White Paper
The Trade Promotion Operating Model Simplified and Demystified
You Can Only Optimize the Promotions You Can Measure, Prove and Govern: A 7-Step View of Maturity and Where Trade Spend Leaks
- Authors
- Anuj Gupta and Chris Newton
- Published
- November 18, 2026
- Length
- 61 min read · 45-page PDF · version 1.0
On this page
Overview
A trade promotion is a price, a placement or an activity that a manufacturer pays the trade to deliver: a temporary price reduction, an end-cap display, a feature in the weekly circular, a distributor volume bonus, 0 percent financing on a riding mower sold through a dealer. The customer executes it, and the money moves afterward: off the invoice, as a deduction, as a bill-back, as a payment on scanned units, as a consumer rebate, as a subsidy to the dealer's floor plan lender. Trade promotion management (TPM) plans, funds, approves, sets up, accrues and settles those promotions. Trade promotion optimization (TPO) estimates what each one returns and chooses the ones that pay.
At any balance sheet date much of the money is still an estimate. General Mills carried $547.0 million of accrued trade and consumer promotions at May 31, 2026; Deere carried $828 million of accrued dealer sales incentives at November 2, 2025, plus $1,892 million of incentive accruals with a right of set-off against trade receivables; General Motors carried $6,485 million of dealer and customer allowances, claims and discounts at the end of 2025. None of the large manufacturers read for this paper discloses trade spend as a share of gross sales. The Promotion Optimization Institute says consumer goods companies have historically spent “between 11% and 27% of revenues” on trade promotion, a survey claim without a stated method, and Drèze and Bell call trade promotion “second only to cost of goods sold as a profit-and-loss item”.
This paper makes two claims. First, trade spend is approved on lift and paid on claims, and the money leaks in between: on units that would have sold anyway, on volume borrowed from later weeks, on deals not passed through to the shelf, on displays paid for and never built, on deductions nobody disputed in time. In the authors' implementation experience, much of it is created in design and setup and found late, in a post-event review that arrives after the next calendar is locked, in deduction aging, in a true-up at close, or never. Second, a manufacturer can only optimize the promotions it can measure, prove and govern: an optimization model is as good as the baseline under it, the settled cost fed back into it, and the controls that keep an approved plan intact on the way to the shelf. The paper sets out a seven-step workflow and a five-level maturity model, and treats packaged goods sold through retailers and durable goods sold through dealers side by side.
Source: General Mills FY2026 Form 10-K, filed July 1, 2026; Deere FY2025 Form 10-K, filed December 18, 2025; General Motors FY2025 Form 10-K, filed January 27, 2026; Promotion Optimization Institute release, January 30, 2023; Drèze and Bell, Marketing Science 22(1), 2003
Section 1
What is trade promotion, and what does it buy?
Manufacturers describe trade promotion by what it buys: Simply Good Foods says its “trade promotions focus on obtaining retail feature and display support”. The people who run the programs describe it by how the money moves. Both views are needed, because the same display can be paid off the invoice, by bill-back or by scan-back, and each route changes who carries the risk. The route is a design decision with a price attached, which is why this paper treats it in step 2, design.
| Vehicle | What it buys | How the money moves | Who carries the risk |
|---|---|---|---|
| Off-invoice allowance | A lower purchase price for a period | Off the invoice at purchase; Toro: “an immediate reduction in the selling price” | The manufacturer, on every unit bought in the window |
| Bill-back | A price cut, display or feature, once proven | Claimed after the event; the X12 812 notifies “an adjustment or billback” | Shared: the manufacturer validates; the customer waits or deducts |
| Scan-back | Units sold to consumers at the promoted price | Paid on units scanned during the event | The customer, for stock not sold on promotion |
| Lump sum and slotting | Placement, shelf space, a feature or an event | A fixed fee; for slotting, per the FTC, paid “as a condition for the initial placement” | The manufacturer, whatever the item sells |
| Cooperative advertising | Advertising by the customer | Against claims with “evidence of the advertising material procured/produced” (Toro) | The manufacturer, if proof is not checked |
| Markdown and margin support | The retailer's margin on a price cut or clearance | Credits or deductions; Walmart lists “markdowns, margin protection”, Target “markdown allowances” | The manufacturer, often without the consumer price record |
| Retail media | Ads on the retailer's sites, apps and stores | Paid to the retailer's media business; Target books some of that income as a reduction of cost of sales | The manufacturer, unless results are measured |
| Coupons and rebates | A lower consumer price | Redeemed through a clearing house or rebate processor | The manufacturer, until redemption settles |
| Dealer programs | Sell-through, stock and selling effort | Customer cash, financing, dealer cash, bonuses, floor plan support (section 2) | Shared: the dealer carries stock and tier risk; the manufacturer, the estimate |
Source: Simply Good Foods FY2026, Toro FY2025, Walmart FY2026 and Target FY2025 Forms 10-K; X12 812 transaction set purpose statement, x12.org; FTC press release, November 14, 2003
Section 1
Who pays whom, and on what document?
Toward the shelf move product, price and the terms of the deal: item data (EDI 888), the promotion announcement (889), which X12 describes as carrying “promotion allowance amounts, dates and performance conditions”, and the invoice. Back toward the manufacturer move data and money already taken: point-of-sale and inventory data (852), distributors' resale reports (867) and their claims against contract prices (844, answered on the 849), the remittance advice that shows what the customer short-paid (820), and the debit or credit notice that explains the deduction (812). In the dealer channel the dealer reports the retail sale that triggers customer cash and its own incentives, and the manufacturer pays for the interest-free period it promised on the dealer's stock. The manufacturer sees its own shipment the day it happens and the consumer's purchase weeks later, if it buys the data. Everything in sections 2 and 3 about measurement follows from that gap.
Source: X12 transaction set purpose statements (812, 820, 844, 849, 852, 867, 888, 889), x12.org; Deere and Toro FY2025 Forms 10-K
Section 1
What do customers' rules say about deductions, proof and timing?
Most deductions settle terms both sides agreed; the customer takes the money when it believes it is owed, and the manufacturer validates afterward, on the customer's clock. The published rules show the shape every program carries.
| Rule | As published or reported | Source |
|---|---|---|
| Dispute window | Kroger: “All claims are due within 180 days of the deduction date”; a claim without supporting documentation “will result in the claim being denied” | Kroger supplier training document, 2022 |
| Distributor deductions | US Foods: “Vendor consideration is typically deducted from invoices or collected in cash within 30 days of being earned” | US Foods FY2024 Form 10-K |
| Compliance charges | Target charges vendors for “merchandise shipments that do not meet our requirements”, such as late or incomplete shipments | Target FY2025 Form 10-K |
| Delivery fines | Walmart's on-time, in-full program, per trade press: a fine of 3 percent of the cost of goods that miss; thresholds rose from 75 percent (2017) to 98 percent (2020) and were reported in 2024 as 90 percent on time and 95 percent in full | Talk Business & Politics, 2017 and 2020; Supply Chain Dive, 2019; SupplyChainBrain, 2024 |
| Proof for co-op | Toro's dealers and distributors submit claims “along with evidence of the advertising material procured/produced” | Toro FY2025 Form 10-K |
| Consumer rebate deadlines | Texas: pay within the time promised or, if the offer is silent, within 30 days of a complete request (rebates of $10 or more). Washington: allow at least 14 days to apply; pay within 90 days | Tex. Bus. & Com. Code 605.001, 605.002; RCW 19.315.020 |
Deduction work reconciles money already taken: an invalid deduction not disputed inside the window is lost by default, and a condition never recorded cannot be proved. Dealers usually file claims that are validated before payment. A 2012 Credit Research Foundation survey, the only non-vendor figure found, put deductions at a median 2 percent of sales for respondents selling mainly to retailers. Across all 493 respondents in 23 industries, a median 3 to 5 percent of deduction dollars were found invalid and charged back, not counting deductions written off without dispute, and by dollars pricing was the largest deduction type, ahead of advertising, promotion and rebates.
Source: as listed in the table; Credit Research Foundation, 2012 Customer Deduction Survey (493 companies in 23 industries, fielded April 2012), presented at the Federal Reserve Bank of Minneapolis Remittance Coalition, Fall 2012
Section 1
How is trade promotion accounted for, on both sides of the shelf?
Under ASC 606 a payment to a customer reduces the transaction price “unless the payment to the customer is in exchange for a distinct good or service”, and any amount above that service's fair value reduces revenue as well. In practice almost all trade promotion reduces revenue. Deloitte's and PwC's examples treat slotting fees as not distinct, and KPMG expects that “Most slotting fees will continue to reduce revenue”. Consumer coupons and rebates redeemed through a retailer are treated as if paid to the retailer: PwC concludes the manufacturer “should account for the rebate in the same manner as if it were paid directly to the Retailer”. Cooperative advertising turns on judgment: it is expensed only when it buys a distinct service whose fair value can be reasonably estimated, and only up to that value. Campbell's nets co-op against revenue; Generac expenses it “when claimed by the customer”.
| Manufacturer | Retailer | |
|---|---|---|
| Where the dollar lands | A reduction of revenue, as variable consideration | A reduction of cost of sales. Walmart: payments from suppliers “are accounted for as a reduction of cost of sales” |
| When it is recognized | At the sale, for programs offered or expected under customary business practice; later only if no promise existed at the sale (ASC 606-10-32-27). Deere's estimate covers “announced and expected incentive programs” | When the related product is sold; at Kroger, where systems cannot track it, “based on inventory turns” |
| What is estimated | Participation, performance, redemption and timing. Clorox lists “customer participation rates”, the rate customers meet performance criteria, product availability and historical redemption | The amount earned and not yet received. Target computes its vendor income receivable “by forecasting vendor income collections” |
| Where the difference surfaces | A deduction the manufacturer disputes, accepts or writes off | A receivable the retailer collects by deduction |
The same dollar is estimated twice, from different data, by two parties with opposite interests, and deductions are where the two estimates meet. The constraint on variable consideration (ASC 606-10-32-11) works in one direction for a promotion: because the promotion reduces the price, an uncertain estimate is resolved toward a larger accrual and less revenue, so that a significant reversal of revenue already recognized is unlikely later. The estimation method is whichever of expected value and most likely amount better predicts the amount (ASC 606-10-32-8), applied consistently.
Source: ASC 606-10-32-8, 32-11 and 32-25 to 32-27 as reproduced in Deloitte, Roadmap: Revenue Recognition (2025) and FASB Transition Resource Group memos 14 and 19; PwC Viewpoint, Revenue guide, section 4.6; KPMG, Revenue for consumer products companies, 2016; Campbell's FY2026, Generac FY2025, Deere FY2025, Clorox FY2026, Walmart FY2026, Kroger FY2025 and Target FY2025 Forms 10-K
Section 1
How large are the accruals, and how are they audited?
| Company | What the filing shows | Filing |
|---|---|---|
| General Mills | Accrued trade and consumer promotions of $547.0 million at May 31, 2026 ($527.2 million a year earlier); differences between estimate and actual “are recognized as a change in estimate in a subsequent period” | FY2026 Form 10-K |
| Hershey | Unsettled promotional costs of $227.7 million in accrued liabilities at December 31, 2025; actual costs “have not deviated from the estimated amount by more than 3%” in 2023 to 2025 | FY2025 Form 10-K |
| Clorox | A 10 percent change in its trade promotion accrual estimates would move net sales by about $21 million | FY2026 Form 10-K, Exhibit 99.1 |
| Utz Brands | Promotion reserves of $26.9 million at December 28, 2025, including $15.4 million of customer adjustments awaiting final processing | FY2025 Form 10-K |
| Deere | Dealer sales incentives of $828 million in accrued expenses, and $1,892 million of incentive accruals with a right of set-off against trade receivables, at November 2, 2025; recorded “At the time of the sale to a dealer” | FY2025 Form 10-K |
| General Motors | Dealer and customer allowances, claims and discounts of $6,485 million at December 31, 2025 | FY2025 Form 10-K |
At two smaller filers, auditors describe testing the accrual against history and against later claims. Jones Soda's auditor reported “comparing current promotional offer redemptions to historical estimates” and comparing allowances applied after year end with the year-end accrual; Bridgford Foods' auditor recalculated the reserves independently, ran sensitivity analysis and “Considered transactions submitted by customers subsequent to year-end”. Clorox's FY2026 auditor's report contains no promotion critical audit matter. The SEC staff asked WD-40 in 2020 why it estimated promotional variable consideration with the most likely amount method instead of the expected value method, and WD-40 concluded that its process was the expected value method. It asked Utz's merger partner why an under-accrued trade spend reserve, trued up in later periods, was not the correction of an error; the response pointed to a system implemented in 2019 that improved the estimate, and treated the change as one in estimate. A promotion accrual is as defensible as the program records behind it.
Source: Forms 10-K as listed, sec.gov; Jones Soda Form 10-K/A, filed May 22, 2026 (2024 auditor's report); Bridgford Foods FY2025 Form 10-K; Clorox FY2026 Exhibit 99.1 (auditor's report); SEC staff correspondence with WD-40 Company (March 2020) and Collier Creek Holdings (August 2020), sec.gov
Section 2
Where do the extra units come from?
A promotion's gross lift is the difference between what sold during the event and what would have sold without it. Where those extra units come from decides whether the promotion made money, and the answer depends on how they are counted.
Gupta's 1988 study of coffee purchases found “more than 84% of the sales increase due to promotion comes from brand switching”, a figure still quoted as proof that promotions mostly take share. It is a share of elasticity, and averaged across the household-panel studies from Gupta to Bell, Chiang and Padmanabhan, switching is about three quarters of it. Counting units in household panel data, van Heerde, Gupta and Wittink found that “on average about one third of the unit sales increase is attributable to losses incurred by other brands”, a group that does not separate the manufacturer's own brands; with store data, van Heerde, Leeflang and Wittink found that cross-brand switching, cross-period effects and category expansion “each of these three parts contribute about one third on average”. Units won from competitors and genuine category growth are incremental to the manufacturer, until competitors respond. Units bought early mostly are not, although Ailawadi, Gedenk, Lutzky and Neslin found that stockpiling can raise consumption. Units switched from the manufacturer's own items are not.
The weeks around a promotion pay part of the bill. In store data for tuna and toilet tissue, van Heerde, Leeflang and Wittink estimated combined pre- and post-promotion dips “between 4 and 25% of the current sales effect”, and noted they are rarely found at the category or brand level. Over the longer run, Nijs and colleagues found across 560 categories in Dutch supermarkets that price promotions rarely change category demand for good: effects fade over about ten weeks and the “long-term impact is essentially zero”. Srinivasan and colleagues found that a price promotion “typically does not have permanent monetary effects for either party”, with “a predominantly positive impact on manufacturer revenues” while “retailer category margins are typically reduced by price promotions”. On this evidence a price promotion should pay for itself inside the event and the weeks around it. The retailer weighs a different trade, its category margin against traffic, which is one reason the two sides rarely agree on what an event was worth.
Source: Gupta, Journal of Marketing Research 25(4), 1988; Bell, Chiang and Padmanabhan, Marketing Science 18(4), 1999; van Heerde, Gupta and Wittink, Journal of Marketing Research, 2003 (working paper version read); van Heerde, Leeflang and Wittink, Journal of Marketing Research 37(3), 2000, and Marketing Science 23(3), 2004; Ailawadi, Gedenk, Lutzky and Neslin, Journal of Marketing Research 44(3), 2007; Nijs et al., Marketing Science 20(1), 2001; Srinivasan et al., Management Science 50(5), 2004
Section 2
What does a promotion's profit look like, line by line?
Lift is not profit. The worked example uses round numbers chosen for clarity; it is no company's data. The item sells to the retailer at $4.00 and has a variable cost of $2.40, a unit margin of $1.60. Without the event, consumers would buy 10,000 units. The promotion takes $0.80 off each unit and adds a $2,000 display fee. Consumers buy 22,000 units during the event, per the retailer's scan data, a gross lift of 120 percent, and the retailer buys exactly what it sells. Of the extra units, 3,000 were bought earlier than they otherwise would have been at full price, and 1,500 came from the manufacturer's own sister items, which earn the same $1.60. The discount is paid on all 22,000 units, $8,000 of it on units that would have sold anyway.
| Measure | Result | What it leaves out |
|---|---|---|
| Gross lift | +120 percent (12,000 units) | Cost, and where the units came from |
| Revenue return | Incremental revenue of $30,400 (22,000 units at $3.20, less 10,000 at $4.00) over trade spend of $19,600: 155 percent | Margin, and the borrowed and cannibalized units |
| Gross profit return | Incremental margin before trade of $12,000 over $19,600: 61 percent, against a break-even of 100 percent | Nothing; the same answer stated before trade spend |
| Net profit return | Incremental profit of minus $7,600 over $19,600: minus 39 percent, against a break-even of zero | Nothing; the same answer stated net of trade spend |
Even if every extra unit were incremental, the event would need 12,500 units of lift, 125 percent, to break even: each unit sold at the deal price keeps $0.80 of margin, so recovering the $8,000 discount on base volume takes 10,000 extra units and the fee another 2,500. With borrowing and cannibalization at 37.5 percent of lift, as here, break-even rises to 500 percent. Depth is the first lever, because break-even lift climbs steeply as the discount approaches the margin. The mechanism matters for a different loss: off-invoice terms with a retailer that buys ahead would add the discount on units sold later at full price, which a scan-back would not pay. Incremental profit is the closest estimate of what the event added to operating profit; no income statement line shows it, because the baseline is a model. Every lever is cheap to pull in step 2 and gone by step 7.
Source: worked example by the authors, with round numbers chosen for clarity
Section 2
How much of a trade deal reaches the shelf?
A trade deal works through the customer's own decisions. Retailers set their own shelf prices, and trade terms pay for the performance the two sides agree: whether to cut the price, by how much and for how long, and how much stock to buy while the deal lasts. Forward buying and incomplete pass-through take money out of a promotion before any consumer sees it.
| Off-invoice | Bill-back | Scan-back | |
|---|---|---|---|
| Paid on | Every unit bought in the deal window | Units bought in the window, once performance is proven | Units scanned at the promoted price during the event |
| Forward buying | Rewarded: stock bought at the deal price can be sold later at the regular price | Rewarded, unless the claim is limited to units sold on promotion | Not paid: units sold after the event earn nothing |
| Pass-through | Up to the customer; the deal is funded whether or not the shelf price moves | As far as the performance written into the terms | Built in: payment requires the consumer price |
| Data the manufacturer needs | Its own shipments | Shipments and proof of performance | The customer's point-of-sale data, and the right to check it |
Forward buying is the customer buying more at the deal price than it sells on promotion. Drèze and Bell start from the industry problem that “manufacturers often lose money on these deals as a result of forward-buying by retailers”; in their analysis, on identical terms, “retailers prefer off-invoice trade deals and manufacturers prefer scan-backs”, but manufacturers can redesign the scan-back “to leave the retailer weakly better off and leave themselves strictly better off”. Pass-through is the share of the deal that reaches consumer prices, and its measured size depends on the definition. Besanko, Dubé and Gupta found own-brand price pass-through “on average, more than 60% for 9 of 11 categories” at one supermarket chain. Ailawadi and Harlam, measuring dollars at one retailer, found that it “passes through more than 100% of the total manufacturer funding it receives in aggregate”, while “the median pass-through rate for individual manufacturers is much lower than 100%”. Nijs and colleagues, across more than 1,000 retailers, estimated mean pass-through elasticities (the percent change in shelf price per percent change in the manufacturer's price) of 0.71 for wholesalers, 0.59 for retailers and 0.41 for the channel. Counting accepted deals, Armstrong, Gerstner and Hess reported in 1994 that “More than one-third of the accepted promotions are pocketed by the retailer”. These measures do not average into one number. The working question for each customer and event is narrower: did the agreed price, display and feature appear, and for how long.
Source: Drèze and Bell, Marketing Science 22(1), 2003; Besanko, Dubé and Gupta, Marketing Science 24(1), 2005; Ailawadi and Harlam, Marketing Science 28(4), 2009; Nijs et al., Marketing Science 29(2), 2010; Armstrong, Gerstner and Hess, NEC-63 conference proceedings, 1994
Section 2
What is a baseline, and who owns it?
Every measure of incrementality starts from a baseline: what would have sold without the promotion. It is a model estimate, never an observation. NielsenIQ defines base sales as “A statistically calculated NIQ-derived measure to determine the expected sales for a product in the absence of in-store promotional activity”. Circana defines them as “modeled volume expected at regular price without merchandising” and calls the promoted volume that is not incremental “Subsidized volume”. Abraham and Lodish's PROMOTER system, published in 1987, estimated “a baseline of what sales would have been had a promotion not been run”. The principle has not changed in thirty-nine years, and neither have the ways it goes wrong.
| Baseline built from | What it sees | What it misses |
|---|---|---|
| Manufacturer shipments | Every unit sold, at once and at no cost | Consumer purchases: forward buying looks like lift, and the dip lands in a later period or another customer |
| Syndicated retail measurement | Consumer sales, price, distribution and promotion in covered stores; NielsenIQ says it collects data “from nearly 900,000 stores globally” | Incremental is total minus base, so dips, consumer stockpiling and own-item switching count as lift; items on deal most weeks leave few clean weeks to fit the base; stores not covered; a delay of weeks |
| Retailer data | The retailer's own point-of-sale data. Walmart Luminate offers a free Basic tier and a paid Charter tier; Kroger's 84.51° Stratum is sold in subscription packages | Other retailers, and the method behind any baseline the retailer supplies |
| The manufacturer's own model on consumption data | Effects by item, customer and week, with dips and cannibalization modeled | Model error: checked against out-of-sample weeks, with incrementality tested against matched control stores, ad zones or dealer regions where the customer allows |
Lock the baseline method before approval and leave the number open. Keep the approval forecast to judge the plan, and judge the event against a baseline re-estimated afterward by the locked method, adjusted only for conditions outside the promotion, such as a competitor's event, a stock-out or a distribution change; report both, and the gap. Count the volume that would have sold anyway as a cost: in the worked example it absorbed $8,000 of the $19,600 spent. Use the latest cost, settled where final and otherwise accrued plus open deductions and claims, flagged by settlement status and re-run when the event settles. An evaluation that pairs this month's baseline with last quarter's planned cost measures neither.
Source: NielsenIQ base sales definition and Retail Measurement Services page; Circana Liquid Data Go CPG Dictionary, 2026; Abraham and Lodish, Marketing Science 6(2), 1987; Talk Business & Politics, October 17, 2023 (Walmart Luminate); 84.51° Stratum launch release
Section 2
How do dealer and durable-goods programs work?
Durable goods sold through dealers carry the same economics with different instruments. The unit is expensive, the purchase infrequent, the dealer holds stock on borrowed money, and the buyer often finances the purchase. Deere offers “sales incentive programs to promote the sale of our products from the dealer to the retail customer” and records their estimated cost “At the time of the sale to a dealer”, from inputs that include “field inventory levels”. The cost is booked at wholesale and settled at retail.
| Program | What it buys | Who receives it | When the cost is known |
|---|---|---|---|
| Customer cash rebate | A lower price to the buyer | The buyer, through the dealer | At the retail sale |
| Promotional financing | A lower rate or payment | The buyer; the manufacturer pays the lender to buy down the rate (interest subvention) | Estimated at the sale to the dealer from take rate and term mix; fixed when the retail contract is written. Ford includes the cost “when the vehicle is sold to the dealer”; GM Financial earned $1,404 million of interest subvention and $1,796 million of leased vehicle subvention in 2025 |
| Dealer cash, volume and stair-step bonuses | Dealer price and selling effort | The dealer | For volume programs, at “the end of the measurement period” (Deere) |
| Pool funds and SPIFFs | Dealer support earned on sales; payments to dealer salespeople per unit | The dealer or its staff | Deere awards “funds based on new equipment sales” |
| Early-order programs | Stock taken before the season, often on extended dating | The dealer | Over the dating period; Deere sets interest-free periods by equipment type and “the time of year of the sale” |
| Floor plan support | Interest-free or reduced-rate periods on dealer stock | The dealer's lender | When the unit sells or the free period ends, whichever comes first. Deere's periods “range from one to twelve months for most equipment”; Toro's inventory financing costs were $79.4 million in fiscal 2025, down from $114.7 million in fiscal 2023 |
| Cooperative advertising | Local advertising by the dealer | The dealer, against proof | When the claim is validated |
Who receives the money decides how much reaches the buyer. In vehicle transaction data, Busse, Silva-Risso and Zettelmeyer found that “Customers obtain 70 to 90 percent of a customer rebate, but only 30 to 40 percent of a dealer discount promotion”. A Federal Reserve study of US vehicle sales found cash rebates in about 57 percent of sales and interest subvention in a little more than 40 percent by the end of 2003, at about equal average value. Vehicle programs add aged-unit support and dealer holdback, and Deere lists “demonstration programs” among its dealer support. In a multi-line dealer, as in much of lawn and garden and outdoor power equipment, each brand's programs compete for floor space, stocking and selling effort, which makes what the money bought harder to define and more important to record. The planning unit is the stack: every program that can apply to one unit, added up at the unit level against the margin the unit carries. Financing cost moves with take rate and term mix, and a stair-step bonus approved on one forecast costs more when dealers pull orders forward to reach the next tier.
Source: Deere, Toro and Ford FY2025 Forms 10-K; General Motors FY2025 Form 10-K (GM Financial segment data); Busse, Silva-Risso and Zettelmeyer, American Economic Review 96(4), 2006; Corrado, Dunn and Otoo, Federal Reserve Finance and Economics Discussion Series 2006-09
Section 2
Where are the regulatory limits on promotional payments?
Most of the rules that bear on trade promotion set conditions on how payments are offered and how offers are described. They are listed in the sources' own words where possible, so that the people who design and approve programs know which questions to route to their advisers.
| Rule | What it looks at | Recent activity |
|---|---|---|
| Robinson-Patman Act, sections 2(d) and 2(e) | Paying for, or furnishing, promotional services or facilities, which must be available “on proportionally equal terms to all other customers competing in the distribution” of the products | An FTC complaint against PepsiCo under 2(d) and 2(e) was filed January 17, 2025 and dismissed in May 2025; Commission statements at filing and at dismissal took different views of the case |
| FTC Guides for Advertising Allowances (Fred Meyer Guides), 16 CFR Part 240 | How a seller may meet 2(d) and 2(e). The Guides say a seller is “well advised to put the plan in writing”, describe alternative terms, notice to competing customers and functional availability, including to retailers that buy through distributors, and state that “The use of intermediaries does not relieve a seller of its responsibility to comply with the law” | Revised in 2014 (79 FR 58245, effective November 10, 2014), adding online advertising to the covered media |
| Robinson-Patman Act, section 2(a) | Price discrimination between competing purchasers, with an element of competitive injury | FTC v. Southern Glazer's, a distributor case under 2(a) and FTC Act section 5. The stipulated decree filed October 2, 2026 runs six years with an independent monitor and defines net price after discounts, rebates, billbacks, allowances and scan rebates |
| Trade cost accounting | How promotion and rebate costs are recorded | Aurora Foods (2001: the SEC alleged trade marketing expenses under-reported by more than $43 million); Monsanto (2016: according to the SEC, rebate costs recorded late; $80 million penalty) |
| Sales pulled forward or reported | Sales moved into earlier periods without adequate disclosure, or reported falsely | According to the SEC: Diageo (2020, distributors “pressured” to buy in excess of demand), Under Armour (2021), Newell Brands (2023), Elanco (2024, discounts, rebates and extended terms at quarter end) and FCA US (2019, dealers paid to report vehicle sales that had not occurred) |
The cases are listed as reported by the agencies, to show the questions regulators have asked. Some of the sources name records: the Guides say a seller is “well advised to put the plan in writing”, and 27 CFR 6.81 requires alcohol industry members to keep three years of records of certain items furnished to retailers.
Source: 15 U.S.C. 13 (govinfo.gov, current through November 2026 per uscode.house.gov); 16 CFR Part 240 and 79 FR 58245; FTC press releases and Commission statements of December 12, 2024, January 17, 2025, May 22, 2025 and October 2, 2026; SEC Litigation Release 16866 (2001) and press releases 2016-25, 2019-196, 2020-36, 2021-78, 2023-210; SEC order in Elanco Animal Health, November 12, 2024; 27 CFR 6.81
Section 2
Which rules apply to consumer offers, financing and alcohol?
| Rule | What it looks at |
|---|---|
| Former price comparisons, 16 CFR 233.1 | A former price is a legitimate basis for comparison if it is “the actual, bona fide price at which the article was offered to the public on a regular basis for a reasonably substantial period of time” |
| Free offers, 16 CFR 251.1 | The seller should not recover the cost of the free item by marking up the regular price; “No more than three such offers should be made in the same area in any 12-month period”, among other limits |
| Mail-in rebates, state law | Payment deadlines and application windows, such as those in Texas and Washington (section 1) |
| Promotional financing, Regulation Z | For closed-end credit, stating a payment amount, number of payments, down payment or finance charge in an advertisement triggers further disclosures, including the annual percentage rate (12 CFR 1026.24(d)). For open-end deferred interest offers, a "no interest" claim must also state “if paid in full” (12 CFR 1026.16(h)) |
| Lease advertising, Regulation M | Advertisements for consumer leases that state certain terms, such as a payment amount, must carry further disclosures (12 CFR 1013.7) |
| State motor vehicle franchise laws | Many states set rules for manufacturer incentive programs offered through franchised dealers, including the timing of claims, payments and audits; they vary by state and were not reviewed for this paper |
| Alcohol, Federal Alcohol Administration Act (27 CFR Parts 6 and 10) | Paying or crediting a retailer for advertising, display or distribution is listed as a means to induce (6.21); purchasing or renting shelf or display space, “i.e. slotting allowance”, is named among practices that put retailer independence at risk (6.152); 6.96 sets coupon conditions: redemption by all retailers in the market, and reimbursement at no more than face value plus a customary handling fee. TTB announced a $12.5 million non-prosecution agreement with Southern Glazer's for bribing retailer employees on September 10, 2026. State alcohol laws add rules not covered here |
Source: 16 CFR 233.1 and 251.1 (govinfo.gov, 2024 edition); Tex. Bus. & Com. Code ch. 605; RCW ch. 19.315; 12 CFR 1026.16, 1026.24 and 1013.7 (consumerfinance.gov, accessed November 18, 2026); 27 CFR 6.21, 6.96, 6.152 and 10.21 (govinfo.gov, 2024 edition); TTB press release, September 10, 2026
Section 3
What is trade spend leakage?
Leakage is trade spend that bought less than it was approved to buy, or paid more than the terms required. Effectiveness kinds explain why an event or program returned less than it cost: the money reached the customer as agreed and the promotion still did not pay. They overlap, so they are measured on events that did not pay back and are never added together. Execution kinds are cash paid that was not owed; they can be counted and recovered.
| Kind | What it looks like | Family |
|---|---|---|
| 1. Subsidized base volume | The discount paid on units that would have sold anyway, beyond what the event's extra margin recovered | Effectiveness |
| 2. Borrowed and cannibalized volume | Lift taken from the weeks around the event or from the manufacturer's own items | Effectiveness |
| 3. Forward buying and diversion | Deal-price units sold after the event at the regular price, or outside the market the deal funded | Effectiveness |
| 4. Pass-through shortfall | An unconditional allowance funded, and the shelf price cut smaller, shorter or not at all | Effectiveness |
| 5. Fixed fees and media, return unmeasured | Slotting, lump sums and retail media bought with no read-out | Effectiveness |
| 6. Tier payouts on pulled-forward volume | Volume and stair-step bonuses earned on orders pulled forward without matching sell-through | Effectiveness |
| 7. Stacking beyond design | More programs on one unit than the plan priced | Both |
| 8. Agreed performance not delivered or not proven | Display, feature, advertising or dealer co-op paid without the performance or proof the terms require | Execution |
| 9. Ineligible and unwound claims | Incentives paid on units, buyers or dates outside the program, on retail sales reported early or falsely, or not recovered when a sale reverses | Execution |
| 10. Invalid deductions written off | Invalid deductions not disputed, or disputed after the window; small balances written off by policy are tracked apart | Execution |
| 11. Duplicate and off-term payments | One promotion paid twice; claims paid on items or dates outside the terms | Execution |
Source: IMA360 analysis; Hershey FY2025 Form 10-K; ASC 250-10-20 (definition of an error)
Section 3
Where is each kind created, and where is it found?
The pattern is the one the earlier papers in this series found for rebates and chargebacks. Effectiveness kinds are created in steps 1 to 3, where the purpose, depth, mechanism, timing and customer of each event are chosen and approved, and they are found in step 7, where events are measured on consumption data at the latest cost; without that measurement they are never found. Execution kinds are created in steps 4 and 5, where terms and conditions are recorded and performance is proved, and they are found in step 6, as a deduction that cannot be matched or a claim that cannot be proved, or in an audit. An organization that fixes leakage where it is found fixes the same event every quarter. One that fixes it where it is created changes the next calendar.
Source: IMA360 analysis, November 2026
Section 3
How do you size trade spend leakage in your own book?
From your own records, as rates by event and by customer. Each test uses data the manufacturer already holds or already buys.
| Kind | Test | Data needed |
|---|---|---|
| 1. Subsidized base volume | Discount paid on baseline units against the event's extra margin, on events that did not pay back | Baseline, deal rate, latest event cost |
| 2. Borrowed and cannibalized | Consumption against baseline in the weeks before and after each event (Pauwels and colleagues found post-promotion adjustment periods of “on average 2 weeks, at most 8 weeks”), and volume shifts across the manufacturer's own items | Weekly consumption by item and customer |
| 3. Forward buying and diversion | Customer purchases in the deal window against its promoted consumption, and shipments against the funded market | Shipments, consumption, ship-to data |
| 4. Pass-through shortfall | Shelf price change against unconditional funding, by event | Retailer or syndicated price data; deal terms |
| 5. Fixed fees and media | Share of fees and retail media spend with a documented read-out | Contracts; post-period results |
| 6. Tier payouts | Tier payouts against retail sell-through in the qualifying period | Dealer orders, retail sales, program terms |
| 7. Stacking | Total program cost per unit by customer and week, against the planned stack | Terms of every program that can apply to the unit |
| 8. Performance and proof | Share of performance payments with proof attached before payment | Claims and proof records |
| 9. Ineligible and unwound claims | Claims tested against eligibility, retail sale records and reversals | Dealer claims; retail sale and registration records |
| 10. Invalid deductions written off | Invalid promotion deductions written off without dispute, or disputed after the window, as a share of the total | Deduction ledger with reason codes and dates |
| 11. Duplicate and off-term | Promotions paid more than once, and payments outside the terms | Payments matched to promotion and customer deal numbers |
Write the definitions down before the first measurement, because every rate moves with the definition behind it. The tests for kinds 1 and 2 need a baseline the organization trusts; until it has one, report them as ranges. Kinds 3 to 6 need consumption, price and sell-through data. The tests for kinds 7 to 11 need only a promotion record that the deduction, the claim and the payment can be matched to, which is why step 4 is usually where the work starts.
Source: IMA360 analysis; Pauwels, Hanssens and Siddarth, Journal of Marketing Research 39, 2002
Section 3
Why do the industry figures disagree?
Industry studies appear in nearly every discussion of trade promotion. They are summarized here with what each one counts, because differences in definition explain most of the differences in result. None is used as a benchmark in this paper.
| Publisher and date | Figure as published | What it counts, and the method stated |
|---|---|---|
| Nielsen, as reported by FoodNavigator, 2014 | “almost three-quarters of promotions don't break even” | Share of promotions; more than one million UPCs and 39 million US promotional events |
| Nielsen, as reported by POPSOP, 2015 | 59 percent of trade promotions in seven countries do not break even; in the US, 29 percent make money | Share of promotions; 212 million events, 2012 to 2014 |
| NielsenIQ marketing material, 2024 to 2025 | “72% of trade spend doesn't break even” | Worded as a share of spend; the same 72 percent appears elsewhere as a share of US promotions; no method stated |
| EY, 2016 | Nearly 70 percent of promotions lose money; average return 95 percent against 100 percent break-even | More than 2,000 events, 25 product groups, more than 14 US retailers; return basis not stated |
| McKinsey, 2019 | About 20 percent of revenue invested in trade promotions; 59 percent of promotions lost money, 72 percent in the US | Cites a 2016 Nielsen study |
| Promotion Optimization Institute, 2023 | “between 11% and 27% of revenues”, historically | Membership survey of more than 130 manufacturers; no method for the range |
| Strategy&, 2017; Deloitte, undated; L.E.K., undated | About 20 percent of gross sales; 20 to 30 percent; 30 percent or more in some categories | A benchmarking study and practitioner experience; no sample sizes |
A count of promotions and a share of dollars answer different questions; a return on revenue and a return on profit can differ by more than the whole margin; and a share of revenue moves with its denominator and with what is counted as trade. A large academic study takes the retailer's side: across CVS's 2003 promotions, Ailawadi, Harlam, César and Trounce found that “approximately 45% of the gross lift is incremental for CVS”, where switching between brands in the store does not count, and that “more than half of CVS promotions are not profitable” for the retailer. A manufacturer's own measure, with its definitions written down, is the benchmark it can act on.
Source: FoodNavigator, October 9, 2014; POPSOP, July 10, 2015; NIQ for Trade Planning; EY, Can smarter pricing and promotion reduce the emphasis on discounting?, 2016; McKinsey, October 23, 2019; Promotion Optimization Institute release, January 30, 2023; Strategy&, March 8, 2017; Deloitte US; L.E.K. Executive Insights XVI(33); Ailawadi, Harlam, César and Trounce, Journal of Marketing Research 43(4), 2006. Industry figures are publisher claims that this paper has not verified independently.
Section 4
What is a trade promotion operating model?
An operating model is the combination of people, process, systems and data that turns a trade budget into promotions in the market and then into settled, measured results. Every manufacturer has one; the question is whether anyone designed it. In the Promotion Optimization Institute's 2025 report, “61% of respondents agree that their companies face difficulties” executing promotions as planned, and 40 percent said data cleansing, harmonizing and staging held them back.
People
- Sales and key account teams negotiate and run events
- Revenue growth management sets price, pack and promotion guardrails
- Finance owns the accrual, the true-up and the profit measure
- Deductions and credit teams validate and dispute
- Supply chain owns the delivery performance behind most fines
- Brokers and retail execution teams enter deals and check stores; captive and floor plan lenders serve dealer channels
Process
- Intent, design, approval, setup, execution, settlement, learning
- Handoffs defined by what each step must pass on
- Customer and statutory clocks owned by name
Systems
- ERP pricing conditions and credit memos
- Trade promotion management for plans, funds and claims
- Optimization models for baselines and scenarios
- Deduction tools for matching and disputes
- Customer portals for deduction backup, claims and repayment
Data
- Shipments, invoices and remittances
- Consumption: syndicated and retailer point-of-sale data
- Deal terms, performance conditions and proof
- Settled cost by event, customer and item
Much of the leakage in section 3 sits in the handoffs: between the terms the key account manager agreed and the ones keyed into the system, between the event cost in the plan and the one settled through deductions, between the baseline method used to approve the event and the one used to evaluate it. The seven-step workflow that follows is organized around those handoffs.
Source: Promotion Optimization Institute release, February 10, 2025 (survey of more than 130 manufacturers); IMA360 analysis
Section 5
What are the seven steps of trade promotion management and optimization?
| Step | The question it answers | When it is weak |
|---|---|---|
| 1. Trade investment strategy and intent | What is the trade budget for, by customer, channel and brand? | Trade spend is last year's plus a growth rate, and nobody can say what it was meant to buy |
| 2. Promotion design and planning | Which mechanism, depth, timing and customer, at what expected profit? | Events are copied from last year's calendar and approved on lift |
| 3. Governance and approval | Who may commit trade dollars, against which limits and guardrails? | Approval is whoever signs, and the plan changes after approval |
| 4. Setup and deal communication | Do the customer, the system and the ledger hold the same terms? | The deal sheet says one thing, the system another, the deduction a third |
| 5. Execution, compliance and accrual | Did the event run as agreed, and what is owed? | Proof is collected when a deduction arrives; the accrual is set from last year's rates |
| 6. Deductions, claims and settlement | Is each claim owed, paid once and on time? | Deductions are written off as they age, or fought one at a time |
| 7. Post-event measurement and learning | What did each event return, against what baseline, at the latest cost? | Results arrive after the next calendar is locked, measured on shipments and accruals |
Steps 1 to 3 decide effectiveness. Steps 4 to 6 protect cash. Step 7 is where the two meet, and it can only be as good as the record the other six leave behind.
Source: IMA360 analysis, November 2026
Section 6
How mature is your trade promotion operating model?
The levels apply to each step independently. Few organizations sit at one level across all seven, and the goal is to find the step holding the others back.
- Executing
- the work gets done by hand, by whoever is available.
- Planning
- standards and a budget exist and are applied unevenly.
- Optimizing
- rules, data and controls make the step repeatable.
- Strategizing
- the step is run against customer, category and channel objectives.
- Maximizing
- the step adapts continuously from evidence.
| Step | 1. Executing | 2. Planning | 3. Optimizing | 4. Strategizing | 5. Maximizing |
|---|---|---|---|---|---|
| 1. Intent | Trade spend is last year's plus a rate. Nobody can say what it was meant to buy. | A budget exists by customer and brand, set from history in the annual plan. | The budget is split by purpose, with a stated objective per customer, and tracked against commitments. | Funds are allocated by expected return by customer, category and channel, and moved in-year as settled results arrive. | Next year's envelope is set from measured event results, and spend that bought nothing is redeployed. |
| 2. Design | Events are copied from last year's calendar. Depth and mechanism are accepted as offered, with no modeled alternative. | Templates and guideline depths exist. Lift is forecast from the last comparable event. | Each event is modeled on a baseline method locked before approval, with decomposition, settled-cost assumptions and a break-even; mechanism and stack are chosen on purpose. | Calendars are optimized by customer within guardrails, against price and competitor plans; dealer stacks are priced at the unit. | Models recalibrate from settled results each cycle, and forecast error is measured and falling. |
| 3. Approval | Approval is whoever signs. Changes after approval go unrecorded. | Thresholds exist by value. Exceptions are granted by email. | Approval runs in a workflow with the event model attached, guardrails enforced and every later change re-approved. | Approvers see expected profit and the customer's performance history; non-standard terms are reviewed before commitment. | Delegated limits are recalibrated from realized returns, and fast-track approvals are audited. |
| 4. Setup | Terms live in deal sheets and email. The deduction team learns of a promotion when the deduction arrives. | Promotions are keyed by hand after agreement, sometimes after the event starts. | One promotion record with an identifier, dates, items, customers, rates and conditions drives customer communication, accrual and claim matching. | Customer terms and dealer stacks are configured once and applied the same way in planning, accrual and settlement. | New mechanisms are configured without development, so setup no longer limits design. |
Source: IMA360 analysis, November 2026
Section 6
Maturity, steps 5 to 7
The three steps below are where execution leakage is found and where effectiveness leakage is measured, if it is measured at all. The four above are where both are created.
| Step | 1. Executing | 2. Planning | 3. Optimizing | 4. Strategizing | 5. Maximizing |
|---|---|---|---|---|---|
| 5. Execution and accrual | Proof is gathered when a deduction is challenged. The accrual is a period-end estimate from last year's rates. | Accruals estimate expected participation and performance at program level and are reconciled to the ledger each period. Compliance is checked on a sample. | Proof is collected during the event against recorded conditions; accruals post by promotion from the record; the method is reviewed each period; the rollforward splits current and prior-period provisions. | Estimates by customer and program draw on point-of-sale, claims and field inventory data; compliance data scores customers and dealers. | Estimates adapt from settlement data; revisions are small, explained in the close where they arise, and show no persistent pattern of releases or top-ups. |
| 6. Settlement | Deductions are written off as they age, or fought one at a time. Claims are paid as received. | Deductions are researched by hand against deal sheets. Disputes live in spreadsheets. | Every deduction and claim is matched to a promotion record before it clears; open deductions are aged and reserved; invalid ones are disputed with evidence inside the window; duplicates are blocked. | Root causes feed setup and supply chain; customers are scored on deduction quality; fines are charged, in management reporting, to the function that caused them. | Clean claims settle without handling, valid disputes are rarely written off, and every window is met. |
| 7. Learning | Results are discussed without being measured. The next calendar is built before the last one is read. | Large events are compared with plan on shipments. | Every event above a spend threshold is measured against a baseline re-estimated by a locked method, on consumption data, at the latest cost, with one profit definition. | Results set next year's mechanisms, depths, customers and fees; leakage is reported by kind. | Incrementality is tested against control stores, ad zones or dealer regions where possible, and spend that bought nothing is redeployed within the year. |
Source: IMA360 analysis, November 2026
Section 7 · Step 1 of 7
Trade investment strategy and intent
Step 1 decides what the trade budget is for, by customer, channel and brand, before any event is planned, so that the key account manager across the table from a buyer, or the dealer manager setting a program, works inside boundaries set on purpose. When this step is weak, trade spend is whatever last year's commitments add up to, and the first time anyone sees the total is at close.
1. Purpose by customer
- Price, display, listing, volume or selling effort
- Where trade spend is the wrong instrument
2. Customer plan and terms
- Joint business plan and the retailer's scorecard
- Everyday terms kept apart from event funds
3. Trade envelope
- Stated by customer, channel and brand
- Fixed, variable and pay-for-performance funds
4. Guardrails
- Everyday low price or high-low, chosen per item
- Maximum depth, frequency and stack; minimum profit per event
5. Program taxonomy
- Mapped to the accounts finance uses
- Applied the same way in every system
6. Success measures
- Incremental profit per dollar by customer
- Signals that intent needs revisiting
| 1 to 2 | 2 to 3 | 3 to 4 | 4 to 5 | |
|---|---|---|---|---|
| People | Name an owner for trade investment. Give finance a seat in setting it. | Bring sales, revenue growth management and finance together before the annual plan. Agree who arbitrates. | Put trade investment into brand and customer planning. Assign owners by customer and channel. | Treat trade investment as a standing capability. Review intent with commercial leadership in-year. |
| Process | Write down what each customer's trade spend is meant to buy. Total current commitments. | Set guardrails for depth, frequency and stack before the negotiating season. Separate everyday terms from event funds. | Allocate the envelope by expected return. Tie unconditional allowances to measurable performance where the customer agrees. | Reallocate in-year as settled results arrive. Feed step 7 results into next year's envelope. |
| Systems | Hold the budget outside email and slides. Keep one current list of programs. | Publish one taxonomy of purposes and mechanisms. Map it to the general ledger. | Version the envelope so downstream teams see changes. Tie funds to approved events. | Track committed, accrued and settled spend against the envelope continuously. |
| Data | Know trade spend by customer and mechanism for the last four quarters. | Measure net price by customer after all trade. Fix item and customer hierarchies first. | Hold the envelope, commitments and settlements in one data set. Report variance monthly. | Model the envelope against volume and price scenarios. Measure the result each period. |
Source: IMA360 analysis, November 2026
Section 7 · Step 2 of 7
Promotion design and planning
Step 2 turns intent into events: a mechanism, a depth, a window, a customer and an expected result, or, in a dealer channel, a program stack priced at the unit. This is where effectiveness leakage is cheapest to prevent. When this step is weak, the calendar is last year's with new dates, the mechanism is whatever the customer prefers, and lift is forecast from the last event that looked similar.
1. Baseline
- Method locked before approval
- From consumption data where available
2. Decomposition
- Competitors, later weeks and own items
- Category expansion, if any
3. Mechanism choice
- Off-invoice, bill-back, scan-back or lump sum
- Forward-buy and pass-through exposure
4. Event profit
- Discount on base volume shown as a cost
- Break-even lift stated before approval
5. Stack check
- Every program that can apply to the unit
- Dealer cash, customer cash, financing and floor plan
6. Calendar fit
- Frequency, depth and competing events
- Price changes and supply limits
| 1 to 2 | 2 to 3 | 3 to 4 | 4 to 5 | |
|---|---|---|---|---|
| People | Give one analyst the event model. Have finance review it before commitment. | Pair key account managers with an analyst on every major event. | Give planners the settled results of comparable events. Train them on decomposition and stacks. | Hold planners accountable for incremental profit. Adjust their latitude on results. |
| Process | List the mechanisms in use. Write a one-page template for each. | Model every event above a threshold on a locked baseline method before commitment. | State break-even lift and stack cost for every event. Choose the mechanism on forward-buy and pass-through exposure. | Optimize each customer's calendar within guardrails. Simulate competitor and price scenarios. |
| Systems | Keep the event model in one file with one owner. | Build a library of standard mechanisms the system can compute. | Plan in the same system that will accrue and settle the event. Reject designs it cannot compute. | Recalibrate models from settled results each cycle. Compare proposals with comparable events automatically. |
| Data | Pull shipments and promotion history by customer and item for two years. | Buy or build consumption data for the largest customers. Hold deal terms as structured data. | Join consumption, shipments, deal terms and settled cost at item, customer and week. | Measure forecast error by customer and mechanism. Act on it. |
Source: IMA360 analysis, November 2026
Section 7 · Step 3 of 7
Governance and approval
Step 3 decides who may commit trade dollars, against which limits, and what happens when an event does not fit. It is also where the questions on section 2's regulatory pages belong: whether payments for promotional services are available on proportionally equal terms to competing customers, in the words of sections 2(d) and 2(e), and whether an advertised price, financing or lease offer carries the disclosures the rules describe. When this step is weak, approval is whoever signs, the quarter-end deal that funds a stretch target goes through on a phone call, and the plan the system holds is not the plan the customer is running.
1. Approval thresholds
- By spend, depth, mechanism and expected profit
- Escalation for non-standard terms
2. Guardrail enforcement
- Depth, frequency and stack checked automatically
- Exceptions visible to finance
3. Program review
- Availability of promotional payments to competing customers
- Wording of consumer, financing and lease offers
4. Change control
- Every change re-approved
- Late changes visible by customer
5. Audit trail
- Decision, approver, date, rationale
- Link from approval to the promotion record
6. Period-end controls
- Rules that hold at quarter end
- Visibility of late-period deals and dealer programs
| 1 to 2 | 2 to 3 | 3 to 4 | 4 to 5 | |
|---|---|---|---|---|
| People | Name who may approve trade spend. Publish the list. | Separate negotiator from approver. Route non-standard terms to review before commitment. | Give approvers the event model and the customer's settled history. Train them on guardrails. | Hold approvers accountable for realized returns. Adjust their limits on results. |
| Process | Set thresholds by spend. Require a written exception above them. | Add program review to the approval path. Set an expiry on every exception. | Clear standard events on a rule check. Reserve human review for exceptions and new mechanisms. | Recalibrate delegated authority each year. Audit a sample of fast-track approvals. |
| Systems | Keep approvals in one place. Retire the email thread. | Route approvals in a workflow with the model attached. Record the decision with the terms. | Block setup of any event without approval. Re-route changes automatically. | Flag events that breach guardrails or stack limits automatically. Attach prior results. |
| Data | Record approver, date and rationale for every event. | Report approvals by approver, mechanism and spend each quarter. | Compare approved with settled cost and profit by mechanism and customer. | Recalibrate limits from realized returns by mechanism and customer. |
Source: IMA360 analysis, November 2026
Section 7 · Step 4 of 7
Setup and deal communication
Step 4 encodes what was approved so that the customer, the trade promotion system, the accrual and the deduction team hold the same terms for the same items on the same dates. This is where execution leakage is manufactured: a deal keyed after the event started, a performance condition never recorded, a promotion number the customer's deduction does not carry. When this step is weak, the agreement is right and every downstream record disagrees with it.
1. Promotion record
- Items, customers, dates, rates and mechanism
- The customer's own deal or agreement number recorded against it
2. Performance conditions
- Price, display, feature or advertising
- The proof that will be accepted
3. Customer communication
- Deal sheet or EDI 889 issued from the record
- Confirmation that the customer loaded it
4. Fund linkage
- Each event tied to a fund and an approval
- Commitments visible as they are made
5. Effective dating
- Ship, order or scan dates, stated
- Treatment of changes and extensions
6. Dealer program setup
- Eligibility by model, region and dealer
- Financing, floor plan and bonus terms with dates
| 1 to 2 | 2 to 3 | 3 to 4 | 4 to 5 | |
|---|---|---|---|---|
| People | Name an owner for promotion setup. Make the role explicit. | Separate setup from negotiation. Train setup staff on mechanisms and conditions. | Give setup staff the deduction results that setup errors caused. | Measure setup on the deduction outcomes it causes. |
| Process | Key every promotion before it starts. Track the ones that are late. | Issue customer communication from the record. Confirm receipt before the event. | Apply amendments once, with effective dates, to every record that uses them. | Close the loop: every deduction mismatch traces to a setup cause and a fix. |
| Systems | Move terms out of deal sheets into a system. | Give every promotion an identifier, and record the customer's deal number against it. | Drive accrual, claim validation and settlement from the same record. | Configure new mechanisms and dealer programs without development. |
| Data | List active promotions with dates, items and customers. | Hold items, customers and hierarchies as maintained master data. | Measure setup latency and mismatch rate by customer. | Reconcile terms with customers' records before every major event. |
Source: IMA360 analysis, November 2026
Section 7 · Step 5 of 7
Execution, compliance and accrual
Step 5 runs the event and records what it owes. The customer cuts the price, builds the display or runs the ad, or the dealer sells the unit and files the claim; the manufacturer collects the proof and books the accrual. When this step is weak, proof is gathered only when a deduction is challenged, the accrual is set at period end from last year's rates, and the variance between estimate and settlement is discovered at close.
1. Proof of performance
- Price, display, feature and advertising
- Matched to the recorded conditions
2. Compliance tracking
- By event and customer
- Fed into the customer's scorecard
3. Accrual by promotion
- Participation, performance and redemption
- Field inventory for dealer programs
4. Accrual method and controls
- The method that better predicts the amount (ASC 606-10-32-8), used consistently
- Cutoff for events that span a period end
5. Rollforward
- Opening, provisions, payments and deductions applied, revisions, closing
- The prior-period line supports the ASC 606 prior-period revenue disclosure
6. Supply performance
- On-time and in-full by customer
- Shortfalls visible before deductions arrive
| 1 to 2 | 2 to 3 | 3 to 4 | 4 to 5 | |
|---|---|---|---|---|
| People | Name who collects proof for each customer. Make finance the accrual owner. | Give field and key account teams the conditions they must verify. | Share compliance results with sales leadership and with customers. | Review compliance and deductions with customers in the joint business plan. |
| Process | Collect proof for the largest events. Book accruals promotion by promotion. | Collect proof during every event above a threshold. Review the accrual method each period against stated thresholds. | Split the rollforward into current-period and prior-period provisions. Explain every revision. | Score customers and dealers on compliance. Use the score in step 2. |
| Systems | Store proof where the deduction team can find it. | Accrue from the promotion record automatically. Test access, change and report controls on the promotion system. | Load point-of-sale, claims and field inventory data to refine the estimate each period. | Adapt the estimate from settlement data continuously. |
| Data | Know open accruals by customer and promotion. | Track accrual against settlement by promotion. | Measure prior-period revisions against the opening accrual, with their sign. | Keep revisions small, and explain them in the close where they arise. |
Source: IMA360 analysis; ASC 606-10-32-8 and 606-10-50-12A
Section 7 · Step 6 of 7
Deductions, claims and settlement
Step 6 decides whether each deduction and claim is owed, pays it once, and disputes the rest with evidence inside the customer's window. It is where execution leakage is found and where the quality of step 4 shows. When this step is weak, deductions are written off as they age because nobody can match them to a promotion, the same event is paid by deduction and by check, and delivery fines disappear into the trade budget.
1. Matching
- By customer deal number, items, dates and amount
- Fines and compliance charges coded apart
2. Validation
- Performance and proof checked
- Rates, items and dates within the record
3. Dispute
- Customer windows tracked, including in customer portals
- Documentation attached to every dispute
4. Duplicate control
- Deductions and checks matched
- Claims compared across channels
5. Deduction accounting
- Open deductions aged and reserved; valid ones cleared against the accrual
- Write-offs above set limits approved
6. Dealer claim validation
- Claims checked against program terms
- Co-op paid against proof
| 1 to 2 | 2 to 3 | 3 to 4 | 4 to 5 | |
|---|---|---|---|---|
| People | Name a deduction owner for each major customer. | Keep promotion setup apart from claim approval, and validation apart from collection. | Share root causes with setup, sales and supply chain monthly. | Hold the functions that cause invalid deductions accountable for them. |
| Process | Research the largest deductions first. Track write-offs by reason. | Match every deduction to a promotion before clearing. Dispute invalid ones inside the window. | Block duplicates before payment. Code fines apart from promotion deductions. | Settle clean claims without handling. Report write-offs of valid disputes as defects. |
| Systems | Pull remittances and deduction backup into one place. | Match deductions to promotion records automatically, with an exception queue. | Validate claims against encoded terms and proof before payment. | Track every customer window and statutory deadline on one calendar. |
| Data | Know open deductions by customer, reason and age. | Measure match rate, dispute rate and recovery by customer. | Measure write-offs without dispute and disputes filed after the window. Reconcile the deduction reserve with the accrual each period. | Score customers on deduction quality. Use the score in step 1. |
Source: IMA360 analysis, November 2026
Section 7 · Step 7 of 7
Post-event measurement and learning
Step 7 measures what each event returned, against a baseline re-estimated by the method locked at approval, at the latest cost, and feeds the answer into next year's intent. It is the only step that can find effectiveness leakage. When this step is weak, results are read from shipments, compared with plan for the largest events only, and arrive after the next calendar is locked.
1. Event measurement
- Against a baseline re-estimated by the locked method
- Settled where final, otherwise accrued plus open claims
2. Decomposition
- Dips and cannibalization measured
- Category expansion, if any
3. One profit definition
- Incremental profit over trade spend
- Written down before measurement
4. Leakage report
- By customer, mechanism and program
- Trend against prior periods
5. Variance explanation
- Change in estimate kept apart from leakage
- Explained in the close where it arises
6. Feedback
- Mechanisms, depths, customers and fees
- Programs retired or redesigned
| 1 to 2 | 2 to 3 | 3 to 4 | 4 to 5 | |
|---|---|---|---|---|
| People | Name an owner for post-event analysis. Give finance a seat in it. | Put analysts and key account managers in the same review. | Make results part of every customer plan and approval. | Hold the commercial organization to measured returns. |
| Process | Review the largest events against plan. | Measure every event above a threshold against the locked-method baseline, at the latest cost. | Report leakage by kind. Feed results into the next calendar before it is locked. | Test incrementality against control stores or regions where the customer allows. Redeploy spend that bought nothing in-year. |
| Systems | Keep results in one place with the event they describe. | Join plan, consumption and latest cost by event automatically. | Feed results back to the design model and the approval workflow. | Recalibrate baselines and models from results each cycle. |
| Data | Keep plan and actual volume for the largest events. | Add consumption data and latest cost by event. | Measure dips, cannibalization and pass-through by event. | Measure model error on out-of-sample weeks. Track it over time. |
Source: IMA360 analysis, November 2026
Section 8
Where should a trade promotion transformation begin?
The system comes later. Start at the step where the last unpleasant surprise originated, which is rarely the step that noticed it. The table maps symptoms most teams recognize to the step that created them and a first move that needs no new system.
| Symptom | Step to fix | First move |
|---|---|---|
| Deduction write-offs rising, mostly promotion-coded and unmatched | 4. Setup | Record the customer's own deal or agreement number against each promotion at setup, issue terms from the record before the event, and measure the match rate weekly. |
| Events that look good on shipments and poor on consumption | 2. Design, then 7. Learning | Lock the baseline method at approval. Measure the weeks around each event and the manufacturer's own items, at the latest cost. |
| A display or feature was paid for and nobody can show it ran | 5. Execution | Define the proof each condition requires at setup, and collect it during the event. |
| Dealer program costs above plan | 2. Design | Price the full stack per unit before approval. Model financing cost under take-rate and term-mix scenarios and stair-step cost under pull-forward, and test claims against retail sale records. |
| Prior-period accrual adjustments that surprise the close | 5. Accrual | Accrue by promotion from the record. Split current- and prior-period provisions in the rollforward and explain each revision. |
| Fines and compliance charges absorbed by the trade budget | 6. Settlement | Code fines apart from promotion deductions, dispute invalid ones, and report valid ones, in management reporting, against the function that caused them. |
| A quarter-end deal nobody can find an approval for | 3. Governance | Block setup without approval and re-approve every change. |
Source: IMA360 analysis, November 2026
Section 9
What role do systems play in trade promotion management and optimization?
Systems decide the ceiling. A manufacturer whose promotion terms live in deal sheets and whose accruals live in a spreadsheet can run the promotions it ran last year and read the results a quarter late. A manufacturer whose terms are computable records, shared by planning, accrual, settlement and measurement, can design, price and prove each event, and optimize the calendar on settled results. The logic lives in one or more of these places, and most manufacturers run a mix.
| Where the logic lives | What it does well | Where it breaks |
|---|---|---|
| ERP pricing conditions and credit memos | Off-invoice discounts; posting credits and accruals | Performance conditions, bill-backs and scan-backs, deduction matching, dealer stacks, anything that changes faster than configuration |
| Spreadsheets and deal sheets | Whatever the analyst who built them understands | Versions, effective dating, audit trail, staff turnover, and the number of events a large calendar produces |
| Separate planning, optimization and deduction tools | Each task inside its own tool | The handoffs: a plan that does not match the setup, a deduction that cannot find its promotion, an optimization model fed planned cost instead of latest cost |
| One governed record of promotion terms beside the ERP | Planning, approval, setup, accrual, claim validation, settlement and measurement on the same terms, with lineage to the ledger | Where master data, configuration or interfaces are poorly controlled, customer data is unavailable, or teams work around it |
What to insist on
One record
Latest cost by event
Configuration without development
Source: IMA360 analysis, November 2026
Frequently asked
Frequently asked questions about trade promotion
What is trade promotion management?
What is trade promotion optimization?
What is the difference between off-invoice, bill-back and scan-back?
How is trade promotion accounted for?
What is a promotion baseline?
How do you calculate promotion ROI?
What is forward buying?
What are deductions, and how should invalid ones be handled?
What does the Robinson-Patman Act say about promotional allowances?
How are dealer incentives different from retail trade promotion?
Source: IMA360 analysis, November 2026
Why IMA360
Where the platform fits
IMA360 is a single governed platform for pricing, rebates, chargebacks, contracts, promotions and incentives. As described on ima360.com, it supports off-invoice discounts, bill-backs, scan-backs, lump-sum payments, volume incentives, display allowances and co-op programs, with configurable rules, budgets and tracking for each; plans budgets, calendars and scenarios and automates accruals; matches deductions against planned promotions, contracts and shipment data with configurable rules, routing unmatched or disputed items to resolution workflows; and reports incremental volume, revenue lift, margin impact and cost per incremental unit for each promotion.
For dealer channels it designs manufacturer, distributor, dealer and consumer programs on one canvas with the stack-up cost per unit; models floor plan interest subsidies and 0 percent financing at the unit level; validates dealer claims for co-op, consumer rebate redemption and SPIFFs against terms, eligibility and proof of performance; and measures the baseline before an event and the lift after it in one engine. This paper argues that a manufacturer can only optimize the promotions it can measure, prove and govern. The platform is built on that principle: one governed record of each promotion's terms, read by planning, accrual, deduction matching, settlement and post-event measurement.
IMA360's customers are listed at ima360.com. IMA360 reports processing more than $400 billion of revenue and 600 million transactions a year.
Source: IMA360 product pages and company figures as published on ima360.com, read November 18, 2026. These are company claims and have not been independently verified.
Sources
Sources and method
The mechanics in this paper reflect implementation experience. Figures, dates and quotations were retrieved from the sources listed below in November 2026 and checked against the version named in each source line, with the exceptions noted in the box at the end of this section. Quotations are verbatim as read; several academic quotations come from published abstracts or the authors' manuscripts, and the van Heerde, Gupta and Wittink quotations from the working paper version of the article. Ratios and the worked example are the authors' arithmetic; the company figures on the Why IMA360 page are IMA360's own.
- 1General Mills (FY2026, filed July 1, 2026), Hershey (FY2025, filed February 17, 2026), Campbell's (FY2026, filed September 24, 2026), Clorox (FY2026, Exhibit 99.1, filed August 7, 2026), Utz Brands (FY2025) and Simply Good Foods (FY2026): Forms 10-K, sec.gov
Supports: Accounting policies, program types, estimation inputs, accrual balances, estimate accuracy and sensitivity
- 2Deere (FY2025, filed December 18, 2025), Toro (FY2025, filed December 17, 2025), General Motors (FY2025, filed January 27, 2026), Ford (FY2025, filed February 11, 2026) and Generac (FY2025): Forms 10-K
Supports: Dealer incentives, accrual at the sale to the dealer, pool funds, interest-free periods, floor plan financing costs, financing incentive timing, subvention, co-op accounting
- 3Walmart (FY2026, filed March 13, 2026), Kroger (FY2025, filed March 31, 2026), Target (FY2025, filed March 11, 2026) and US Foods (FY2024)
Supports: Retailer and distributor accounting for vendor consideration, compliance charges, deductions from invoices
- 4Jones Soda Form 10-K/A (filed May 22, 2026); Bridgford Foods FY2025 Form 10-K (filed January 28, 2026); SEC staff correspondence with WD-40 Company (March 31, 2020) and Collier Creek Holdings (August 6, 2020)
Supports: Critical audit matters on promotional allowances; staff questions on estimation method and reserve true-ups
- 5SEC Litigation Release 16866 (2001); SEC press releases 2016-25, 2019-196, 2020-36, 2021-78 and 2023-210; SEC order in Elanco Animal Health (November 12, 2024)
Supports: Enforcement involving trade marketing expenses, rebates and channel incentives
- 6Deloitte, Roadmap: Revenue Recognition (2025), section 6.6; FASB Transition Resource Group memos 14 and 19 (January 2015); PwC Viewpoint, Revenue guide, section 4.6; KPMG, Revenue for consumer products companies (2016)
Supports: ASC 606 consideration payable to a customer, timing, slotting, consumer rebates through retailers, variable consideration and the constraint; ASC 250-10-20 on errors
- 715 U.S.C. 13 (govinfo.gov; uscode.house.gov); 16 CFR Part 240 and 79 FR 58245 (September 29, 2014); FTC complaints, press releases and Commission statements on PepsiCo (January 17 and May 22, 2025) and Southern Glazer's (December 12, 2024 and October 2, 2026)
Supports: Robinson-Patman sections 2(a), 2(d) and 2(e); the Fred Meyer Guides; recent FTC activity
- 816 CFR 233.1 and 251.1; 12 CFR 1026.16, 1026.24 and 1013.7 and official interpretations (consumerfinance.gov, accessed November 18, 2026); Tex. Bus. & Com. Code ch. 605; RCW ch. 19.315; 27 CFR 6.21, 6.96, 6.152 and 10.21; TTB press release, September 10, 2026
Supports: Former price and free offers, financing and lease advertising, mail-in rebate deadlines, alcohol trade practices
- 9X12 transaction set purpose statements (x12.org); Kroger supplier deduction claim training document (2022); Talk Business & Politics (February 23, 2017; September 3, 2020; October 17, 2023); Supply Chain Dive (March 8, 2019); SupplyChainBrain (February 22, 2024); Credit Research Foundation 2012 Customer Deduction Survey
Supports: Documents in the promotion and deduction flow, customer rules and windows, delivery fines, deduction rates, retailer data access
- 10Gupta (Journal of Marketing Research 1988); Bell, Chiang and Padmanabhan (Marketing Science 1999); van Heerde, Gupta and Wittink (Journal of Marketing Research 2003, working paper version read); van Heerde, Leeflang and Wittink (Journal of Marketing Research 2000; Marketing Science 2004); Ailawadi, Gedenk, Lutzky and Neslin (Journal of Marketing Research 2007); Nijs, Dekimpe, Steenkamp and Hanssens (Marketing Science 2001); Pauwels, Hanssens and Siddarth (Journal of Marketing Research 2002); Srinivasan, Pauwels, Hanssens and Dekimpe (Management Science 2004)
Supports: Where promotion lift comes from, post-promotion dips, long-term effects
- 11Drèze and Bell (Marketing Science 2003); Besanko, Dubé and Gupta (Marketing Science 2005); Ailawadi and Harlam (Marketing Science 2009); Nijs, Misra, Anderson, Hansen and Krishnamurthi (Marketing Science 2010); Armstrong, Gerstner and Hess (NEC-63 proceedings, 1994); Ailawadi, Harlam, César and Trounce (Journal of Marketing Research 2006)
Supports: Trade deal mechanisms, forward buying, pass-through, retailer promotion profitability
- 12Busse, Silva-Risso and Zettelmeyer (American Economic Review 2006); Corrado, Dunn and Otoo (Federal Reserve FEDS 2006-09)
Supports: Pass-through of customer and dealer incentives; use of cash rebates and interest subvention
- 13NielsenIQ base sales definition and Retail Measurement Services page; Circana Liquid Data Go CPG Dictionary (2026); Abraham and Lodish (Marketing Science 1987); 84.51° Stratum launch release
Supports: Baseline and incremental definitions; data for measurement
- 14Promotion Optimization Institute releases (January 30, 2023; February 10, 2025); Nielsen figures as reported by FoodNavigator (2014) and POPSOP (2015); NIQ for Trade Planning; EY (2016); McKinsey (2019); Strategy& (2017); Deloitte; L.E.K.
Supports: Industry figures, shown with attribution and the measure each counts; publisher claims that this paper has not verified independently
Retrieved from SEC EDGAR, ftc.gov, govinfo.gov, uscode.house.gov, federalregister.gov, consumerfinance.gov, ttb.gov, x12.org, journal and author pages, and the other sources named above, November 2026
Download the PDF
The original 45-page white paper, to save, print or share with your team.
Get Your Free White Paper
Fill in your details below to download instantly.

