We just released a new interactive data tool, Amazon Performance Rank, that benchmarks brand performance on Amazon — allowing us to identify category trends. Here’s some insight into the data behind it.
Our initial release (the beta version) benchmarks brands in 34 fashion and beauty categories, offering a mix of high- and low-level classifications. We took a daily sample of the 100 bestselling products in each category during Q3 2017 and developed a 10-point scoring system that takes into account the rank of the bestselling product as well as the share of the list for which each brand accounts. This scoring system attempts to provide a balance between brands with many Best Sellers products and those that rely on a single product to drive the bulk of Amazon sales.
To illustrate how the underlying data looks, here’s a plot of data from Hanes (a brand with a large share of the Women’s Activewear Best Sellers list), Soffe (which has a small share of the list), and icyzone (which is trending upward). Each dot represents an individual product placement.
Next, we’ll visualize a few key metrics we like to use when working with Amazon Best Sellers data. Before we dive in, let’s define some terminology:
Rank: The average rank of the brand’s bestselling product over the time period
Visibility: The fraction of days the brand is present on the Best Sellers list
Real Estate: The share of the Best Sellers list for which the brand’s products account over the time period
The chart below visualizes these metrics for every brand that showed up in the Women’s Activewear Best Sellers list during Q3:
OK, so let’s get to the goods. Our interactive data tool presents the data both as a graph and as a ranking. The graph highlights the category leader and average, as well as the biggest winner and loser, while the ranking provides percent change and a variability score (standard deviation) to provide additional context.
As we continue to build out our Amazon Best Sellers data capabilities, we’ll be exploring how the dynamics of each category affect the ability of brands to gain and maintain status. How can we measure the volatility of a category? Can we identify the point at which increased investment is unlikely to supplant category leaders? How should we interpret broad product categories as opposed to narrow categories of directly competing products?
Keep an eye out for companion L2 research, where we will dive into the rankings and use additional data to unpack how brands are winning and losing.