Insights from Amazon shopping queries

By: Camilla Cristiano, Sr. Analytics and Media Manager, and Ashton Brown, Tech writer

Advertisers want to understand how customers shop on Amazon in order to optimize their media investments. This article analyzes shopping queries from the top five Personal Care Appliances categories in Germany to help advertisers understand how customers shop in Amazon’s store.

Advertisers are often curious at which point in the customer journey they should begin advertising. Similarly, they may be interested in understanding which Amazon Ads solutions may help them to reach their customers at the right time. In this article we attempt to answer those questions by providing insights on how customers shop in Amazon’s store. To conduct our research, we analyzed shopping behaviours associated with products in the top five Personal Care Appliances categories in Germany (men’s shavers, epilators, power toothbrushes, beard and hair trimmers, and hair removal devices).

To better understand if advertisers should engage customers early in their journey or just before they are ready to make a purchase, we set up our research using the following customer actions as parameters:

  1. Began the journey with a generic keyword query on Amazon.
  2. Conducted at least one additional query during the journey (within the same category).
  3. Completed a purchase in the same category within 30 days of the first generic keyword query.

Our analysis reveals that customers tend to purchase a product from one of the last two brands queried. While this behaviour is common during Q1-Q3, what stands out is that customers switch even more actively to branded queries at product purchase during Q4.

On Amazon, the typical shopping journey starts with generic keywords and ends with branded keywords

To better understand the customer product purchase journey, we began with a comparison of what happens when customers shopped on Amazon using generic keywords (e.g. smart speaker) vs. branded keywords (e.g. Amazon Echo). The aim of this portion of the study was to understand if customers are more likely to:

  • Use generic keyword queries and then purchase a product, or
  • Shop for a specific brand and then purchase a product.

We found that, in general, more customers typically begin shopping on Amazon with a generic keyword and then refine their queries to include to branded keywords before purchasing. The visual below shows the percentage of shoppers who refined their queries to include branded keywords prior to purchase.

Shopper journey length and period % of total shoppers

Q1-Q3

Same day purchase

Q1-Q3 Same day purchase, last shopping query: 51%
Q1-Q3 Same day purchase, last two shopping queries: 56%

Purchase within 30 days

Q1-Q3 Purchase within 30 days, last shopping query: 55%
Q1-Q3 Purchase within 30 days, last two shopping queries: 61%

Q2-Q4

Same day purchase

Q2-Q4 Same day purchase, last shopping query: 53%
Q2-Q4 Same day purchase, last two shopping queries: 57%

Purchase within 30 days

Q2-Q4 Purchase within 30 days, last shopping query: 58%
Q2-Q4 Purchase within 30 days, last two shopping queries: 63%
Orange circle: Last shopping query

Last shopping query

Light blue circle key: Last two shopping queries

Last two shopping queries

Conclusion and recommendations

Our analysis reveals that most customers begin with generic queries for products on Amazon and then refine them to branded queries before making a purchase. Our analysis also reveals that customers tend to switch to branded queries in Q4 more than any other time in the year, which may be related to the prevalence of branding campaigns and deals during key shopping events, or may be due to customers researching gift ideas during the holiday season.

To ensure that advertisements are reaching the right audience at the right time, we have come up with a few recommendations.

Recommendations

  • Consider increasing product and brand visibility early in the shopper journey. Increasing visibility early in the customer journey (e.g. when they are still shopping using generic terms) can help ensure that customers are considering an advertiser’s product prior to a purchase decision or brand refinement being made.
  • Use Sponsored Brands and Sponsored Products to reach shoppers early in their journey by bidding on generic keywords, to help raise brand awareness and increase the likelihood that they refine their shopping to your brand’s keywords.
  • Reach shoppers with Sponsored Display ads to further increase brand visibility while shoppers browse the category in Amazon’s store.

Methodology

Below is a breakdown of the steps taken to perform our analysis of the top five Personal Care Appliances categories in Germany (men’s shavers, epilators, power toothbrushes, beard and hair trimmers, and hair removal devices).

  • We isolated customers that made a generic shopping query (a shopping where neither a brand name or model is specified) as the first shopping query of their journey on Amazon.
  • Within that group, we looked at whether customers purchased a product (either on the same day or later) in 2019 and made at least one more shopping query before purchasing a product.
  • We refined our study to exclude branded searches that came after being exposed to an ad – to remove any advertising impact from Sponsored Brands, Sponsored Products, etc.
  • We then analyzed the last one to three queries made before buying the product, breaking these down between branded (a shopping event where either a brand name or model is specified, or where both brand name and model are specified) and generic (a shopping event where neither a brand name nor model are specified) queries. We decided to analyze the last and second to last query before product purchase as that is the stage where the branded/generics split reaches its saturation.
  • Outliers were removed based on the following parameters:
    • Number of queries during the customer journey. To remove the outliers, we applied the median rule methodology (Q2±2.3 inter-quartile range or IQR). As a result of doing this, only the following data ranges were included: (i) [0,30] days; (ii) [0,10] queries.
  • We split the data between purchases that take place in Q1-Q3 2019 and Q4 2019 This is to assess if the customer shopping journey looks different during Q4 than during the rest of the year.