BrandsEye: Driving Accurate Opinion Through Advanced Sentiment Analytics

Happier customers are more loyal to a brand and the key to more profitable business. The best organizations understand sentiments of their customers. Real-time insight into a customer's feelings through advanced analytical techniques offers an effective tool to drive engagement and improve bottom line. One such company which is pioneering the field of sentiment analytics and helping organizations make sense of public opinion is BrandsEye.

BrandsEye is the world's leading opinion mining company. The company uses a proprietary mix of artificial intelligence and crowdsourcing to mine online conversations for sentiment and the topics driving that sentiment. In 2016, BrandsEye used this technology to predict the outcomes of the Brexit referendum and US presidential election. Today, the company provides accurate and granular opinion data to organizations around the world.

BrandsEye was founded in the age of the blogosphere before social media became popular. The company's mission has always been trying to understand what people said – understanding opinions both individually and broadly. In doing this, BrandsEye found that algorithms were good at some things and poor at others. They were particularly poor at understanding sentiment, which is key to understanding opinions. This led the company to create its crowdsourcing approach to deliver really accurate sentiment (and later topics that sentiment was based on) to its clients.

BrandsEye's algorithms draw on large volumes of public data from social media platforms in order to understand how people are feeling at scale. BrandsEye analyses data on a range of issues that in turn help governments and businesses make better decisions, based on the volunteered feedback of their constituents and customers. The data has a number of important uses, from delivering superior customer experience improvements to mitigating political risk in an election campaign.

A Passionate Leader and Entrepreneur

Craig Raw, Founder of BrandsEye is an engineer with a passion for creating new and unique products that combine the respective strengths of humans and machines. Craig was CTO of the digital agency Quirk from inception to its sale to WPP in 2014. During this time, he created BrandsEye to solve a growing need to understand online conversations deeply and accurately. After Quirk, Craig joined BrandsEye in a full-time capacity to steer the business towards its global ambitions.

Driving Innovation to Improve Customer Experience

Craig feels there are serious limitations of natural language processing algorithms in understanding the way humans talk online. Machines are still not particularly good at accurately understanding the nuances of written language, particularly online conversation which is rife with sarcasm and slang. To overcome this challenge, BrandsEye employs a Crowd of human verifiers to improve the accuracy of its sentiment analysis.

Once the algorithms have mined the relevant public data from social media, it is distributed to the BrandsEye Crowd of trained and vetted human verifiers. The Crowd of human verifiers read hundreds of thousands of online conversations every day, enabling the company to accurately understand not just how people feel but what particular issues are driving them to feel that way.

Each individual social media post is distributed to multiple Crowd members who codes it for sentiment. With this information, clients are able to hear and serve their existing customers and citizens better and reach out to new prospects when they will be most receptive.

The company's Crowd verified approach allows it to reach industry-leading accuracy levels of 95%, making the data trustworthy and actionable. This Crowd verified data then train its algorithms to be more accurate over time.

"From there we can look at the emotive (opinion led) conversation and further analyze for topics – that is, what people are talking about. This allows us to understand their opinions accurately both individually and on aggregate" added Craig.

Recognition for Excellence

BrandsEye gained global recognition in 2016 for analyzing public tweets to predict both Brexit and the US elections providing meaningful, predictive insights gained from analyzing social media at scale. Since then, the world has viewed social media in a different light and the company now caters to clients globally. BrandsEye has also been part of the Global Endeavor Program since 2015. One of the achievements the company is most proud of is the value they are able to create for its crowdsourced contributors, many of them would have been otherwise unemployed, who are able to make a living through the work BrandsEye provide. For the company, creating jobs is a very important outcome, particularly in poorer areas of the world.

Looking Towards the Future

Craig believes the information age is now truly upon us, with the rise of technology companies like Apple dominating the business world. These companies dominate because of the data network effect (also expressed as Metcalfe's Law) which states an exponential relationship between the number of points of data, and the value of that data. Big data is another term used to explain this effect. What we typically see are emergent truths that come from being able to analyze big data sources, truths that were often not previously available. This is part of the general trend from a human-driven, heuristic approach to understanding the world, towards a more algorithmically driven one.

"The rise of the internet of things and other trends has seen the amount and type of data we are collecting expand rapidly. This, combined with AI techniques that have recently had a revival under the name of 'deep learning', is permitting us to understand patterns and categorize trends in ways that were not previously available. This is particularly true in understanding our own behavior at scale. We are now starting to explore how we talk and interact with each other on a scale and accuracy not previously possible" Craig said.