Syntha Lab RU

Plan versus actual and profit factor analysis (bridge)

An analysis of why actual profit differs from plan: how much came from volume, price, cost of sales and expenses. It drives the decision on what to fix first: the demand forecast, price, purchasing or expenses, and whom to ask about the deviation.

Finance Brand and retail Verified: 2026-11-13

Open interactively →

Why this document

The statement that profit is 5 million below plan does not say what to do. A factor analysis splits the deviation into causes: fewer units sold, deeper price reductions, dearer purchasing or bloated expenses. The finance director learns whether the forecast or the execution was wrong, and the commercial director gets a specific owner. That is why plan versus actual must be regular and on the same lines as the budget, otherwise the comparison is meaningless. Without it the same mistake repeats next season.

At a glance

What is this document and why is it needed?
An analysis of why actual profit differs from plan: how much came from volume, price, cost of sales and expenses. It drives the decision on what to fix first: the demand forecast, price, purchasing or expenses, and whom to ask about the deviation.
Who owns the document and how often is it rebuilt?
A financial analyst prepares, the finance director interprets; monthly (by the 7th to 10th) and at season end; one page for the owner, a breakdown by factors and owners for the commercial team.
What mistakes are the most common?
Actuals compared with an “updated” plan → deviations vanish, no learning → compare with a frozen plan. Factors do not add up to the total → trust is lost → sequential substitution and a sum check. Everything is blamed on the “market” → management errors are invisible → separate market and management factors.

For whom

Brand
A brand splits the deviation by channel and collection: volume (wholesale, site, marketplace), price and channel mix, unit cost (FX, fabric, sewing), returns, commissions, fixed costs. In Model B actual EBITDA is 20,594 against plan 35,048: a deviation of -14,454; the main factor is volume.
Retail
A retailer splits by store and category: traffic and conversion (volume), basket (price and discount depth), IMU and purchase price, shrink and returns, rent and staff. In Model R actual EBITDA is 13,029 against plan 34,800: -21,771. With a thin EBITDA margin a 1 to 2 pp deviation of gross margin means hundreds of thousands of roubles, so discount depth is shown as a separate factor.

What it consists of

What a bridge is

  • A bridge is a chain: plan profit → deviation factors → actual profit. Each factor is shown as a number; the sum of factors equals the total deviation exactly.
  • The sequential substitution method: take the plan model, replace one block of assumptions with actuals and recompute, then the next; the increase in result after each step is the factor effect. Factor order affects the split, so it is fixed in the methodology.

Factors for fashion

  • Volume: units sold at plan prices and margin. For a brand it is shipments and preorders, for a retailer traffic × conversion.
  • Price and discount depth: actual price against plan; end-of-season price reductions; channel mix (a cheaper channel grew faster).
  • Purchase cost: FX, fabric price, sewing, duties, freight; for a retailer purchase price and IMU.
  • Retailer: Returns and losses: return share, unsellable units, shrink, defects.
  • Marketing, commissions, acquiring: variable costs depending on revenue.
  • Fixed costs: payroll, rent, IT, management; deviations here come from decisions, not from the market.
  • Retailer: revenue = traffic × conversion × average check, so the bridge uses three factors. A store example: plan 100,000 visitors, conversion 20%, check RUB 5,000 excluding VAT (100,000 thousand); actual 95,000, 19%, RUB 5,200 (93,860). By sequential substitution: traffic -5,000, conversion -4,750, check 3,610; the sum -6,140 equals the deviation -6,140.

How to read and what to do

  • Market factors (demand, FX) are separated from management factors (price, purchasing, costs): the first correct the forecast, the second change actions.
  • For each factor an owner and an action are assigned; without this plan versus actual is a report, not a tool.
  • Guide thresholds (set your own): revenue differs from plan by more than 5% or EBITDA by more than 10%: a mandatory factor breakdown; an unexplained remainder above 5% of the total deviation: the substitution method is redone; the same direction of deviation two months in a row: the plan assumptions are revised, not only the actuals; one factor above half of the total deviation: an action and an owner are assigned the same day.

Brand and retailer: what is computed differently

  • Brand: the brand bridge includes preorders against actual buy-through, the channel mix (wholesale, site, marketplace) with different margins, and the FX part of cost separately from the supplier price.
  • Retailer: the retailer bridge includes traffic, conversion, check, markdown depth, losses (shrink) and rent; fixed store costs are shown as the result of decisions, not as a market effect.

Brand: how the document differs and what to add

  • Brand: the brand bridge starts with volume by channel: wholesale, site and marketplace are computed separately because price and returns differ.
  • Brand: price and discount for a brand are the wholesale price list, platform promotions and own-site discounts: the effect on revenue and on margin differs.
  • Brand: purchase cost is broken down into factory price, FX rate, freight and duty: FX and freight can be separated from the management decision.
  • Brand: platform returns and the write-off of unsellable units form a separate factor; a 3 pp rise in returns in a channel shows in the bridge before it reaches the balance sheet.
  • Brand: platform commissions, logistics and marketing form a “channels” package; plan-versus-actual on DRR and commission is checked against the platform report.
  • Brand: wholesale shipments move between months by shipment and partner acceptance; therefore the monthly bridge is read together with the cumulative season-to-date one.
  • Brand: deviation causes are kept in a reference list: factory, FX, partner, platform, demand; without it the report becomes a list of explanations.
  • Brand: a factory deadline miss is shown as its own factor: it moves season sales and the cost of money but not the purchase price.

Retailer: how the document differs and what to add

  • Retailer: the retailer bridge starts with traffic, conversion and average ticket: sales volume is decomposed into visitors and buyers before price.
  • Retailer: price and discount depth is the retailer’s main factor: a 1 pp rise in price reductions (15% regular) changes revenue and margin and is read by category.
  • Retailer: comparable stores and new points are separated in the bridge: opening a new store distorts volume without any change in old ones.
  • Retailer: shrink (1.2% of the base) and online returns (22%) form a separate factor; shrink growth shows via stocktakes with a lag, so an estimate is accrued.
  • Retailer: rent, staff and utilities form the fixed factor: an overspend there is almost always the result of indexation or an opening, not random noise.
  • Retailer: promotions and card discounts are shown apart from the general price reduction: their effect on margin and traffic is compared with the promotion plan.
  • Retailer: plan-versus-actual by store is built on four walls: revenue, gross margin, direct store costs; head-office costs are shown apart.
  • Retailer: a weekly plan-versus-actual is needed in season: a lag at week six can still be corrected by re-supplies and price, at week twelve only by clearance.

Control checks

  • The factors add up to the EBITDA deviation (brand -14,454, retailer −RUB 21.771 million): the residual is zero.
  • The plan version is frozen at the approval date: if the plan is rewritten retroactively, the bridge shows the team’s work, not the model.
  • Typical traps: the order of factors changes their sizes, the mixed effect is attributed to volume, actual and plan data in different units and without VAT in one part.

Where the data comes from

Data sourceFields
The plan model (the version approved at the start of the period)all assumptions and results; a frozen copy
Actual period datasales by channel and SKU, returns, purchases, FX, expenses
Reference list of deviation causescause code, owner, action

How to build it step by step

  1. Step 1: Freeze the plan model at the start of the period; make all comparisons against it, not against an “updated” plan.
  2. Step 2: Collect actuals in the same cuts and rates (excl. VAT, the same channels and terms).
  3. Step 3: Compare totals: revenue, gross margin, EBITDA, net profit; record the total deviation.
  4. Step 4: Fix the factor order (volume, price, cost, losses, variable costs, fixed costs).
  5. Step 5: Substitute actuals into the plan model one block at a time and recompute; record the increase in result after each step.
  6. Step 6: Check that the sum of factors equals the total deviation; if not, a block was missed.
  7. Step 7: Break down the largest factor deeper: by channel, category, store, collection.
  8. Step 8: Separate factors into market and management ones and assign owners and actions.
  9. Step 9: Update the forecast for the rest of the period (rolling forecast) and show the gap to budget.
  10. Step 10: Prepare a one-page report for the owner: the bridge chart, the three main causes, decisions and deadline.

Formulas

MeasureFormulasNote
Volume effectΔVolume = (Q_actual − Q_plan) × Price_plan × Margin_planUse contribution margin, not gross revenue: volume without margin is not profit.
Price effectΔPrice = (Price_actual − Price_plan) × Q_actualFor a retailer price = regular price minus discount depth.
Mix effectΔMix = Σ(Share_actual − Share_plan) × Q_actual × Unit marginIsolate it if channels or categories differ strongly in margin.
Cost effectΔCost = −(Cost_actual − Cost_plan) × Q_actualShow the FX part separately from the supplier price part.
Bridge checkΣ factor effects = Profit_actual − Profit_planSequential substitution guarantees equality; a gap means a missed block.
Retail bridge: traffic, conversion, checkΔtraffic = (T_a − T_p) × C_p × K_p; Δconversion = T_a × (C_a − C_p) × K_p; Δcheck = T_a × C_a × (K_a − K_p)The order of factors is fixed in advance: another order gives another split between factors with the same total. For margin add a step for unit gross margin and variable costs.
Analytical form and substitutionΣ analytical effects + the cross term = Profit_actual − Profit_plan; the substitution method allocates the cross term by orderThe formulas above give effects in analytical form while the example uses substitution: totals agree and only the cross term differs. Name the method in the report header and do not change it from period to period.

Formulas in notation

Deviation
Δ = Fact − Plan
For revenue and profit a positive value is overperformance, for costs overspend.
Volume and price effect
ΔRev = Σ (U_1 − U_0) · P_0 + Σ U_1 · (P_1 − P_0)
U_0, U_1 are planned and actual units, P_0, P_1 are prices; the first sum is the volume effect, the second the price effect.
Factor bridge
EBITDA_fact = EBITDA_plan + Σ_f Δ_f
Δ_f is the effect of factor f computed by substituting actual for plan along a chain; the order of factors is fixed in advance.
Bridge residual
ε = Σ_f Δ_f − (Fact − Plan) = 0
If the residual is not zero, a factor is missing from the chain or one effect is counted twice.

Worked example

EBITDA bridge from plan to actual by six factors, RUB thousand (sequential substitution)

LineBrandRetailer
Plan EBITDA35,04834,800
Factor: Sales volume−4,610−5,364
Factor: Price and discount depth−1,653−5,372
Factor: Purchase cost−3,682−3,165
Factor: Returns and losses−800−1,475
Factor: Marketing, commissions, acquiring−2,195−2,937
Factor: Fixed costs−1,514−3,458
Actual EBITDA20,59413,029
Sum of factors − deviation (must be 0)00
Revenue: plan206,665510,159
Revenue: actual190,856489,400
Net profit: plan21,05111,662
Net profit: actual10,210−6,221

Actuals are set as deviations of the assumptions from the plan: the brand sold 4 to 6% less, channel prices are 3 to 4% lower, purchasing is 4.2% costlier, returns are 2 to 3 pp higher, marketing and commission are higher, fixed costs grew; the retailer got 4% lower store sales with online growth, discounts deeper by 1 pp, a worse IMU, higher shrink, marketing and rent. The sum of six factors equals the total deviation (a check in the table: 0). The factor order is fixed, with another order the split changes but the total is the same.

Figures

From plan to actual: brand EBITDA by factor
Waterfall: Plan EBITDA 35,048, Sales volume −4,610, Price and discount depth −1,653, Purchase cost −3,682, Returns and losses −800, Marketing, commissions, acquiring −2,195, Fixed costs −1,514, Actual EBITDA 20,594010,00020,00030,00040,000Plan EBITDAPlan EBITDAPlan EBITDA: 35,04835,048Sales volumeSales volumeSales volume: −4,610−4,610Price and discount depthPrice and discountdepthPrice and discount depth: −1,653−1,653Purchase costPurchase costPurchase cost: −3,682−3,682Returns and lossesReturns and lossesReturns and losses: −800−800Marketing, commissions, acquiringMarketing,commissions,…Marketing, commissions, acquiring: −2,195−2,195Fixed costsFixed costsFixed costs: −1,514−1,514Actual EBITDAActual EBITDAActual EBITDA: 20,59420,594RUB thousand
  • positive
  • negative
  • total
Data as a table
StepChangeLevel
Plan EBITDA—35,048
Sales volume−4,61030,438
Price and discount depth−1,65328,785
Purchase cost−3,68225,103
Returns and losses−80024,303
Marketing, commissions, acquiring−2,19522,108
Fixed costs−1,51420,594
Actual EBITDA—20,594

From plan to actual: brand EBITDA by factor. Each step is one factor recomputed in sequence from plan to actual; the sum equals the deviation.

From plan to actual: retailer EBITDA by factor
Waterfall: Plan EBITDA 34,800, Sales volume −5,364, Price and discount depth −5,372, Purchase cost −3,165, Returns and losses −1,475, Marketing, commissions, acquiring −2,937, Fixed costs −3,458, Actual EBITDA 13,029010,00020,00030,00040,000Plan EBITDAPlan EBITDAPlan EBITDA: 34,80034,800Sales volumeSales volumeSales volume: −5,364−5,364Price and discount depthPrice and discountdepthPrice and discount depth: −5,372−5,372Purchase costPurchase costPurchase cost: −3,165−3,165Returns and lossesReturns and lossesReturns and losses: −1,475−1,475Marketing, commissions, acquiringMarketing,commissions,…Marketing, commissions, acquiring: −2,937−2,937Fixed costsFixed costsFixed costs: −3,458−3,458Actual EBITDAActual EBITDAActual EBITDA: 13,02913,029RUB thousand
  • positive
  • negative
  • total
Data as a table
StepChangeLevel
Plan EBITDA—34,800
Sales volume−5,36429,436
Price and discount depth−5,37224,064
Purchase cost−3,16520,899
Returns and losses−1,47519,424
Marketing, commissions, acquiring−2,93716,487
Fixed costs−3,45813,029
Actual EBITDA—13,029

From plan to actual: retailer EBITDA by factor. For the retailer the main contributions come from price and discount, online returns and fixed costs.

Specifics by category

Apparel
Apparel: in the breakdown show end-of-season price reductions and size returns; category mix (tops, bottoms, outerwear) changes margin.
Footwear
Footwear: show the size run and returns by last separately; forecast error in extreme sizes hits stock, not sales.
Accessories
Accessories: the price factor is stronger (gifting periods, promotions), returns are small; shrink and theft in stores are shown on a separate line.
Kids
Kids: the weather and school calendar factor creates a volume deviation, not price causes; split by age group.
Men
Menswear: low volatility, the bridge looks “cleaner”: the largest factors are usually purchase price and FX.
Women
Womenswear: mix and discount depth explain most of the deviation; split core, fashion and image.

Common mistakes

  • Actuals compared with an “updated” plan → deviations vanish, no learning → compare with a frozen plan.
  • Factors do not add up to the total → trust is lost → sequential substitution and a sum check.
  • Everything is blamed on the “market” → management errors are invisible → separate market and management factors.
  • Actuals on another basis (incl. VAT, other channels) → an artificial deviation → the same basis and cuts.
  • The analysis comes two months later → decisions are late → monthly by the 10th.
  • A factor has no owner → nothing changes → record an owner and an action deadline.

Russian specifics

  • The law does not require plan versus actual; but if a bank or investor requests it, agree the form and basis (incl. or excl. VAT, management or statutory profit) in advance.
  • Exchange differences and reserves in statutory actuals differ from management ones: show them as a separate reconciliation line. Rule: Order 180n of 15.11.2019: Ministry of Finance Order of 15.11.2019 No. 180n (FSBU 5/2019 Inventories) (rule check not completed).
  • If part of the deviation is caused by a change in platform terms or the VAT rate, state the effective date: for VAT of 22% it is 2026 (Law 425-FZ). Rule: Law 425-FZ of 28.11.2025: Federal Law of 28.11.2025 No. 425-FZ (amendments to parts one and two of the Tax Code; VAT 22%) (rule check not completed).

Open interactively →