Retail math by segment: IMU, MMU, markdowns, sell-through, GMROI, turnover
How initial markup, maintained margin, markdown share, sell-through, turnover and GMROI relate across segments and how to read other people’s figures. These are formulas and orders of magnitude, not norms; they decide whether a margin is achievable in the chosen segment and what share of markdowns can be afforded.
Why this document
Retail metrics are linked: the more markdowns, the lower the maintained margin at the same initial markup, and the faster the turnover, the lower the markup needed for the same GMROI. The finance director and the commercial director read this document when they set margin targets and compare themselves with the market. It is needed so that a figure from another segment is not taken as a norm. That means targets are set by formulas with your own data, not by the industry average.
At a glance
- What is this document and why is it needed?
- How initial markup, maintained margin, markdown share, sell-through, turnover and GMROI relate across segments and how to read other people’s figures. These are formulas and orders of magnitude, not norms; they decide whether a margin is achievable in the chosen segment and what share of markdowns can be afforded.
- Who owns the document and how often is it rebuilt?
- Kept by a buying analyst or finance controller; once per season; passed to category managers and buying heads.
- What mistakes are the most common?
- Comparing a company IMU with another’s reported gross margin → false conclusions → reduce to MMU and one base. Computing turnover from two dates → distortion under seasonality → use the monthly average. Mixing markdown share in units and money → a wrong MMU → state the base.
For whom
- Brand
- A brand reads retail math to understand what the retailer expects: initial markup and maintained margin, markdown protection, sell-through; wholesale terms, bonuses and exchanges arise from it.
- Retail
- A retailer uses retail math for buying plans (OTB), brand and category evaluation, markdown decisions: the goal is to protect MMU and GMROI.
What it consists of
Definitions
- Retailer: IMU = (initial retail − cost) / initial retail; keystone = IMU 50% (price twice the cost).
- Retailer: MMU = (net sales − COGS) / net sales; MMU = 1 − (1 − IMU) / (1 − d) with markdown share d in money.
- Sell-through = units sold / (units sold + units on hand) for the period (that is, sold ÷ everything that was on sale; with no opening stock it equals sold ÷ received); full-price sell-through is counted separately.
- Turnover T = COGS / average stock at cost; days of inventory = 365 / T.
- Retailer: GMROI = gross margin in money / average stock at cost = T × MMU / (1 − MMU).
- Markdown share d: markdown amount / revenue before markdowns; different bases give different numbers.
How to read public numbers
- Reported gross margin includes markdowns, returns and sometimes logistics, so it is closer to MMU than to IMU: another company reported margin cannot be compared with a list initial markup, the base, period and cost composition differ.
- Segments differ in the mix of (IMU, d, T): luxury is high IMU, low d and low T; mass is a medium IMU, large d and high T; off-price is low purchase cost and high T.
- There are no universal “norms” for sell-through and markdown share: treat numbers of others as hypotheses and test them on your own ledger.
Brand and retailer: what is done differently
- Brand: wholesale price from the retailer target IMU: W = P0 x (1 - IMU_r), where P0 is the retail price excl. VAT; the brand wholesale margin GM_b = 1 - C / W. Example: P0 = RUB 8,000 and a retailer IMU of 55% give W = RUB 3,600, and at C = RUB 1,800 the brand margin is 50%.
- Brand: support for retailer markdowns (compensation of part of the markdowns or a discount on the next order) raises retailer MMU and lowers brand margin: count compensation as a retroactive reduction of the wholesale price W.
- Brand: in own retail IMU = 1 - C / P0 and MMU follows the own-retail d; a unit brings more margin than in wholesale but needs rent and staff and holds stock: compare GMROI and cash frozen in stock, not only the margin percentage.
- Retailer: buying from a target MMU: IMU_needed = 1 - (1 - MMU_target) x (1 - d). Example: a target MMU of 50% at d of 20% needs IMU = 1 - 0.5 x 0.8 = 60%.
- Retailer: target GMROI = T x MMU / (1 - MMU): at slow category turnover compensate with a high MMU, at fast turnover a low MMU is acceptable (off-price logic). Compare categories by GMROI, not by margin.
- Retailer: reconcile management and accounting gross margin: the inventory impairment reserve under FSBU 5/2019 changes reported margin, so management MMU and accounting gross margin diverge; keep a bridge between them.
How to read the result and which decisions follow
- Read the IMU and MMU bars as the price of markdowns: the kids gap is 8.6 pp at d 16%, accessories 2.4 pp at d 6%. The portfolio GMROI is 2.74; footwear (1.83) pulls it down, men (3.42) up.
- Brand: the sensitivity table shows turnover moves GMROI more than markdowns (at +20% to T the portfolio gives 3.29, at -20% to IMU 1.61). Decision: it pays a brand more to cut the batch and speed up turnover than to build more initial markup that the retailer will cut with markdowns anyway.
- Retailer: footwear holds 22.2% of stock against 16.4% of revenue, accessories 10.1% against 8.2%. Decision: in categories where the stock share exceeds the revenue share, cut depth and frequency of deliveries; a 10% cut in d gives less GMROI than a 10% faster T.
- Brand: if the retailer asks for an IMU 5 pp higher (60%), the wholesale price falls from 3600 to RUB 3,200 and the brand wholesale margin from 50 to 43.8%. Before agreeing, recompute whether the order volume covers this difference.
- Retailer: set the buying goal through MMU and GMROI at once: a high IMU at low T gives a low GMROI, while a low IMU at high T can give the same result (off-price logic).
Formula derivation and sensitivity tables
- Derivation of MMU. Let P_0 be the price before markdowns, C cost, IMU = 1 - C / P_0, hence C = P_0 x (1 - IMU). Markdowns of share d of turnover give revenue P_0 x (1 - d). Then MMU = (P_0 x (1 - d) - C) / (P_0 x (1 - d)) = 1 - (1 - IMU) / (1 - d).
- Derivation of GMROI. Gross margin GM = S x MMU, cost of sales COGS = S x (1 - MMU), average stock at cost Z = COGS / T. Then GMROI = GM / Z = S x MMU x T / (S x (1 - MMU)) = T x MMU / (1 - MMU).
- The inverse formula: from MMU = 1 - (1 - IMU) / (1 - d) it follows that IMU = 1 - (1 - MMU) x (1 - d); at a target MMU of 50% and d of 20% IMU = 1 - 0.5 x 0.8 = 60% is needed.
- Sensitivity of the total portfolio GMROI (parameter change in relative percent, code values): IMU: -20% -> 1.61; -10% -> 2.11; 0% -> 2.74; +10% -> 3.57; +20% -> 4.68. markdown d: -20% -> 2.9; -10% -> 2.82; 0% -> 2.74; +10% -> 2.66; +20% -> 2.58. turnover T: -20% -> 2.2; -10% -> 2.47; 0% -> 2.74; +10% -> 3.02; +20% -> 3.29.
- Brand: moving from the retailer IMU to your wholesale margin goes through W = P_0 x (1 - IMU_r) and GM_b = 1 - C / W; at P_0 RUB 8000, IMU 55% and C RUB 1,800 the brand margin is 50%, at IMU 60% 43.8%.
- Brand: compensation of retailer markdowns reduces W retroactively: a compensation share x of markdowns gives W_eff = W - x x d x P_0 x (1 - IMU_r); count it as part of the wholesale price, not a marketing expense.
- Brand: fast turnover replaces high markup: for women at twice the T the same GMROI of 3.14 needs MMU of 34.4% instead of 51.2%; this is how off-price logic works.
Control checks
- Total rows are sums of roubles: MMU 50.4% is the sum of margin over the sum of revenue, GMROI 2.74 the sum of margin over the sum of stock (the code matches the example table).
- Brand: IMU computed on the wholesale price and the retailer IMU on the retail price are not equal; before comparing bring both sides to one base (price excl. VAT).
- Retailer: take the markdown share in money from turnover before markdowns, not in pieces; the share of pieces on discount and the share of money differ.
- Brand: reconcile management and statutory gross margin: the inventory impairment reserve under FSBU 5/2019 pulls them apart, and this bridge must be recorded.
- Retailer: compute T from the monthly average stock; from two dates seasonality overstates or understates turnover in footwear and accessories.
Where the data comes from
| Data source | Fields |
|---|---|
| Sales and stock ledger | revenue before and after markdowns, cost, stock at period start and end |
| Management accounts | gross margin, inventories, cost of sales (for turnover) |
| Calculation method | definitions and calculation order |
How to build it step by step
- Step 1. Fix the bases: price or cost, units or money, period.
- Retailer: Step 2. Compute IMU from the segment buying plan.
- Retailer: Step 3. Set the planned markdown share d and compute MMU.
- Step 4. Compute T from the monthly average stock, not two dates.
- Step 5. Compute GMROI and compare segments.
- Step 6. Reconcile with the public gross margin and explain the gap (logistics, returns, rent).
- Step 7. Set limits: minimum MMU and GMROI, maximum d.
Formulas
| Measure | Formulas | Note |
|---|---|---|
| MMU from IMU and markdowns | MMU = 1 − (1 − IMU) / (1 − d) | If IMU is 50% and d is 20%, MMU = 37.5%. |
| GMROI | GMROI = T × MMU / (1 − MMU) | T at cost; turnover at retail prices needs another formula. |
| Days of inventory | DIO = 365 / T | Compare with the seasonal cycle. |
| Brand: Wholesale price from retailer IMU | W = P0 x (1 - IMU_r) | P0 is the retail price excl. VAT; the brand wholesale margin is GM_b = 1 - C / W. |
| Retailer: Needed IMU from target MMU | IMU = 1 - (1 - MMU_target) x (1 - d) | The inverse of MMU = 1 - (1 - IMU) / (1 - d); d in money. |
Formulas in notation
- Wholesale price from retailer IMU
W = P_0 × (1 − IMU_r)
P_0 is the retail price excl. VAT; the brand wholesale margin is GM_b = 1 - C / W.- IMU needed for a target MMU
IMU = 1 − (1 − MMU_t) × (1 − d)
The inverse formula: how much to build in at the start to reach the target after markdowns.
Worked example
Category cut: IMU, markdowns, turnover and GMROI, net revenue RUB 146 m (exercise assumptions)
| Category | Revenue, RUB m | IMU, % | Markdown d, % | MMU, % | T | DIO, days | Average stock at cost, RUB m | GMROI |
|---|---|---|---|---|---|---|---|---|
| Women | 62 | 58 | 14 | 51.2 | 3 | 122 | 10.1 | 3.14 |
| Men | 28 | 57 | 11 | 51.7 | 3.2 | 114 | 4.2 | 3.42 |
| Kids | 20 | 55 | 16 | 46.4 | 2.8 | 130 | 3.8 | 2.43 |
| Footwear | 24 | 52 | 12 | 45.5 | 2.2 | 166 | 6 | 1.83 |
| Accessories | 12 | 62 | 6 | 59.6 | 1.8 | 203 | 2.7 | 2.65 |
| Total | 146 | 50.4 | 2.7 | 135 | 26.8 | 2.74 |
Revenue of RUB 146 m is the sum of five categories; total MMU 50.4% is the sum of gross margin in roubles over the sum of revenue, not an average of percentages, total T 2.7 is cost of sales over the sum of average stock, total GMROI 2.74 is margin in roubles over stock at cost. Footwear and accessories hold more stock per rouble of sales, so their weight in total T is heavier than their revenue share. All inputs are exercise assumptions, not market norms.
Figures
- IMU
- MMU
Data as a table
| Measure | IMU | MMU |
|---|---|---|
| Women | 58 | 51.2 |
| Men | 57 | 51.7 |
| Kids | 55 | 46.4 |
| Footwear | 52 | 45.5 |
| Accessories | 62 | 59.6 |
IMU and MMU by category. The gap between IMU and MMU is the price of markdowns: widest for kids because of d = 16%.
- Revenue share
- Share of average stock
Data as a table
| Measure | Revenue share | Share of average stock |
|---|---|---|
| Women | 42.5 | 37.7 |
| Men | 19.2 | 15.8 |
| Kids | 13.7 | 14.3 |
| Footwear | 16.4 | 22.2 |
| Accessories | 8.2 | 10.1 |
Revenue share and stock share by category. A category is heavier in stock than in revenue if its stock share is higher: footwear and accessories hold stock beyond their revenue.
Data as a table
| Measure | GMROI |
|---|---|
| Women | 3.14 |
| Men | 3.42 |
| Kids | 2.43 |
| Footwear | 1.83 |
| Accessories | 2.65 |
GMROI by category. Gross margin per rouble of stock: footwear is lowest because of T = 2.2, men highest because of T = 3.2.
- 1.61 … 4.68 x
Data as a table
| -20% | -10% | 0% | +10% | +20% | |
|---|---|---|---|---|---|
| IMU | 1.61 | 2.11 | 2.74 | 3.57 | 4.68 |
| Markdown d | 2.9 | 2.82 | 2.74 | 2.66 | 2.58 |
| Turnover T | 2.2 | 2.47 | 2.74 | 3.02 | 3.29 |
Portfolio GMROI sensitivity. Total GMROI when a parameter changes by ±10% and ±20% in relative terms: turnover moves the result more than markdowns.
- 1.72 … 3.59 x
Data as a table
| -20% | -10% | 0% | +10% | +20% | |
|---|---|---|---|---|---|
| Women | 3.34 | 3.24 | 3.14 | 3.04 | 2.94 |
| Men | 3.59 | 3.51 | 3.42 | 3.34 | 3.26 |
| Kids | 2.63 | 2.53 | 2.43 | 2.33 | 2.23 |
| Footwear | 1.94 | 1.89 | 1.83 | 1.78 | 1.72 |
| Accessories | 2.71 | 2.68 | 2.65 | 2.62 | 2.6 |
Category GMROI as markdown d shifts. Each category when the markdown share changes by ±10% and ±20% in relative terms: shows where markdowns are more dangerous.
Specifics by category
- Apparel
- Apparel as the base for the formulas. Take IMU of 55-62% and d from 6 to 16%. Example: IMU 58% at d 14% gives MMU = 1 - 0.42 / 0.86 = 51.2%. Read segments as a triple (IMU, d, T): luxury is high IMU and low d, mass is medium IMU and high d, off-price is low buying cost and high T. Do not compare initial markup with another company reported gross margin: these are different bases.
- Footwear
- Footwear: turnover is below apparel (T = 2.2, DIO 166 days in the example) because the size run holds stock in edge sizes; IMU is lower because of the high cost. GMROI at low d is still comparable: IMU 52% and d 12% give MMU 45.5% and GMROI = 2.2 x 0.455 / 0.545 = 1.83. Compute T on the monthly average stock, not on two dates: seasonal AW footwear deliveries distort the result strongly.
- Accessories
- Accessories: IMU is high (62% in the example), d low (6%), T modest (1.8), so GMROI rests on margin, not turnover: MMU = 1 - 0.38 / 0.94 = 59.6%, GMROI = 1.8 x 0.596 / 0.404 = 2.65. Gift seasonality makes annual-average T imprecise, so compute by month and by subcategory (bags, scarves, costume jewelry): a category average hides slow items.
- Kids
- Kids: IMU 55% and d 16% give MMU 46.4% at T = 2.8. School and New Year peaks inflate average stock, so T is computed by month. Take price excl. VAT: the 10% rate under the decree no. 908 list or 22%, and at the same shelf price IMU differs between items; a rate error shifts all indicators. Compute sell-through to the school date and to New Year separately from the annual one, otherwise size tails get lost in the average.
- Men
- Men: IMU 57%, d 11%, MMU = 1 - 0.43 / 0.89 = 51.7%, T = 3.2. Steady demand and a high basics share give a higher T and a lower d, so GMROI is above women at a close initial markup. This depends on your assortment structure: compare men and women d over 2-3 seasons of your own data rather than rely on general claims about conservative demand.
- Women
- Women: IMU 58%, d 14%, MMU 51.2%, T = 3.0. Fashion widens the d spread between styles: core at d 5% and fashion at d 30% give one average that hides both. Compute MMU and GMROI separately for core and fashion. Take markdown volume in money from pre-markdown turnover, not in units: the share of units on discount and the share of money differ, and the base error is the most common here.
Common mistakes
- Comparing a company IMU with another’s reported gross margin → false conclusions → reduce to MMU and one base.
- Computing turnover from two dates → distortion under seasonality → use the monthly average.
- Mixing markdown share in units and money → a wrong MMU → state the base.
- Using “norms” from other publications without testing → decisions on unverified numbers → mark “unverified” and use your own ledger.
Russian specifics
- There are no unified Russian norms for IMU, MMU, sell-through and markdown share; consultancies and associations publish guides whose freshness and method differ: unverified.
- For Russian data take price excl. VAT (22% from 2026; 10% on part of kids goods), otherwise IMU and MMU are distorted. Rule: Tax Code Art. 164: Tax Code, Art. 164 (VAT rates).
- The inventory write-down reserve under FSBU 5/2019 affects reported margin: reconcile management and statutory gross margin.