Decide.com now assigns a score to more than 22,000 products that the service tracks.

Decide.com already tells consumers when they should buy cameras, phones, refrigerators and other gadgets and appliances, analyzing price drops and increases. Now, the Seattle startup and brainchild of University of Washington computer scientist Oren Etzioni and others is getting into the recommendation business, assigning a score of 1 to 100 on more than 22,000 products that it tracks.

The new Decide Score feature, launching today, puts products in four categories from “we love it” (assigned to nine percent of products) to “don’t buy it” (assigned to 12 percent of products).

With the launch of Decide Score, the young upstart appears to be taking direct aim at Consumer Reports. The 76-year-old magazine analyzes and ranks thousands of different products, from automobiles to gas grills to baby strollers. Asked about competition, Decide.com’s Shauna Causey agreed that other entities do offer in-depth consumer information and recommedations.

Lawn mowers are one of the new product categories from Decide.com

“We’re taking a different approach by using big data and the wisdom of the users to determine if products are the best or worst,” Causey said. In order to compile the score, Decide.com is analyzing 7,000 expert reviews and two million user reviews from sources such as Amazon.com, Best Buy and others.

Decide.com pulls that review information from third-party sources, using a “crawling” technology similar to what search engines employ.

“It entails crawling publically available reviews, and presenting users with pointers to read the full reviews as appropriate,” said Causey. “The heart of our algorithms is in normalizing the reviews and removing biases due to different sources, changes in review scores over time, and attempts to “game” the review system.
By aggregating multiple sources and statically analyzing the reviews and their scores we are able to deliver very high-quality scores that have been benchmarked carefully against multiple sources.”

In addition to the new Decide Score, the company also is expanding its offering of products to include sports equipment (tennis rackets, tents, treadmills, skis, etc.); lawn & garden items (hedgers, leaf blowers, lawn mowers, etc.); and tools & hardware (drills, saws, generators, etc.)

Launched in June 2011, the company says its price predictions have been accurate 77 percent of the time.

Decide.com raised a $6 million venture round last year from Maveron, Madrona,  early Google investor and former Amazon.com executive Ram Shriram; former Expedia CEO Erik Blachford and former Farecast CEO Hugh Crean. Former Farecast executive Mike Fridgen, who helped sell the online travel search company to Microsoft for $115 million in 2008, serves as CEO of Decide.com

Previously on GeekWireDecide.com unveils daily deal site, promises cash if prices later fall on gadgets

Comments

  • Oren Etzioni

    A huge amount of blood, sweat, and passion has gone into this major expansion of our product. I couldn’t be more proud of the dedicated and talented team at Decide. Look for rapid iteration and expansion in the next 8 weeks!

  • jeff

    Isn’t it easier to just go to amazon and read the reviews?

    • MikeFridgen

      Hi Jeff – thanks for asking. Only Decide applies technology across sources – expert and user generated – to provide a completely unbiased, data driven recommendation. We’ve found that individual sources, such as Amazon, do not tell the whole story. Check out this post by our VP Engineering Kate Matsudaira on how we do it
      http://cacm.acm.org/blogs/blog-cacm/153962-data-mining-product-reviews/fulltext

  • http://twitter.com/adamloving Adam Loving

    When there are a lot of choices, reading all the reviews can be really difficult. I’m looking forward to trying this for buying a projector for my office.

  • David Shim

    Congrats to the team at Decide! They shaking up a space that has had limited innovation in the last decade.

  • Doos

    I had a bad experience where they recommended a buy. When the price dropped, I requested the money they promised. They came back after 10 days and said that the recommendation was ‘wait’. Their website was changed to reflect that. They have awful customer service (10 days to reply back, no phone number to call) and an awful business ethic (not being honest, defrauding customers). I cannot return the product since I am past the 14 day return period with the store I bought. So, I lost the money in two ways – one the membership fee I have paid so far, second money I would have saved by otherwise returning and purchasing from the other cheaper store. There is no point in getting their membership. Their price gaurantee is for 2weeks. You can return or get price matched in most of the stores. You can get the same and more detailed reviews on CNET.com or other similar websites. After being a member for 2 months, I not only felt cheated but realized that this website is useless.

Job Listings on GeekWork

Find more jobs on GeekWork. Employers, post a job here.