The DataHawk Sales Estimator estimates the units sold and the sales made (in your marketplace's currency) for a certain product during the last 30 days.

We also display a historized visualization of daily estimates for a certain product. These features are available provided you track the product on the DataHawk platform.

At DataHawk, we try to provide as close to precise as possible product sales estimates. There are two ways to go about it;

The Best Seller Rank of a product generally indicates a product’s sales performance. It is safe to assume that the better a product ranks, the more times it’s been sold. At DataHawk, we have a huge bank of this data collected over a long period. Our Data Science team combined this historical data with their knowledge and created a model that could link the best seller rank to the quantity sold.

Furthermore, they created sales estimating models for all the categories available on Amazon (any depth in the category tree accepted). This enables us to estimate any product we get the listings for (tracked products) and any product that ever came within the top 100 Bestsellers of any category on Amazon.

Therefore, we could compute an estimated sales volume per category (monthly and historically) to analyze category trends and compare categories with each other.
These estimates account for 89% of tracked products' estimates.

They account for 10% of all estimates. We use the review-based estimates whenever a product doesn't rank in a root category (product not covered by 1.).

We use our first sales estimation algorithm based on the number of reviews and their evolution for all the products that do not fall under the first model parameters (i.e., don't rank in a category). With all the different information we have on a product (reviews, ratings, price, category, marketplace, etc.), we attribute specific rates of sales per review per day and leverage this PDP data into a sales estimator algorithm.

Although this is not as precise as the BSR-based sales, it is still a good proxy.

Our algorithm uses other metrics and indicators, such as changes in the Best Sellers Rank and Keywords Position, but it's our secret sauce :) !
This allows us to give sales estimates for 99% of the products we track.

The missing one percent of estimates because we need a certain amount of historical data for a review-based estimate.

Based on the computations we ran since April 2021, the sales estimator shows a median error rate of around 35%. Meaning, that at least half of the estimations given by the software, as of now, have an accuracy of 65% and above. Our estimates are more accurate for products with a higher number of reviews.

We computed the error rate of 20 products in the Amazon-US marketplace across 7 main categories and compared the results with estimates from our competitors. The following is the result:

DataHawk’s estimator is undoubtedly the most stable sales estimator (see average error differences).

Simply put, even our worst estimates (3rd quartile of error rates) have a 56% error rate which is almost twice better than Jungle Scout and four times better than Helium 10.

DataHawk’s estimator doesn’t allow estimates to be way off. This chart shows that you can trust the vast majority of our estimates.
When looking at our best estimations (1st quartile and median figures), we can confidently say that in addition to being stable, DataHawk’s estimator has one of the best accuracies, especially with top-selling products.