CPG companies spend 15–25% of gross revenue on trade promotions — discounts, in-store displays, retailer incentives, BOGO deals, shelf placement fees, co-op advertising. For giants like Coca-Cola or Lakme (HUL), this represents billions. Yet studies consistently show 50–70% of trade promotions don't break even.
The $4B TPM software market — Anaplan, o9, Accenture's tools — has tried to solve this for two decades. They've failed. Here's why.
A company like Coca-Cola operates across 200+ countries, with thousands of SKUs, millions of retail touchpoints, and hundreds of promotion types running simultaneously. Evaluating one promotion requires data from ERP systems (SAP, Oracle) for shipment data, retailer POS systems (Nielsen, IRI) for sell-through, TPM tools for planned spend, field sales apps for compliance, distributor portals for secondary sales, and finance systems for deductions.
⚠️ THE CRUSHING REALITY
For Coca-Cola, a single national promotion with Walmart involves data sitting across Walmart's Retail Link, Coke's SAP, Nielsen syndicated data, and the local bottler's ERP — none of which talk to each other natively. By the time you consolidate, cleanse, and reconcile, the promotion is over. You're always flying blind or looking in the rearview mirror.
For Lakme in India, the problem compounds because kiranas (small independent shops) — forming 80%+ of Indian retail — have zero digitised POS data. You simply cannot know if a promotion ran, at what price, or whether it drove incremental volume at the last mile.
To measure a promotion's true incremental lift, you need to know what would have sold without the promotion. This "baseline" is deceptively hard to construct.
Pantry Loading
Consumers buy 3 months of Coca-Cola during a BOGO deal, then disappear. Sales spike during the promotion, crater afterward. The post-event dip is almost never attributed back to the promotion's cost.
Cannibalization
A promotion on Lakme's lipstick might steal sales from its own kajal. Internal cannibalization makes ROI look better than it is — and is almost never measured.
Halo & Spillover
A Coke Classic promotion lifts Diet Coke too. How do you attribute that? Most companies don't even try.
Seasonality Layering
Coca-Cola in India runs promotions during IPL and summer and Diwali — often overlapping. Disentangling seasonality from promotional lift is a genuine statistical nightmare.
Most companies end up using overstated baselines because field sales teams and brand managers have incentives to show promotions "worked."
Promotional Compliance Gap
Retailers commit to running a feature display. In reality, 40–60% of planned displays never execute as agreed. Companies pay for compliance they don't get — and have almost no visibility into it.
Deduction Disputes
Retailers deduct claimed promotional amounts directly from invoices. Validating whether a claimed deduction matches an actual agreed promotion is a multi-billion dollar reconciliation problem globally. Coca-Cola reportedly has hundreds of millions in open deductions at any point.
Slotting Fees
Paying for shelf space is non-negotiable for launches. For Lakme launching a new serum in 5,000 modern trade outlets, this is pure sunk cost before a single unit is sold.
Customization Pressure
Walmart wants a different pack size. DMart wants a different mechanic. Costco wants a bundle. Each customization fragments demand forecasting and supply chain planning.
Channel Stuffing
When Coca-Cola offers distributor trade incentives, distributors buy aggressively then dump inventory at discounted prices, destroying brand price integrity.
Promotional Pass-Through Failure
Trade funds given to distributors are supposed to be passed to retailers as discounts. In reality, distributors often pocket the margin without passing the promotion through. The promotion reaches the end customer at full price.
Diversion / Grey Market
Products promoted at a low price in one region get bought in bulk and re-sold in premium markets. Persistent for Lakme across India's price-sensitive Tier 2/3 vs. premium urban markets.
Secondary Sales Invisibility
For Lakme via HUL's distribution network, the company sees primary sales in real time but secondary and tertiary sales data has a 2–4 week lag at best, and is often estimated.
Volume-Compensated Sales
A Coca-Cola Key Account Manager's bonus is tied to volume sold to Walmart. Running a deep promotion that sells 1M extra cases but destroys margin is good for the KAM's bonus, terrible for the P&L.
Finance Blind Spot
Trade spend is managed as a top-line deduction, not a marketing investment. A $50M TV campaign needs a business case. A $50M trade promotion fund gets allocated based on historical precedent.
Brand vs. Sales Conflict
Brand teams set price architecture and equity guardrails. Sales teams make real-time deals that violate those guardrails. Nobody has a clear authority line.
Local vs. Global Tension
Global HQ wants consistent pricing. The India team is under pressure to match a Pepsi promotion in Karnataka. Multiply this by 200 countries and you have permanent policy drift.
A 20% volume uplift predicted; 60% actual uplift occurs. Coca-Cola runs out of stock (lost sales, retailer penalties, brand damage). Or they over-produce and excess inventory must be liquidated. Promotional pack sizes require production line changeovers planned months in advance — but promotions are often confirmed by retailers only 4–6 weeks out. Last-minute demand spikes require expedited freight at premium rates, negating a large portion of the volume uplift's margin.
Too Late
By the time data is consolidated (3–6 months), the next year's promotional calendar is already being negotiated.
Politically Contentious
If a promotion shows negative ROI, the brand team blames execution. Sales blames the baseline model. Finance blames cost assumptions. Nobody owns the loss.
Learning Doesn't Transfer
Coca-Cola might learn that a free-standing insert doesn't work for Sprite in the US Northeast. That learning lives in a PowerPoint deck the next KAM never sees.
Mix Complexity
A "promotion" is a combination of price discount + display + feature ad + sampling + social media. Attributing ROI to any single element is nearly impossible.
| Cost Layer | Visibility |
|---|
| Face value of discount / price reduction | High — always counted |
| Slotting / display / feature fees | Medium — often budgeted |
| Unpassed-through distributor trade funds | Low — rarely measured |
| Supply chain premium costs (expediting, changeovers) | Low — rarely attributed to the promo |
| Compliance leakage (promotions that didn't execute) | Very Low — underreported |
| Post-promotion demand dip (pantry loading effect) | Very Low — never attributed |
| Internal cannibalization | Very Low — almost never measured |
| Deduction disputes / write-offs | Medium — Finance tracks, but siloed |
| Organizational time cost of planning & reconciliation | Near Zero — treated as overhead |
Why This Problem Persists Despite Billions Spent on Technology
Companies have spent hundreds of millions on TPM/TPO tools. Yet the problem persists because:
1. Technology can't fix misaligned incentives — if sales reps are still compensated on volume, they'll game any system.
2. Data quality upstream hasn't improved — AI models on bad data produce confident wrong answers.
3. Retailer systems remain proprietary — Walmart's Retail Link and Amazon's Vendor Central don't natively integrate with CPG systems.
4. Organizational change is harder than software — deploying a new tool doesn't change how a negotiation happens between a KAM and a Walmart buyer.
This is ultimately a problem at the intersection of data engineering, behavioral economics, organisational design, and channel strategy — which is exactly why it remains one of the most expensive unsolved problems in consumer goods.