CPGmatters: Kellogg Optimizes Promotions to Stimulate Brand Sales
Through a special arrangement, presented here for discussion is a summary of a current article from the monthly e-zine, CPGmatters.
Kellogg is moving ahead with a custom TPO (trade promotion optimization) system which combines vendor-supplied technology as well as advanced analytics from an in-house team. As Kellogg escalates from its TPM (trade promotion management) foundation to TPO, it aims to build customer-specific calendars, simulate and optimize events and analyze post-events.
During a recent webinar, Amjad Malik, vice president of business analytics at Kellogg, provided an overview of the company’s transition to TPO. He said "having planning and analysis in one place helps your business processes and helps make sure promoted prices and everyday prices are aligned and executed the way headquarters wants you to execute."
His experience shows three elements are key to successful TPO practices:
- integration with the TPM system;
- advanced analytics and;
He considers integration the most important to "avoid dual entry, minimize inputs, work within one system, and reduce errors and complexity. … The account executive or trade planner … should be able to look at what the event is going to do. That’s one of the biggest keys to making sure we have the right TPO solution."
For its part, Kellogg has embedded advanced analytics to conveniently show base consumption, lift simulations and effects of sales cannibalization. With TPO, Kellogg can also properly shape consumption to shipments for more accurate forecasts and simulate the revenue, lift and profits that will generate from different promoted price points.
Mr. Malik also noted the post-event analysis capability "allows us to decompose volume into buckets like trade, coupons and base volume to understand what’s coming from where and what’s cannibalized."
The analytics team has "built all models for all the channels and categories we play in," he added. Variables are many, including: accounts with syndicated data, shipment-based data or POS data; product and customer attributes; profitability measures; and promotion frequency and track records (products promoted a lot or sometimes, new items with brief histories, future items with no histories, and special packs such as bonus or multipacks).
"For TPM/TPO to fly, you can’t put in too many things. If too complicated, people won’t use the system, which means they won’t use the advanced analytics," said Mr. Malik. "Restrict your thoughts to optimization activities which really add value. These could be your starting points. Then bring new levels of optimization into your tool. … It really is a journey."
Training also helps users to understand the system’s value, said Mr. Malik: "We share best practices. We work with user groups. We have modeling teams work with field teams to make sure we can leverage and build models to their advantage. We use retailer successes as case studies."
Discussion Questions: What are the most important factors in achieving trade point optimization? What do you think of the potential of using advanced analytics and demand data to model trade promotion variables?