How behavioural and demographic audiences perform on Amazon
By: Jessie Liu, Sr. Analytics and Media Manager
Many advertisers are curious about the performance of various audiences that are available to them in streaming media. You can use the learnings in this article to help inform your own streaming audience strategies.
First-party behavioural insights can help better inform media decision making
Athleisure, sports, athletics, fitness, yoga, cycling, running
Relying on demographics can provide a one-dimensional view of audiences.
Amazon Ads first-party insights give brands a more complete picture of their audiences.
Since the emergence of paid television advertising, advertisers have had little choice but to base their audience definition strategy mainly on demographics. In recent years, however, digital advertising has revolutionized the audiences that advertisers can reach online, including through streaming media or over-the-top (OTT). Nevertheless, many TV buyers, even those who have shifted television advertising spend to streaming media, default to demographic-first audience definition.
While demographic audience selection can help brands achieve their reach objectives, advertisers who solely use demographic audiences risk serving ads to customers that may find them irrelevant. They may also miss opportunities to reach interested customers outside their demographic audiences.
This paper uses Streaming TV ads campaign data from Q2 to Q4 2019 to analyze which audience selection method, behavioural or demographic, is more effective at driving consideration, as indicated by detail page views (DPV) on Amazon. Behavioral audience selection includes in-market and lifestyle segments, while demographic audience selection includes characteristics like gender, age and income.
1. On average, behavioural audiences were more likely to view product detail pages on Amazon than demographic audiences on Streaming TV ads
While the size of a demographic audience was roughly the same as that of a behavioural audience in this study, behavioural audiences for the brands we analyzed had a +44% higher median DPV rate vs. demographic audiences.
Most behavioural audiences showed higher median DPV vs. demographic audiences, ranging between +37% and +101% vs. demographic audiences. Across verticals, hardlines (which includes items such as hardware, housewares, automotive, sporting goods and toys) showed the highest percentage delta between behavioural audiences and demographic audiences, with behavioural audiences showing a median of +101% higher likelihood to view detail pages vs. demographic audiences.
The visual below displays a box plot distribution of detail page view rate (DPVR) for behavioural audiences, compared to demographic audiences across 28 campaigns in softlines (which includes apparel), hardlines and consumer packaged goods (CPG).
DVPR of behavioural vs. demographic audiences by vertical
x = detail page view rate (indexed)
2. The effectiveness of behavioural audiences depends on relevance.
Overall, data suggest that using appropriate behavioural audiences can help create more effective marketing campaigns for driving consideration on Amazon. However, insights also show that not all behavioural audiences work for every brand.
Behavioral audiences performed better when they were aligned with the product or brand. This means that while, on average, using behavioural audiences can help improve campaign performance, behavioural audiences that are not aligned with the product or brand can lead to poor campaign performance.
For example, early tech adopters had lower DPVRs vs. the median DPVR of demographic audiences for a softlines campaign. This could have been due to a misalignment between the fast fashion brand and the tech audience in this campaign. This is supported by the fact that this same behavioural audience was among the best-performing audiences for a hardlines campaign (+581% higher median DPVR compared to that of demographic audiences in hardlines).
Advertisers should use standard Amazon Ads campaign reports to regularly fine-tune their audience selection strategies to include the audiences that are most receptive to their campaign messages and exclude those that are not aligned with the product or brand – and so unlikely to respond to their brands’ messages.
In essence, we have found that demographic audiences on Streaming TV ads may not be the most efficient strategy to achieve advertisers’ consideration objectives. It suggests that incorporating behavioural audience selection may be a better strategy, either by defining audiences based solely on behavioural information, or by using behavioural information to complement demographically defined audiences.
As an example of this second application, in addition to using demographic audiences for reach, Streaming TV advertisers can also exclude lower-performing behavioural audiences from demographic audiences to reach the most responsive audiences. With the technologies and tools available today, can use behavioural insights to more effectively impact customer consideration towards their brands.
Amazon helps brands reach relevant audiences on billions of shopping and streaming signals. Incorporating these behavioural audiences, along with ongoing audience optimization, can help advertisers develop a holistic audience selection strategy and obtain better results from their Streaming TV ad campaigns for which driving consideration on Amazon there is the objective.
This study looked at detail page view rate (DPVR) for all audiences from the 28 Streaming TV ad campaigns included in this study. TV advertisers have traditionally used reach as the primary metric to measure the effectiveness of TV advertising campaigns.
While we recognize that reach is important, this analysis shows how advertisers can most efficiently reach the audience interested in their brands to drive consideration. We chose DPVR in this analysis as a proxy for customers’ intention to purchase, and also because DPVR provides advertisers with an actionable view into how their advertising tactics contribute to activity on Amazon.