Kyle
Washington DC startup Review Signal launched last week after 19 months in the making. Founder Kevin Ohashi has developed a product review platform that can be more trusted than traditional platforms like Yelp or Google reviews. Review Signal analyzes data points across social media to gauge the pulse of what people think about a particular product or service.
While there are hundreds of verticals that Review Signal could be applied to Ohashi has started with web hosting reviews as his first full on demonstration of how the system works.
Review Signal mines the data across social networks, and then separates the “mentions” for a specific product or service into good or bad. As Ohashi tells us in the interview below, if there are 100 tweets about a product with 50 being positive and 50 being negative, the raw review score would naturally be 50%. Now people aren’t forced to read long, sometimes biased long-form reviews.
One of the biggest problems we’ve seen with traditional reviews stems from the reason the average person would write a review anyway. Aside from career reviewers and those types that have to review absolutely everything, the bulk of the rest of reviews on review sites come when a customer is raving about a company, product or service after having a great experience or after they’ve had a bad experience.
You don’t get a lot of “average” everyday folks that take the time to write a review about a run of the mill or average experience.
For example say you went to a decent (not exceptional, just decent) restaurant and had the best chocolate cake ever. You’re much more likely to tweet “had this awesome chocolate cake” with a photo, rather than write a 5 paragraph review on the restaurant itself or, heck, even the chocolate cake.
One of the most fascinating parts about this story is how it came about in the first place. Ohashi actually based Review Signal off the idea he used for his masters thesis.
Check out our interview with Ohashi below. He does a really great job of explaining how the system actually works.
Review Signal is a review site powered by conversations in social media. It turns conversations into reviews by filtering and analyzing what people are saying about a company publicly. The first industry we decided to cover is web hosting. Our web hosting reviews launched with over 100,000 reviews about hosting companies. Our goal is to create a transparent and accurate picture of what consumers think about a company and give that information away for free.
The easiest way to explain Review Signal is our 45 second video: http://www.youtube.com/watch?v=VPpwbZWLwJQ&feature=player_embedded
Basically, we collect opinions people share about companies on social media (Twitter only right now). We run a lot of filtering to remove spam and make sure each opinion is relevant. Then we analyze whether people were happy or not with the company. The overall rating is the average experience of users. If 50 people were happy with a company and 50 people were not, the company would have a 50% rating. We have a detailed How It Works page explaining how everything is calculated.
I, Kevin Ohashi, am the founder of Review Signal. My academic background is I attended The University of Vermont and got a BA in Economics with a minor in Computer Science. I also received two MSc’s from Lund University in Sweden in Entrepreneurship and International Marketing & Brand Management. This is one of many startups/projects/businesses I’ve run since I started my career in high school. I’ve been involved in the domain name business, I started an outsourcing company in Indonesia, a gift recommendation engine (GiftLizard.com), and many other projects/businesses.
Washington, DC
Washington DC has a small startup community. There are a lot of technical people in the area but there is a lot of competing demand from all the jobs the government creates. It also doesn’t help that it’s a very transient city that many people use a stepping stone to other things. I don’t think DC has quite revived itself from its AOL days. There are a couple big names like LivingSocial, but nobody as big as AOL was in the 1990’s-early 2000’s. However, I think there are a certain type of startups that belong in the city and would benefit most from being here. The popularity and growth of DC startups is definitely trending upwards, maybe we will see another AOL growing out of the area soon.
The technology and idea mainly came out of my master’s thesis. The topic was predicting box office sales using tweets. It used a lot of data mining, sentiment analysis and statistics to try and model word-of-mouth and use that to estimate how much money a movie would make. The obvious question then becomes why didn’t you do movie reviews? The reasoning is quite simple, movie reviews is a very crowded and established market where people don’t seem to be that unhappy. So I decided to review something else, something scummy and awful. Web hosting reviews were a perfect fit. They are something I’ve never been satisfied with and it’s an industry I knew a lot about.
I originally wanted to call it Mention. I thought I had a deal for the domain name Mention.com but it fell through. So I went back to the drawing board and generated a lot of names. I was heavily involved in the domain name business for many years and built a lot of tools to generate brand names. They came in very handy. I thought Mention was really clever because the reviews were what people mentioned about a company. Since Mention was unavailable I wanted something that reflected what the company does. It filters down social media and finds the signal embedded within every day conversations. The signal of course being our opinions of companies. Sticking review in front turned into Review Signal. It’s reviews based on the signals we, the consumers, send in every day conversations on social media.
Web hosting reviews are horrible. Nobody trusts them, nor should they.More generally though, Review Signal is about rebuilding our faith in online reviews. It uses data mining techniques to gather more opinions than anyone else, which is helpful for the statistical significance of our results. We also make it very transparent that these reviews came from people’s public social identities. Every single review can be verified at the source it was posted. Nobody ever comes to Review Signal and ‘writes a review,’ we listen to the natural conversations happening all day, every day. So our information is always up to date and tracking changes over time. We want your trust, but you can also verify everything.
Trying to strike a balance between monetization and trust. Nearly every review site monetizes with affiliate links (among other things). Review Signal isn’t an exception. How do we convince users that we’re different and actually giving honest reviews? We’re going with an honesty and transparency approach. Whether that works in the long run, time will tell.
I honestly wasn’t sure I could turn my idea into a reality. I didn’t know if I had the knowledge and ability to create the systems required to take raw conversation and turn it into a high quality review site. It took 19 months of reading, learning, testing and implementing ideas to get Review Signal live. It started out as an ‘I wonder if…’ situation into a ‘this might actually work’ and the rest is history.
Improving our process and expanding into other areas beyond web hosting. If all goes according to plan, Domain Registrars will be the next niche covered. I also have to balance development and marketing. Review Signal is a bootstrapped company, so it needs to generate some revenue to fund further development.
