The shelf is a central concept in Gradient. A digital shelf, or just “shelf”, refers to a slice of the market on Amazon. A shelf can refer to a big department like Appliances, a narrow niche category like Collector’s Baseball Cards, or something more complex like all things school age kids need for going back to school.
All metrics provided by Gradient are relative to a specific shelf. What is the product rank in the shelf? What is the product share of the shelf?
One of the most basic metrics for a brand is the Shelf Presence, which measures the amount of interest consumers express for products in that shelf. Gradient uses direct observations and triangulation to estimate the overall interest in the shelf. The portion of interest occupied by a particular brand or product is the called the brand or product's share of shelf.
Together with the Shelf, the Brand is a central concept in Gradient. All your analytics activity is organized under one or more brands which you add to your account by Following.
You can follow one or more brands in your account. You can follow any brand you want, even a competitor, as long as it is present on Amazon. The brand's name is taken from the actual product listings on Amazon.
When listing their products on Amazon, brand managers can decide how to name their brands on Amazon. That means that a seller may list multiple brands or one single brand, depending on their strategy. For example a company Acme might list everything they sell under the name "Acme", or they may list some products under "Acme Toothbrush" and some under "Acme Toothpaste" as separate brands.
As a user of Gradient, you organize your measurements and insights by following brands in specific shelves. Different subscription levels provide access to more or less brands in your account. See subscription plans to learn more. The most common way to follow a shelf is by searching for a product or brand and visiting the score page, as shown below.
The Gradient Score™ is a quick way to find out how well a product or brand is performing on Amazon. The Gradient Score™ measures the relative strength of a product or brand in relation to the other participants in the shelf. The Gradient Score™ measures performance by blending observations of over 100 different elements for each product or brand in a category and comparing them to the population of other products within the same category. A Gradient Score™ of 500 represents the mean performance, 600 is 1 standard deviation above the mean, 700 is 2 standard deviations above, 400 is 1 standard deviation below the mean and 300 is 2 standard deviations below the mean.
Learn more about Gradient Score.
The Gradient Console™ is our self-serve web-based platform that brand leaders can use to follow brands and track their performance on Amazon. Creating an account in Gradient automatically gives you access to the Gradient Console.
The Gradient Audit is a summary of metrics for a brand in specific shelf between two specific times periods. Metrics are shown for the brand and its top products, but also the top competitors in the same shelf.
Audits are automatically generated for Gradient Plus customers for each brands and shelf you are following. As your brand audits are completed, you may view your audits in the Gradient Console.
Learn more about Audits.
Gradient Recommendations provide you with a set of optimizations to improve brands and products performance on Amazon. Recommendations are unique to every product and category. Thanks to our proprietary model, not only you are provided with specific actions but also you can have an idea of the impact each recommendation can have on sales.
Recommendations should be used together with the Gradient Score and the component scores (Presence, Response and Positioning scores). The latter will give you an idea of the area you should be focusing on. Depending on the type of recommendation, you will be able to see results in the short, medium or long term. Based on your business goals you will want to focus on one type or the other.
Learn more about Recommendations.
KPIs are the primary measure of success for a brand or product. These include Sales, Sales Rank, Gradient Score. Gradient reveals which factors have the highest impact on your KPIs.
Gradient Console users can enable daily alerts to get notified about significant changes in your brand's products on Amazon.
Learn more about Daily Alerts.
A digital shelf, or just Shelf, is a slice of the eCommerce market, specifically on Amazon. A shelf can refer to a big department like Appliances, a narrow niche category like Collector’s Baseball Cards, or something more complex like all things school age kids need for going back to school.
The two important measures of a shelf are (a) the amount of interest consumers express for products in the shelf and (b) the amount of sales revenue received by products in the shelf. Interest is measured in terms of the number of product searches and resulting listing impressions (both paid and organic) in the funnel. Whereas, sales revenue for the shelf is simply the sum of all sales in the category.
With respect to interest, Gradient uses direct observations and triangulation to estimate the overall interest in the shelf. We refer to the portion of interest occupied by a particular brand or product as Shelf Presence. For sales revenue, actual sales figures are known only to the individual sellers, Gradient may sometimes offer an estimate of a brand’s Revenue Opportunity using a brand’s Presence to suggest a ballpark potential the brand might aim to achieve.
The Gradient Score™ varies between 1 and 1000 but most scores are between 400 and 700.
Refers to the “amazon best sellers rank” for a product in the shelf category. The lower the number the better as it indicates higher sales with respect to other products. For a brand, the sales rank is the weighted average of sales ranks for all the brand’s products in the shelf category.
Percent of sponsored impressions for a given shelf that are occupied by a particular product or brand. Paid Presence uses the same calculation method as blended Presence except only includes paid impressions. Related to but not the same as Paid % of Presence.
Brand Presence, Product Positioning, Brand Presence, Customer Response are sub-components of the overall Gradient Score™, which roughly align to different brand management disciplines. Specifically, Presence encompasses promotional activities, Positioning encompasses merchandising activities, and Response encompasses product and customer service quality. Gradient incorporates numerous raw measures related to these disciplines for each component score. Your brand’s Gradient Score components are a reflection of your measures as compared to the other participants in the category. The overall score is a weighted blend of the three component scores.
Brand Presence encompasses promotional activities, which primarily includes shelf share, paid/organic composition, and seller mix. Your brand’s component scores are a reflection of these measures as compared to the other participants in the category.
Product Positioning encompasses merchandising activities, which includes price, inclusion of keywords in listings, image quality, and special listing features such as Amazon Prime. Your brand’s component scores are a reflection of these measures as compared to the other participants in the category.
Customer Response encompasses product and customer service quality, which includes ratings, reviews, and stock management. Your brand’s component scores are a reflection of these measures as compared to the other participants in the category.
When available, the estimated potential sales given the observed Shelf Presence for the brand or product in the selected category over the time frame. Note, this is not an actual sales forecast, and should be used for comparative purposes only.
The total number of customer reviews received by the product. At the brand level, this is a weighted average. Note that the number of reviews can sometimes go down! If this happens, it doesn’t necessarily require action. Amazon periodically monitors the reviews and may remove large numbers of reviews if they deem them to be suspicious.
The average product rating between 1 to 5 stars, with 5 stars being the best. Amazon calculates this star rating as a weighted average, to give more weight to more recent and more qualified reviewers.
Because the average star rating sometimes does not tell the whole story, these metrics provide insight into how the product's ratings spread out. For example, even if your average rating is steady, a rising % of 1-star ratings could require attention.
The estimated number of distinct sellers observed in the buy box winning position for each product in a given category and time frame. The lower the number of sellers the better as the brand has more control on its sales channel. At the brand level, this is a weighted average of each of the brand’s products.
Indicates the percentage of days that the product was observed to be in-stock. This indicator provides insight into the inventory management and potential lost sales of a brand or product. A precipitous drop could indicate attention is needed.
Brand or product average price within the shelf category. At the brand level, differences between time period average prices can sometimes be explained by changes in the product mix. For example, the format sold by each brand (e.g. single bottle pack versus 24 bottle pack), or the inclusion of small accessories in the category.
Measures the range of movement of a product's price within a given time frame. Includes max and min, starting and ending price.
This is the percent of the product or brand’s total impressions that were sponsored rather than organic. Related to but not the same as Sponsored Shelf Presence.
In general, the larger and more numerous the images for a product listing, the better. The image quality metric reports the total image surface area for all images in the product listing, as a sum in megapixels. Therefore, higher numbers here are better.
This indicator reports how well a brand’s products titles and descriptions contain the top keywords in the category. The higher the index the better.