How accessories and avid customers impact brand growth on Amazon

By: Bratiraj De, Senior Program Manager

Some electronics customers have bought recently, some have bought frequently, and some have a high dollar spend. Those that do all three can be called 'avid' customers. In this article we explore what that means for advertisers.

Story highlights:

With long sales cycles in Consumer Electronics (CE), advertisers often have difficulty obtaining meaningful insights into what drives repeat purchases. In this article we provide insights using CE order and spend data over 30 months (from 2017 through 2019) that suggest CE advertisers can also grow sales through accessories and through increasing their number of avid customers (customers in the top 40% in terms of dollar spend, number of orders, and purchase frequency).

Our analysis shows that avid customers disproportionately drive brand sales, and that these customers tend to purchase accessories (e.g., a TV soundbar, a phone case, or a Bluetooth keyboard) more than non-avid customers, especially after purchasing a standalone item. Furthermore, data show that ads that promote accessories can encourage more customers to become avid customers.

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

1. How avid and non-avid customers compare

Insight: Avid customers spend disproportionally more than non-avid customers

Our analysis shows that one avid customer contributes as much in sales as about seven non-avid customers. The visual below highlights the importance of avid customers. Even though, on average, only 5% of CE customers are avid customers, they contribute 25% of brand sales.

Index campaign performance (Baseline = bottom quartile)

Blue circle key: avid customers

Avid customers

Green circle key: non-avid customers

Non-avid customers

95% non-avid customers and 5% avid customers

% of customers

79% non-avid customers and 21% avid customers

% of orders

75% non-avid customers and 25% avid customers

% of sales

2. A higher investment in accessories was associated with a higher percentage of avid customers


The following visual is a scatterplot of how much each CE brand invested in advertising accessories for the two years covered in the analysis. The y-axis shows the percentage of avid customers, and the x-axis shows the ad investment in accessories as a percentage of total brand sales in Amazon’s store. For the brands studied, the scatterplot shows that a higher investment in accessories was associated with a higher percentage of avid customers.

Correlation between % avid customers and Accessories investment as % Retail sales for CE brands

Leading brands reach the highest sales benefit in the week they increase display spending (week 1), while challenger brands reach the highest benefit in weeks 2 and 3. Overall, challenger brands received a higher benefit over leading brands during the 7 weeks observed.

Y = % avid customers
X = accessories investment as % retail sales


To attract new avid customers, advertisers should consider:

  • Remarketing accessories to audiences who purchased a standalone product.
  • Marketing to audience who have purchased a standalone product but have not purchased complementary accessories through:
    • Amazon DSP remarketing helps you reach customers wherever they spend time after tentpole events like Prime Day, Black Friday, and Cyber Monday
    • Using sponsored ads and keywords to reach customers shopping for accessories after tentpole events
    • Upselling through remarketing and bundling standalone products with accessories
    • Creating Stores featuring your full accessory catalog to help drive greater awareness of your brand's portfolio


Our analysis shows that to grow brand sales, CE advertisers should not only focus on standalone items but also on accessories and increasing the number of avid customers. As seen in the article, avid customers tend to purchase accessories more often and spend more than non-avid customers. On Amazon, CE advertisers can use sponsored ads and Amazon DSP, and make use of tentpole events like Prime Day, Black Friday, and Cyber Monday to market accessories and attract new avid customers, which may lead to brand growth.


The analysis defines transactional customer loyalty towards a brand using the Recency, Frequency, and Monetary (RFM) framework. Analysis sees how recently and frequently customers visit Amazon’s store to purchase a brand in a given time period and also looks at how much a customer spends on a brand on Amazon in that time period. Transactional loyalty is then defined by the following three steps:

  • Step 1 – Assigning a percentile score between 1 to 5 (score of 1 denoting bottom 20 percentile) based on how frequently customers purchase from a brand on Amazon (Frequency Score), how recently they have visited to purchase (Recency Score) and how much they spend on brand (Monetary Score) in the analysis time period.
  • Step 2 – Combine Frequency and Recency scores to form visit segments – Customers who are Lapsing, have Tried and Left, were Once Frequent, are Newly Engaged, are Frequent, and are Highly Frequent (Exhibit 1).
  • Step 3 – Combine Visit segments with Monetary Score to create various Transactional Loyalty Segments – Premium, Growth, Potential, Uncommitted and Lapsed customers (in descending order of loyalty). (Exhibit 2) For example: According to exhibits 1 and 2, avid customers of a brand are in the top 20th percentile of spend, frequency of purchases, and how purchase recency.

This analysis also groups CE brands’ customers across three areas:

  • Dollar amount spent on the brand.
  • Number of orders made from the brand.
  • How recently a purchase was made from the brand.

For each area, we look at highest to lowest (i.e., highest spend to lowest spend, most orders to fewest orders, most recent to least recent), and break the group into percentiles. The top 20th percentile is group 5, the top 20%-40% range is group 4 and so on. The bottom 20% is group 1. For recency of purchase, we group customers based on time since last purchase. Customers whose most recent purchase was within six months are group 5, customers whose most recent purchase was between six months and a year are group 4, and so on. Customers who last purchased from the brand more than 24 months ago are group 1.

Avid customers are customers who are in the groups 4 or 5 for all three areas. Therefore, each brand has a theoretical maximum of 40% avid customers. This maximum is hit when there is complete overlap in the Venn diagram of the top 40% of customers in spend, orders, and recency. Each brand has a theoretical minimum of 0% avid customers when there is no overlap in the Venn diagram.