (as-of 2026-06-26).
• NRF 2026 (official event recap, January 2026): described the retail industry as entering an "agentic" phase of AI deployment in which systems act autonomously to "optimise inventory, enhance customer journeys and power personalised engagement," with NRF 2026 sessions explicitly covering AI-powered autonomous pricing that reacts to demand and competitor pricing in real time (NRF Big Show 2026 recap; Impact Analytics NRF 2026 summary, vendor blog) (as-of 2026-01).

[!unverified] Aggregator/snippet figures not fetched at source: AI improving markdown performance "by an average of 26%," a fashion retailer "saving $32 million annually through reduced terminal markdowns," and AI cutting markdown depth 5–10% while lifting inventory turnover 10–25% (NetSolutions snippet, date unknown). Single-source, vendor/aggregator, unverified — recorded as a claim, not as fact.

Vendor / tooling landscape

Gartner frames the competitive category as "Unified Price, Promotion, and Markdown Optimization Applications" (Gartner reviews, date unknown) — a useful umbrella for where these tools compete.

• McKinsey uses a proprietary tool, Periscope Price Advisor, to guide apparel companies through forecasting, markdown, pricing and promotion across the product lifecycle (McKinsey, search snippet, date unknown). McKinsey Periscope was named a Leader in the IDC MarketScape: Worldwide Retail Price Optimization Solutions 2023 Vendor Assessment (McKinsey.com press release, 2023) (as-of 2023).

• RELEX Solutions was named a Leader in the IDC MarketScape: Worldwide Retail Price Optimization Solutions 2025–2026 Vendor Assessment (container ID: US52989825, 21 vendors assessed), as announced on 10 December 2025. IDC's VP Ananda Chakravarty stated: "RELEX brings together strong automation and data-driven insights that help retailers make smarter, faster pricing and promotions decisions." RELEX's roadmap is focused on building a unified dashboard integrating base pricing, promotions management, and markdown optimisation (RELEX press release via BusinessWire, 2025-12-10) (as-of 2025-12-10).

• Nextail — AI-native merchandising platform for fashion/apparel (inventory optimisation, allocation, replenishment, SKU-store demand prediction), reporting up to 40% sales growth for brands such as Guess and Gina Tricot (CBInsights company profile / vendor claim, date unknown) — low confidence.
• Centric Software, Toolio, Markmi — fashion merchandise-planning suites with markdown modules (all vendor sources used above).

Margin improvement potential — what McKinsey actually says

[!unverified] Vendor implementation case figures of "4–6% margin improvement" / "~200 bps" / "up to 4% gross-margin uplift" from various WebSearch snippets. Not strictly contradictory with McKinsey ranges (ceiling-of-discipline vs specific deployments) but the smaller figures come from vendor case studies with conflict of interest.

Regulatory / structural context (2026)

EU ESPR — destruction ban on unsold textiles (primary-sourced)

The prohibition on destruction of unsold apparel, clothing accessories, and footwear was adopted as a Delegated Regulation and Implementing Regulation on 9 February 2026 under the Ecodesign for Sustainable Products Regulation (ESPR), which itself entered into force on 19 July 2024 (European Commission Environment, 2026-02-09).

In-force dates by company size (European Commission Environment, 2026-07-17):

Tier	Criteria	In-force date
Large companies	Exceeds at least 2 of 3: >250 employees, >€50M turnover, >€25M total assets	19 July 2026
Medium companies	50–249 employees, turnover ≤€50M	1 January 2030
Small and micro enterprises	<50 employees, balance sheet <€10M	Permanently exempt

Product scope: apparel, clothing accessories (hats, headgear), and footwear of all materials (leather, rubber, plastics, textile). (as-of 2026-07-17)

Permitted exceptions — destruction is allowed only when items are (European Commission Environment, 2026-07-17):

• Unsafe or severely damaged
• Non-compliant with EU or national law
• Counterfeit or IP-infringing
• Rejected by charities or donation schemes

All exceptions must be documented. Companies must retain records for five years to allow national authority inspections.

Disclosure: a standardised format for reporting volumes of unsold goods discarded applies from February 2027 (European Commission Environment, 2026-02-09).

Enforcement: national market-surveillance authorities are responsible; fines must be "effective, proportionate, and dissuasive" — each member state sets its own fine schedule. No EU-harmonised fine quantum was published as of 2026-08-08 (Anthesis Group consultancy commentary, 2026, secondary source).

Scale context: the Commission estimates 250,000–600,000 tonnes of unsold clothing are destroyed annually in Europe, representing 4–9% of total garments placed on the EU market (as-of 2026-07-17) (European Commission Environment, 2026-07-17).

Implication for markdown strategy: with destruction no longer available as a terminal inventory clearance route for large EU fashion companies, the calculus on how low to go on terminal markdowns shifts — units that previously might have been destroyed must now be sold (however cheaply), donated, or repurposed. This structurally changes the floor on clearance markdown depth.

• Markmi also reports that supply-chain lead times have "roughly doubled to ~67 days" (Markmi, vendor, 2026) (as-of 2026-06-26).

[!unverified] The "~67 day" lead time doubling figure is vendor-stated (Markmi) and not independently corroborated here.

• Eightx reports US online apparel return rate at ~24.4% (attributing Coresight Research, "The True Cost of Apparel Returns"), with discount-driven buyers returning at higher rates than full-price buyers — implying markdowns should be modelled on net units after returns (Eightx 2026-06-08) (as-of 2026-06-26). See Returns Management.

Key terms

Term	Meaning (as sources use it)
Markdown	A permanent price reduction to clear inventory (vs a temporary promotion)
Markdown ladder	The sequence of discount percentages a style steps through over a season (Centric)
Terminal / clearance markdown	The final, deepest cut to exit remaining stock, floored at landed cost + min margin (Toolio)
Sell-through trigger	Marking down based on a style falling below its sell-through curve rather than a calendar date (Eightx)
Markdown reserve / budget	The planned markdown allowance carried in Merchandise Financial Planning (MFP) / Open-to-Buy (OTB)

What practitioners report

Gap (2026-08-08): Reddit MCP unavailable in Cowork cloud (known recurring gap). Apify YouTube transcripts unavailable (known recurring gap). Calendar-vs-trigger timing debate, "one deep cut vs incremental ladder," and whether small/mid retailers find optimisation tools worth the cost remain unfilled. Five YouTube videos were identified from PI Apparel Merchandise Planning USA 2025 and NRF 2026 — these are high-signal conference content candidates for the next run that can access Apify. See YouTube — Markdown Optimisation 2026-08-08 for the video list.