Volume reconciliation tells you how many units moved. Trade spend reconciliation tells you whether the money attached to those units was correct. Most FMCG teams run the first every month and the second only when a deduction arrives that nobody recognises. A retailer deduction is not the problem. It is the visible tip of a trade term that was never written as a rule.

This guide covers the deduction cycle in five steps: why the money side runs on a different clock, how to encode terms as rules before the claims arrive, what a deduction line actually contains, the three passes that decide a claim, and the numbers that tell you whether the process is under control.

The key point Reconciling trade spend means comparing three things: what was agreed, what was claimed and what was earned. The first comparison is only possible if trade terms are written as rules before the claims arrive.

Why the money side runs late

Volume reconciliation has a fixed rhythm and a shared dataset. Every month, the same sources are compared at the same grain: sell-in against sell-out, shipments against POS, stock against ERP. The money side has neither. A retailer does not send a claim for each promotion at the end of the period; it deducts from the next payment, sometimes without a line item, sometimes with a promotion reference that does not match anything on your side.

The result is that by the time finance sees a deduction, the promotion period has already been reconciled, signed off and filed. The data needed to verify the claim — qualifying sell-out units at SKU level — has been archived under the previous month. The team ends up reconstructing a decision after the fact, against a retailer that holds the invoice and the payment. Public material on deduction management describes the same pattern from the retailer-facing side: claims arrive as short pays and chargebacks, and the work begins only when the money is already missing.

Three ownership problems make it structural. Sales negotiates the term, demand planning owns sell-out, finance owns payments. Nobody owns the chain from term to claim to settlement, so a deduction that mixes a valid allowance, a timing difference and an error in one line is argued invoice by invoice instead of being decomposed.

Write trade terms as rules first

The fix starts before the promotion runs. Write every trade term as a rule with the same discipline as a comparison rule: the product scope, the qualifying period, the qualifying movement (sell-in or sell-out, and at what point in time), the rate, any cap, and the exclusions or conditions. Give the rule an ID, an owner and effective dates.

When the rule exists, the expected amount is a calculation, not an opinion: qualifying units times the rate, subject to the cap. When the rule does not exist, every deduction is a one-off negotiation conducted on the retailer's timeline. The same logic that makes volume comparisons testable applies here — a term that can be written as a rule can be checked before the money moves.

The product scope is the first thing that fails. Retailer item codes rarely match your SKUs, and the mapping is what connects a claim line to a product set. If the scope is not resolved, the rate is applied to the wrong units and the expected amount is wrong before any claim arrives.

What a deduction actually contains

Retailers do not usually send a bill for a promotional allowance. They withhold the amount from a later payment, sometimes with a line item, often without one. The names vary — short pay, chargeback, billback, co-op, scanback, markdown support, price protection — but the structure is the same. A deduction is a financial claim that a trade term was met and that a specific amount is owed. The amount exists to pay for a behaviour: units sold, display kept, price lowered, feature advertised.

Reconciling a deduction is therefore not an accounting cleanup. It is checking three things in order: whether the term exists, whether the behaviour was earned, and whether the amount follows the agreed rate. Most disputes fail on the first check and should never reach the third. Analysts who work on the retailer side describe the same sequence: verify the deduction exists in the agreement, verify the event occurred, then verify the arithmetic. When a claim cannot pass existence, it should be disputed with the agreement as evidence, not paid under protest.

The three-pass reconciliation

Once terms are rules and claims are linked to promotions, the money side can be reconciled in three passes, each with a question and an action.

PassQuestionAction
ExistsDoes the claim correspond to a promotion and a term?Route to review or reject with evidence
AmountDoes it match the expected amount within tolerance?Approve for payment or flag
ResidualWhat is left, and why?Classify by cause, assign an owner, age it

Pass one — exists. Every claim line must carry the promotion ID and therefore the rule. If it does not, match by period, product and rate, and route anything that still does not match to review. An unlinked deduction is the same disease as an unmapped SKU in a sell-out file: the point where the process stops being traceable. This pass measures the quality of your term encoding, not of the retailer's claim.

Pass two — amount. The expected amount is a computation from the rule: qualifying units times the rate, capped. Compare it with the claimed amount within a defined tolerance. A claim that matches is approved and paid fast. A claim that does not is flagged before it reaches the residual, with the expected value attached, so the conversation is about a variance, not about a number out of nowhere.

Pass three — residual. What remains after the first two passes is classified by cause: timing (units still settling), retailer error, rate disagreement, or missing data. Every residual gets an owner and an age, and it is reviewed under the same exception lifecycle you would use for a volume exception. The point of the classification is that each cause has a different fix and a different evidence pack.

A worked promotional claim example

The following is illustrative. It shows one promotion where the claimed amount is above the capped earned amount, so the residual can be explained without treating every difference as a dispute or a loss.

Term or claim elementIllustrative valueWhy it matters
Promotion and claimP-2026-05 / D-8841Connects the deduction to one agreement and retailer reference
Retailer and product scopeRetailer ES-042 / SKU MX-2041Prevents the rate being applied to the wrong account or product
Qualifying period2026-05-01 to 2026-05-31Sets the dates used to select qualifying activity
Agreed rate and cap€0.12 per qualifying unit; cap €2,000Defines the expected calculation and maximum earned amount
Qualifying units18,000 unitsSource volume used to calculate the earned amount
Expected earned amount18,000 × €0.12 = €2,160; capped at €2,000Applies the term before comparing the retailer claim
Claimed deduction€2,050Amount withheld by the retailer
Residual€2,050 claimed − €2,000 earned = €50Amount requiring classification and evidence
Classification and evidenceCap overclaim; agreement v2, sell-out extract v3, deduction ref D-8841Shows why the €50 is not earned under the agreed rule
Settlement status€2,000 settled by credit note CN-7714; €50 disputed and closedSeparates what was earned and settled from what was rejected

The arithmetic is deliberately short: qualifying units × agreed rate gives €2,160, then the cap limits the earned amount to €2,000. The claim is €2,050, so the residual is €50. The evidence trail links the rule, the qualifying volume, the retailer reference and the settlement document.

What should balance?

The agreed term defines what can be earned. The earned amount is the result of applying that term to qualifying activity. The claimed deduction is what the retailer withholds, and the settled amount is what is finally credited or accepted.

  • Agreed term. The rule, scope, period, rate, cap and conditions.
  • Earned amount. The calculated entitlement after eligibility and caps.
  • Claimed deduction. The retailer's requested or withheld amount.
  • Settled amount. The amount paid, credited, disputed or rejected after review.

These values do not always become equal at the same moment. Timing, eligibility, caps, disputes and credit processing can leave a documented residual; the control is to classify it and retain the evidence, not to force every balance to zero.

The numbers that matter

Report the deduction run the way a demand planner reports a forecast: claims received, matched within tolerance, approved, disputed, paid, and the aging of open disputes, all by promotion and by retailer. Two ratios carry the diagnosis. The share of claims that pass pass one is the quality of your term encoding. The share that pass pass two within tolerance is the quality of your qualifying-volume data.

When pass one is low, the fix is the rule book: the terms exist but are not encoded, so the claims cannot be matched. When pass two is low, the fix is the volume reconciliation itself — the same sell-out and stock work the rest of these guides describe, because the expected amount is only as good as the qualifying units behind it. When both ratios are healthy and deductions still appear, the residue is commercial: a real negotiation about a real earned allowance, and it should be paid.

This is why trade spend reconciliation belongs in the same operating model as volume reconciliation rather than in a separate finance inbox. The volume side provides the qualifying movement that the money side depends on, and the money side provides a market test of whether the terms are working. Both use the same product, store and calendar mapping, and both are useless if the underlying volume data does not balance.

Where the deduction cycle usually breaksIn most teams it breaks in pass one: terms are negotiated but never encoded, so no claim can be matched and every deduction becomes a one-off argument. The first step is not a tool. It is writing the existing terms as rules and linking each claim to a promotion.

Talk to Marksyte

What this guide does not claim

Three limits. First, some deductions are correct. The retailer may have earned the money under the term as agreed, and the reconciliation should pay it faster, not fight it. Second, trade spend reconciliation is downstream of volume reconciliation: if sell-out and stock do not balance, expected amounts will not either, and no amount of rule-writing fixes that. Third, tolerances, dispute windows and approval levels are commercial decisions, not technical ones. The process makes those decisions explicit; it does not replace them.

Frequently asked questions

What is a promotional deduction in FMCG?

It is an amount a retailer withholds from a payment claiming it is owed under a trade term, such as a promotion allowance, co-op, markdown support or rebate. The names vary, but the structure is the same: a financial claim that a trade term was met.

How is trade spend reconciliation different from volume reconciliation?

Volume reconciliation balances units: sell-in against sell-out, shipments against POS, stock against ERP. Trade spend reconciliation balances money: agreed terms against claimed deductions and against amounts actually earned and settled. They share the same product, store and calendar mapping, but the sources and the owners are different.

Why do retailer deductions keep exceeding what we expected?

Usually because the expected amount was never computed. The term was not encoded as a rule, the claim is not linked to a promotion event, or the qualifying volume is measured differently by the retailer than by your sell-out data. Each cause has a different fix, and the first step is making the mismatch visible by rule and by promotion.

Sources

  1. BlackLine, What Is Deduction Management?.
  2. BlackLine, How to Handle Retailer Deductions: A Step-by-Step Guide.
  3. HighRadius, Retailer Deductions — The Hidden Risk to Your Margins.
  4. McKinsey, Trade Promotion Management and Optimization (TPM/TPO).
  5. ESM Magazine, Missing Data: How to Prevent Collaborative Profit Leakage.

The article separates published process descriptions, commercial interpretation and the site's own methodology. It does not claim that a specific share of retailer deductions is erroneous, because the sources do not provide a verifiable basis for that figure.