Social media is fast becoming a significant influence on the path to purchase. Businesses are taking notice and looking for ways to source crowd insights, sentiment and authentic user experiences to leverage the sway social content has over prospective buyers.
But there’s no effective social media search engine, especially on the product end, and the landscape is vast. It takes the likes of Big Data and machine learning to meaningfully scratch the surface of social.
Enter Feelter, a technology startup that looks to bring the power of social media sentiment to brands.
“The most trustworthy, meaningful source of information out there right, the place where people willingly share their experiences, are social networks,” said Saar Székely, VP of research and development at Feelter.
Founded in 2013, the Tel Aviv-based startup begins by sifting through social media, looking for content that mentions or features client brands and products.
“It’s an enormous amount of data,” Székely said, “but the text is usually short. You need to conclude quite a lot about the relevance and value from quite a small amount of information.”
From the unstructured social data it gathers, Feelter uses machine learning and advanced language processing to analyze the vocabulary, grammar and sentiment within each mention of a brand across social media. This information is combined with the history and quality of a user’s content along with a metadata analysis of likes, shares and the comments on each post.
“We take into account all of these different factors so that we can clear away all of the noise,” said Székely, “so we can put all of your efforts into analyzing the most meaningful parts of the social discussion.”
Feelter scores each relevant social media post. Then, its machine learning algorithms use the scores to curate Feelter’s social feed widget embedded at strategic locations on brands’ websites.
“We’re trying something completely new,” said Smadar Landau, Feelter cofounder and CEO. “We bring the truth. Customers want to understand social conversations. Unlocking the most relevant content behind all of those selfies so this is what makes us unique.”
Feelter’s social curation has recently attracted the attention of major industry players, said Landau, across a number of verticals as companies increasingly realize the importance and value social conversations have in supporting and influencing consumer decisions.
“What’s interesting,” said Székely, “is that people aren’t always looking to social content to see what other people said. They want to see what a product says about them, what kind of people are using it and in what situation. You get a fuller experience of how the product will fit within your life.”
Székely noted that, while Feelter sees participation and success across demographics, what data the company has suggests that it is especially effective on younger generations.
“Maybe it’s because the younger generation shares more,” he said, “but the information is relevant across demographics.”
Every vertical and every brand are different, noted Székely. “Each one of them has unique features, and our algorithms identify and optimize for them based on the different user engagement in different verticals.”
Easy Click Travel, an early adopter of Feelter’s service in the travel industry, saw a 39.5 percent rise in conversion rates, a 48 percent increase in time spent on the site and nearly a 50 percent reduction in bounce rates after embedding Feelter’s widget at two key spots on its website — the hotel search results page and individual hotel pages.
Moving forward, Feelter finds itself in a unique position to fine-tune the relationship between businesses and their customer base. By collecting social data, Feelter is gaining insight and access to what specific parts of the social conversation best drive consumer action — insight that the startup has in the form of new solutions.
“We’re exposed to both sides of the equation,” Székely said. “Being right in the middle puts us in a very interesting position. We’re building more tools for the retailers and the brands behind the scenes.”
Additionally, Feelter is working to grow its image and video recognition technology, both to improve the specificity of its analytics capabilities but also to keep pace with the changing social media landscape — especially as it pertains to video content.
“We’re focusing heavily on developing image and video recognition, moving with the changes in social networks,” Székely said. “It’s surprising how fast the landscape changed in terms of video, how quickly everyone got into this content war. It’s fascinating to see from our position how content shifts from place to place, how sharing behavior changes.”