About DealScience

My name is Matt Dombrowski and I spent the past 10 years at Amazon and Spotify developing machine learning algorithms, most recently to help social media influencers find the best deals for their audiences. I saw firsthand how difficult it is, even for professional shoppers, to navigate the overwhelming number of "deals" online. E-commerce companies like Amazon employ hundreds of thousands of scientists and engineers that work to manipulate pricing and promotions with the specific goal of extracting the maximum amount of dollars from your wallet. In a given year, the average consumer is exposed to tens of thousands of experiments designed to drive more spending. Retailers frequently leverage legal but deceptive practices such as “price-increase and list-price synchronization” (PILPS), wherein an item's price is raised while simultaneously introducing a new higher list price to create the illusion of a discount. I created DealScience to do my small part to level the playing field, putting the power back in the hands of us consumers where it belongs. DealScience uses simple but powerful statistical methods, not black box AI algorithms, to sift through the noise and find the quantifiably best deals that are actually worth your time and money. So we can all stop doing this:

How It Works

We continuously scan the Amazon catalog to find best selling products that are currently at their lowest price ever on Amazon. Next, we evaluate products on a scale of 0 to 100 based on various factors including:

  1. Price score: Compares a product's current price on Amazon with its typical recent price across a number of retailers
  2. Brand score: Evaluates a product brand's name recognition, desirability, and quality from independent sources
  3. Trend score: Assesses a product's trendiness based on its rate of increase in Amazon sales as well as engagement on this site
  4. Product score: Assesses a product's quality based on star rating, review count, Amazon category sales rank, price volatility, image quality, return rate, and many other factors

We combine these inputs to compute an overall numeric deal score between 0 and 100 then rank deals accordingly, ensuring you only see the absolute best deals on Amazon right now. We make the component scores transparent to the user in the deal score badge in order to convey exactly why a given product is a great deal. Users can also sort, filter or search based on factors most important to them.

Mission and Community

The mission of DealScience is to empower consumers to make better, faster shopping decisions with data. We aim to save consumers valuable time, money, and stress while also reducing unnecessary purchases that harm the planet. A portion of our profits are contributed towards consumer advocacy organizations. We plan to contribute original academic research in the area of consumer protection and develop "Deal Science" into a rigorous scientific discipline. As a community-driven platform, we are genuinely thrilled to hear your feedback and suggestions in service of this mission. Please reach out to us on Instagram or email me directly at [email protected]. We have lots of other exciting community features in the works such as personalized deal alerts that are actually worth your attention, a Discord server, and a browser extension, so stay tuned by subscribing to our newsletter.