Display ads can help challenger brands drive long-term sales growth

By: Daniela Yu, Sr Analytics and Media Manager

Challenger brands in Amazon’s store may benefit from using display ads to help improve sales.

Story highlights:

Challenger brands – defined here as small- to medium-sized businesses that sell in Amazon’s store – are often curious about the effectiveness of display and sponsored ads. Without proper insights, challenger brands may focus only on sponsored ads and miss opportunities to grow ad-attributed sales using display ads, or display ads and sponsored ads together.

We conducted a comparative study between challenger brands and leading brands to see whether display or sponsored ads performed better. We found that challenger brands may benefit from combining display and sponsored ads, as there is a synergistic “bonus” sales impact beyond each of their individual benefits. Our analysis also found that the synergy effect is more common for challenger brands than for established brands.

Our study includes five challenger brands and nine established brands. These 14 brands cover six categories from multiple verticals: Nutrition and Wellness, Baby Care, Coffee, Cold Beverages, Mass Skin Care and Mass Cosmetics. For each brand, we observed ad spend compared to sales benefit over 77 weeks, from January 2018 to July 2019, for a total of 1,078 observations. Also, for each brand, we used dynamic models to calculate the long-term sales benefits of advertising, including the sales elasticities of sponsored ads and display separately, and the interaction of sponsored ads and display to reveal media collaborations.

For more information, see the Methodology section at the end of this article.

1. The long-term sales effect of display is higher for challenger brands than for established brands

The study shows that leading brands see the highest sales benefit in the week they increase display spending, while challenger brands see the highest benefit in the second and third weeks after increasing display spending. But, overall, challenger brands received a higher benefit over leading brands during the period observed.

The visual below shows the aggregated sales elasticities over time.

Sales elasticity of display over time by leading versus challenger brands

Leading brands reach the highest sales benefit in the week they increase display spending (week one), while challenger brands reach the highest benefit in weeks two and three. Overall, challenger brands received a higher benefit over leading brands during the seven weeks observed.

Y = Sales elasticity
X = Weeks (Week one = week of increased impressions)

Blue circle key: leading brands

Leading brands

Light blue circle key: Challenger brands

Challenger brands

2. Combining display and sponsored ads can yield positive results for challenger brands

The analysis shows that combining either display and sponsored ads or display and retail (“retail” includes merch placements, product recommendations, and any other impressions that a user may view during their Amazon shopping experience) saw the highest yields in terms of ad-attributed sales for six of the 14 brands (three for display and sponsored ads, and three for display). Among these six brands, four are challenger brands and two are leading brands. This breaks down to two out of nine (22%) of leading brands and four out of five (80%) of challenger brands showing synergy between display and sponsored ads.

In the visual below we quantify the synergy for each of the six brands. For each brand, we calculate the ad-attributed sales the brand generated during weeks when both its impression types were above their respective average impression levels versus when both impression types were below their average levels (baseline sales). We compared that to weeks when the brand had above average impressions from only one of the two, instead of both.

Our analysis shows that there is an 87 percentage point ad-attributed sales impact when both sponsored ads and display are used together. In weeks when Brand 1 had only above average display impressions, ad-attributed sales increased by 55% over baseline. In weeks when Brand 1 had only above average sponsored ads impressions, ad-attributed sales increased by 47%. However, in weeks when Brand 1 had above average impressions in both sponsored ads and display, ad-attributed sales didn’t increase by 102% (55% + 47%), but by 189%. Thus, synergy is responsible for an 87 percentage point (189 ppt – 102 ppt) ad-attributed sales benefit above the sum of sponsored ads and display’s independent sales effects. On average, the synergy effect is larger among challenger brands than that of leading brands.

Display generated synergy across brands and media types

Brand 1: Challenger brand, display and search

Brand 1: Challenger brand, display and search: 87%

Brand 2: Challenger brand, display and retail

Brand 2: Challenger brand, display and retail: 75%

Brand 3: Challenger brand, display and search

Brand 3: Challenger brand, display and search: 14%

Brand 4: Challenger brand, display and retail

Brand 4: Challenger brand, display and retail: 4%

Brand 5: Leading brand, search and display

Brand 5: Leading brand, search and display: 32%

Brand 6: Leading brand, search and retail

Brand 6: Leading brand, search and retail: 4%

Conclusion and recommendations

Amazon display ads can have a long-term sales impact that extends beyond the first one or two weeks when display investment is increased. This total long-term sales impact can be higher for display than for sponsored ads, especially for challenger brands.

For challenger brands, display’s impact on sales peaks later and lasts longer, reaching a long-term sales impact similar to that of sponsored ads. Therefore, we conclude by recommending that challenger brands use display together with sponsored ads to help gain the synergistic “bonus” impact of sales beyond the sum of their separate benefits.

Methodology

Sampling

To balance the generalisability and efficiency of the research outcome, we first sampled three verticals. We then sampled two categories from each of the three verticals using strict criteria: we sampled verticals and categories that were different in seasonality, average selling price, retail readiness and business size. We expect that the diversity of vertical and category choices reasonably represents businesses of different natures, which enables us to generalize the findings to other categories with more confidence.

Dynamic model methodology

We estimated the advertising effects on sales using Vector Autoregressive (VAR) models. This is an approach that is widely used in economics, finance and marketing for its comprehensive and efficient estimation of dynamic effects.