Bringing promotional planning into one analytical cockpit

Challenge

A food manufacturer selling through the main modern retail chains ran a busy promotional calendar, but the information needed to manage it lived in different places. Invoiced sales came from the ERP, store sales arrived in a different format from each retailer and from a market panel, and campaign details were kept in a spreadsheet with free-text descriptions.

Campaigns were not recorded consistently. Mechanics, communication, discount depth and funding terms were described differently from one action to the next, so comparing promotions was slow and often inconclusive. There was also no agreed way to separate the extra volume generated by a promotion from normal sales, which made it hard to tell whether a campaign had paid for itself.

Planning tended to happen channel by channel and customer by customer. Without a consolidated view, the commercial team had limited visibility of how promotional spend was affecting volume, value, margin and market share overall.

Solution

We started by mapping the promotional process from planning to post-campaign review, working with the commercial, campaign management and IT teams. This showed where data was being lost or typed in twice, defined the indicators needed to evaluate a campaign, and led to a governance model with clear responsibilities, from the yearly promotional guidelines set by marketing to the day-to-day set-up of each action. Mock-ups of the tools were validated with users before development began.

Every campaign follows the same structure: mechanic (direct discount, loyalty card discount, multibuy, promotional pack, free product), communication type and funding model. Teams build reusable action templates by customer, products and discount depth, and then place them on a promotional calendar. Sell-in campaigns are pushed to the ERP in real time, so commercial conditions no longer have to be entered by hand.

It also holds sell-out from a market panel and from four retail chains, loaded through standard imports that validate each file and flag new products for review. Product transitions and observed shelf prices are kept in the same place.

For evaluation, each campaign is split into pre-, in- and post-campaign windows, defined separately for sell-in and sell-out. Actual sales are compared with a baseline that removes promotional peaks to estimate what would have sold anyway. From there, a Power BI dashboard calculates uplift in volume and value, promotional versus regular margin, investment and net profitability for each action. Further views cover sales evolution, promotional versus regular sales, and market position by segment and retailer, and the dashboard is sent to the team automatically every week.

Development ran in short cycles with user testing, and feedback from the commercial team shaped the final versions. After go-live, a support phase kept refining the tools, improving data quality and adjusting the dashboard as teams built it into their routine.

Results

The company now has one place to plan and review its promotions. Campaigns follow a common structure, so actions can be compared across customers, products and mechanics, and the history grows with every new campaign.

Performance is assessed with a consistent method that looks at shipments to retailers and sales to consumers, and at margin as well as volume. Commercial and management teams now discuss which actions are worth repeating based on the same set of numbers.

With structured promotional data in place, the company is ready for the next step: forecasting the effect of planned campaigns and testing alternative scenarios before committing budget.