The key role of full-funnel advertising for Grocery advertisers in 2020

By: Kapil Goyal, Sr. Analytics and Media Manager

This analysis compares actual sales in 2020 to sales predicted by using an econometric model. It also examines the relationship between advertising investment and sales performance and identifies the changes in advertising investments made by top- vs. bottom-performing Grocery brands.

Story highlights:

Since the first quarter of 2020, grocery advertisers on Amazon have experienced an increased demand for their products. In fact, Inmar polled over 300 US Grocery consumers and found that 78.7% of them purchased groceries online, and 51.4% reported buying groceries from Amazon, an increase of +32% since 2019.

With this increase in demand, brands may be curious to know how they can optimise their advertising budgets. To answer these questions, we utilised Inmar’s report to contrast:

  • The actual sales vs. predicted sales in 2020 by using an econometric model.
  • The relationship between advertising investment and sales performance.
  • The changes in advertising investments made by top-performing advertisers vs. bottom-performing advertisers on Amazon.

Results show that top-performing brands (total sales) in the Grocery category invested more than bottom-performing brands overall, and used a full-funnel advertising strategy on Amazon.

For more on how we collected our data, see the Methodology section at the end of this article.

1. Grocery sales were higher than forecasted in aggregate as well as for many verticals in 2020

Insight

The analysis produced predicted sales for 2020, after controlling for the organic growth of advertisers and Amazon, seasonality, changes in Amazon Fresh/Whole Foods delivery and pickup policy, and advertising spend (via advertising impressions). The difference between the actual sales and the predicted sales is the unexpected change in the sales for 2020 grocery. Using predicted sales and actual sales, we were able to index the actual sales against the average predicted sales. The resulting indexed actual sales allowed us to examine the effect of shopping behaviour changes on sales and determine if actual sales were above or below expected sales.

Illustration of actual sales and predicted baseline sales

Y-axis: Sales (0-180), X-axis: Months (March – December)

Graph of sales lift over 12 weeks. Brands gained 100 (indexed) sales during week one (week of campaign activation); in week two sales are at 70; in the following weeks sales continued to be higher than baseline (indexed as 0).
Dark blue: Actual sales

Actual sales

Light blue Predicted sales

Predicted sales

2. Growth of full-funnel impressions differed sharply between the top- and bottom-performing brands, especially for upper-funnel impressions

Insights

Throughout this analysis, we define the top-performing brands to be those in the top 25th percentile of absolute sales growth and the bottom-performing brands to be those in the bottom 25th percentile. Worth noting:

  • The aggregate sales for top-performing brands were 1.3X the aggregate sales of the bottom-performing brands in 2019. Then, in 2020, top-performing brands grew to 3.5X in 2020 vs. 2019, while the bottom-performing brands decreased to 0.9X during the same period (Figure 2a).
  • Figure 2b demonstrates that Amazon upper-funnel impressions for the top-performing brands grew to 3X in 2020 vs. 2019, while upper-funnel impressions of the bottom-performing brands stayed flat during the same period.
  • Figure 2c shows that the lower-funnel impressions of the top-performing brands grew to 2X in 2020 vs. while the lower-funnel impressions of the bottom-performing brands decreased to 0.9X during the same period.

Amazon Grocery sales

Figure 2a. Sales growth of grocery advertisers

Amazon Grocery sales: 2019 base

2019 base

Amazon Grocery sales: Bottom brands 2020

Bottom brands 2020

Amazon Grocery sales: Top brands 2020

Top brands 2020

Amazon upper funnel

Figure 2b. Sponsored Brands impression growth

Amazon upper funnel: 2019 base

2019 base

Amazon upper funnel: Bottom brands 2020

Bottom brands 2020

Amazon upper funnel: Top brands 2020

Top brands 2020

Amazon lower funnel

Figure 2c. Sponsored Products impression growth

Amazon lower funnel: 2019 base

2019 base

Amazon lower funnel: Bottom brands 2020

Bottom brands 2020

Amazon lower funnel: Top brands 2020

Top brands 2020

Conclusion and recommendations

Overall, Grocery sales were higher than predicted through the end of 2020. We modelled sales using a predictive modelling approach and analysed sales growth vs. predicted sales for every Grocery vertical. Separately, we analysed two groups of brands, top-performing brands (top 25th percentile) and bottom-performing brands (lowest 25th percentile), based on the growth of their sales in 2020 vs. 2019.

The sales of the top-performing brands increased to 3.5X in 2020 vs. 2019, whereas the sales of the bottom-performing brands declined to 0.9X during the same period. Results show that the top-performing brands invested 3X more in Amazon upper-funnel impressions and 2X more in lower-funnel impressions in 2020 vs. 2019. The bottom-performing brands were flat in their upper-funnel impressions and declined to 0.9X in their lower-funnel during the same period.

Recommendations

  • Increase Amazon Ads upper-funnel investment to help shoppers find your brand.
  • Use Amazon full-funnel advertising to help grow your sales.

Methodology

This analysis includes 2,500 advertiser brands from 15 Grocery verticals. There are 94 weekly observations for each brand between Mar 2019 and Dec 2020. The first 52 observations make up the baseline period while the remaining 43 form the study period. We aggregated data at the vertical level (e.g. produce, lollies and chocolate, etc.) and built vertical-wise predictive models to forecast sales in 2020. We then compare this forecast to actual sales. Separately, we also analyse the year-over-year increase in advertising impressions for top- vs. bottom-performing brands.