The challenge
Arnott's range was growing and competition was getting tougher. The business needed its promotions to work and to make money, but it had no reliable way to use its data for those decisions. As a result, promotion plans were not as good as they could be, revenue was at risk and operations were less efficient.
What was delivered
Our team was involved in delivering a trade promotion optimisation system that uses artificial intelligence. It works through large amounts of data and gives the team practical advice, so they can adjust promotion plans to meet revenue goals and decide quickly.
The program also covered the groundwork. It improved data quality, so the advice would be consistent and reliable, and it brought the separate processes around promotions into one way of working.
What it included
- AI that refines promotion plans against revenue goals
- Large amounts of data turned into practical advice
- Data quality work so the advice was consistent and reliable
- Faster decisions, with the important data available straight away
- Promotion planning and the work around it simplified
Results
Arnott's put the system into use, and its promotion plans improved. Data quality rose, so the advice was more consistent and reliable. Decisions came faster because the important data was available straight away, and planning promotions took less effort. The improved plans also lifted revenue. Figures are under NDA.
Why it matters
AI advice can only be trusted if the data and the process under it are sound. Much of this program went into data quality and a clean process, which is why the recommendations could be trusted. Any organisation that wants AI recommendations in its planning should expect the same order of work: fix the data first, then add the model.