Trade promotion forecasting for FMCG suppliers

What is trade promotion forecasting?

Trade promotion forecasting is the practice of predicting how a promotional activity will affect product demand before it runs. It estimates the volume uplift a promotion is likely to generate, the cannibalization risk across other SKUs in the category, and the net margin impact after accounting for promotional costs.

The distinction that matters is between baseline sales and incremental sales. Baseline is what you would sell without any promotion. Incremental is the additional volume the promotion generates. Trade promotion forecasting is the process of modeling that increment accurately, per SKU, per retailer, per promotion type, before any budget is committed.

This is different from general sales forecasting, which looks at overall demand. Trade promotion forecasting isolates the specific effect of a promotional activity from the underlying baseline, so teams can evaluate the promotion itself rather than the total volume picture.

‍

Why trade promotion forecasting is difficult in practice

The baseline problem

The most underestimated challenge in trade promotion forecasting is determining the baseline accurately. In categories with normal promotional pressure, baseline estimation is relatively straightforward: average the volume between promotions and you have a reasonable starting point.

But in categories where some FMCG suppliers sell 60% of their volume in promotion, the sales pattern changes fundamentally. Instead of gentle peaks and recoveries, you see sharp spikes and deep troughs, sometimes with sales close to zero between promotions. The baseline is no longer the average volume between promotions. It is a hidden variable masked by the constant promotional cycle.

Manually estimating the baseline in this environment is unreliable. Human analysis cannot separate true underlying demand from the noise of continuous promotional activity. AI-driven baseline modeling, trained on granular historical sell-out data per SKU and per retailer, is what makes accurate trade promotion forecasting possible in high-promotion categories.

Fragmented data across retail sources

Trade promotion forecasting requires data from multiple sources that rarely arrive in a compatible format. POS data from SIS or 7EVEN. Syndicated market data from Nielsen or Circana. Pricing and promotion history from IPV or Superscanner. Internal ex-factory and trade spend data. Each source uses different product hierarchies, different update cadences and different category structures. Before any forecasting model can run, all of that data needs to be harmonized into a single consistent view. Without that foundation, the model is working on incomplete or inconsistent inputs. As a result, it produces forecasts that cannot be trusted.

EAN changes breaking historical trend lines

Every packaging update, reformulation or relaunch introduces a new EAN code. In a manual data environment, that break in the product history means the forecasting model loses the historical trend line for that SKU. The promotion history of the old EAN does not automatically carry over to the new one. EAN changes that are not processed automatically create gaps in the data that systematically undermine forecast accuracy.

‍

What good trade promotion forecasting enables

Know the ROI before you commit

The most direct benefit of trade promotion forecasting is the ability to evaluate a promotion before budget is allocated. A supplier who can model the expected uplift, the cannibalization risk and the net margin impact per SKU before a retailer conversation is not guessing. They are negotiating from a position of data.

That changes the dynamic of the retailer meeting entirely. Instead of agreeing on a promotion and hoping it performs, the supplier arrives with a model showing projected performance under different scenarios. That is the conversation retailers want to have with strategic partners, not with vendors asking for shelf support.

How to forecast trade spend ROI for promotions

Forecasting trade spend ROI for a promotion requires four inputs: the historical baseline for that SKU at that retailer, the price elasticity of that SKU, the expected promotional mechanic (price discount, buy-one-get-one, display placement), and the planned duration and depth of the discount.

With those inputs, an AI model can estimate the incremental volume uplift, calculate the gross revenue impact, subtract the promotional cost, and project the net margin effect. That is the ROI of the promotion, calculated before it runs. The category manager reviews the model output, applies commercial judgment and context the model cannot see, and decides whether to proceed, adjust or reject the proposal.

Identify cannibalization before it costs margin

A promotion on one SKU rarely exists in isolation. A discount on the 400g format cannibalizes the 200g. A promotional mechanic on the premium range shifts volume from the standard range. These cannibalization effects are predictable if the historical data is clean and complete enough to model them.

Trade promotion forecasting at SKU level surfaces those effects before the promotion runs. A category team that can show a retailer that their proposed promotion will grow the category without cannibalizing adjacent SKUs is making a materially stronger case than one that presents volume projections in isolation.

Plan promotions that grow the category, not just the brand

The promotions that win retailer support are the ones grounded in category-level thinking. Trade promotion forecasting makes that possible by modeling the impact of a promotion on the full category, not just on the supplier's own range. That category-level perspective is exactly what Joint Category Development requires: shared data, shared analysis and proposals that create value for both parties. A promotion that the retailer's own data confirms will grow category revenue and reduce waste is a promotion that gets approved.

‍

How Captain makes trade promotion forecasting possible

Captain's promo simulator is built on exactly the data foundation that makes trade promotion forecasting reliable: all retail sources harmonized into a single consistent product model, EAN changes processed automatically so historical trend lines stay intact, and near real-time sell-out data from SIS, 7EVEN, Nielsen, Circana and internal sources.

The promo simulator calculates price elasticity per SKU from actual promotional history at each retailer. It models the expected uplift, cannibalization risk and net margin impact of a proposed promotion before it is agreed. Category managers can run multiple scenarios, compare projected outcomes and arrive at the retailer meeting with a data-grounded proposal rather than a volume estimate based on intuition.

Having a human in the loop remains essential throughout. The model surfaces the analysis. The category manager applies commercial judgment, account knowledge and context that the model cannot see, and makes the final call. Data quality is the foundation of all of this: a promo simulator working on fragmented or inconsistent data produces projections that undermine the credibility of the proposal rather than strengthening it.

‍

Results in practice

At Johma, trade promotion forecasting grounded in price elasticity per SKU produced a promotional plan that was adopted directly by Hoogvliet, Vomar and Plus. The retailer recognized the quality of the analysis and acted on it without further negotiation.

At Elho, automating the data foundation freed the category team from 60% of harmonization work, creating the capacity to apply proper forecasting to 25+ retail accounts rather than the 10 largest. That is what trade promotion forecasting looks like when the data foundation is in place. Read the full Elho case.

‍

Ready to forecast your next promotion before you commit?

Request a demo to see how Captain's promo simulator helps your team model promotion impact per SKU before budget is allocated, and come away with practical tips for your specific situation.