Upper funnel ad effectiveness and seasonality in consumer durable goods

Abstract

Brands usually invest in a portfolio of ad products for brand consideration and conversion. To gauge performance, brands often use ad-attributed metrics to compare return on advertising spend (ROAS) across different ad channels. There are two shortcomings with this approach. First, it relies on a predetermined attribution window (eg 1 day, 14 days, 30 days).1 The resulting ROAS does not only favour lower funnel ads, such as paid search, but could change with different attribution models. Secondly, attributed performance metrics usually do not consider seasonality, and organic consumer demand changes over the course of the year, which could bias the estimates of the effectiveness of ads. This could result in mistakenly attributing high organic demand to campaign performance. These issues are addressed with a seasonal autoregressive integrated moving average with x/exogenous variables (SARIMAX) model, accounting for seasonality and brands' past performance. This analysis compares ad efficacy on total retail metrics, regardless of attribution methods. This method is applied to Amazon Ads for 15 brands in US consumer durables. While the lower-funnel ad product (Sponsored Products) sees a lot more current usage, higher-funnel ad products such as Sponsored Brands, demand side platform (DSP) display, and Streaming TV Video are all found to have higher efficacy. Brands should evaluate their ad-product performance at regular intervals to avoid under-utilising high-performing ad products. Future research is also encouraged to replicate this methodology in different verticals and locales for generalisability.