Interpreting social messages for clues of trader sentiment to predict directions in the market and individual stocks has become a growing niche of Fintech innovation. Examples of using social sentiment include analyzing spikes of messages to spot breaking news before it is revealed to the greater public, as well as revealing where investor and trader money appear to be headed. Among new offerings, startups have been creating solutions for both retail investors as expensive bid data social analysis services for financial institutions like hedge funds and banks.
A new entrant, Brand Radar, has a goal of providing the retail market social sentiment analysis tools that it believes until now have been limited to institutions. Revealing its product to the public after nine months of development, Brand Radar is taking the unique approach of launching a Kickstarter campaign to offer the product to traders.
Connecting with Mike Handley, Project Organizer for Brand Radar, he explained that despite having interest from angel investors, they chose to use Kickstarter as it “has a good track record of helping start-ups develop products.” In this regard, backers will be able to access the initial beta solution immediately, with funding from the Kickstarter campaign going towards developing their finished Next Generation product.
Nonetheless, Brand Radar is taking a chance that customers will back its project, as other similar offerings are typically provided with a trial period. Handley explained that they believe Kickstarter could work for them as the community has an “investment focused mentality” that is interested in help funding new startups and innovative products. In addition, Handley stated that they are preparing free trial data to support the claims of their product. Overall, like other Kickstarter projects, there is a leap of faith that backers must be prepared to take before receiving the product.
In relation to how Brand Radar compares to existing products on the market, Handley expressed that they believe they will be able to provide improvements by integrating new advancements in the Machine Learning and data analysis. According to Handley, at the root of the product there are data collection and analysis techniques used by the Planck satellite and include using “neural tensor networks to get a fairly significant edge on sentiment understanding.”
Currently being offered to backers is Brand Radar’s Beta version, as well as the Next Generation version when completed later this year. The sentiment focuses on producing expected odds of rising and falling for Nasdaq listed technology stocks. In addition to the odds, Handley stated, “We will also publish how robust we think the prediction is.” Looking ahead, Handley expressed that beyond evolving its analysis, the firm has plans to produce a mobile app for on-the-go access of its sentiment results.
