Advertiser: QSR Stores Sdn Bhd Brand: KFC Creative Agency: - Credits: -
In the first half of 2017 KFC experienced a concerning sales low despite having the highest top-of-mind recall and highest media spend. Clearly there was a disconnect, we tasked data to pursue insights that would stem the slide, and make the second half of 2017 profitable.
Insight, Strategy and the Idea
We sifted through multiple datasets such as holidays, weather, traffic, pricing, media mix (including datasets not conventionally associated with impacting sales) to uncover 1. Malaysians are 74X more likely to stay indoors during bad weather 2. Online video buys nudge sales during bad weather: In good weather, online video contributed to a mere 6% of sales while TV contributed a whopping 78% of sales. However during bad weather days, the numbers flipped with online video accounting for 64% of sales & TV to 13%. In H1 our media mix had failed to account for changing audience behaviour during bad weather. Introducing the Agile Media Predictor, an engine that learned weather data and recommended media mix against predicted sales.
1. We set up a weather data bank that collected day by day temp, rainfall info etc. Based on historical data, the bank also had thresholds for the temperature/rainfall that could impair sales 2. The weather bank API was fed into the Econometric model. When data hit thresholds flagging a bad weather day – the model adapted and sent alerts to planners to pause TV buys & trigger the DSP to initiate video buys. The paused TV spots would be used up by other segments in the KFC portfolio. The process was sound, but only time would tell if it was sound enough to reclaim sales.
Results and Effectiveness
H2 2017 came with 6 in 10 days being bad weather days. We achieved • Revenue growth of 8% • Servings increase of 6% • And a media ROI unseen in 4 years