The company, founded in 2012, was one of the first to bring together analysis of both traditional media and social media. It’s an area that has long been siloed; traditionally, social and traditional media have often been treated and sampled separately – and in many cases were devoted to specific outlets. For instance, before it was acquired by Twitter in 2014, Gnip competed on its aggregation from a few select social media outlets. Until now, the company operated below the radar, carrying only a modest $60 million of venture financing, but it has some very high profile clients including federal agencies and household brand companies. But it just appointed veteran technology executive Guy Churchward as CEO, with the intent to broaden its focus. Our interest was piqued by the confluence of two major news stories that broke barely 15 minutes apart yesterday. Just before noon, the US Senate passed a bipartisan infrastructure bill; on a normal news day, bipartisan agreement on any legislation would have been the top story in a country that has become more polarized. Barely 15 minutes later, New York governor Andrew Cuomo announced his resignation, subsequently burying the infrastructure bill in attention. We asked Zignal to show us their analysis of how the competing headlines played out, and the chart at the top of this post shows how the Cuomo resignation quickly eclipsed the Senate bill and became the day’s top story. Zignal Labs aggregates content from hundreds of millions of traditional and social media sources, ingesting text content and metadata (the company did not give a precise number) at a rate averaging 6000 articles per second. It describes its services as “narrative intelligence,” and differentiates in its ability to provide real-time pictures of trending topics, sentiments, and risks. It provides monitoring and analytics, which can report response to PR communications strategies, sentiment analysis, identification of influencers, and assessments of risk. From each feed, Zignal Labs ingests text and metadata, and conducts up to a couple dozen enrichments of incoming documents, articles, transcripts, or tweets. It uses machine learning techniques to distinguish human content from that of bots; its models account for factors such as time of day, frequency of posting, and content of posts. It uses a variety of streaming engines, along with Kafka for publishing alerts and updates, and Elasticsearch as its search engine. It provides a variety of visualizations that can aggregate the number of mentions of a certain topic, keyword, or sentiment across time; graphs showing connections between different mentions; and the ability to conduct pinpoint searches, such as who was the first Twitter user to post the hashtag #BlackLivesMatter. The company is hardly alone in casting such a wide net. Its better known rival, Cision claims to have “the most complete collection” of online news, blogs, social, print media, broadcast channels and online forums, but it tends to focus on generating reports on a daily basis. By contrast, Zignal Labs plays up the real-time nature of its analyses. Such real-time social barometers may be common with specialized services monitoring social media, but rarely do they also combine traditional media. Meanwhile. Recorded Future, which was acquired by private equity firm Insight Partners in 2019, also monitors social media firehoses, but concentrates on identification and prediction of security threats. Zignal Labs’ differentiation is that it can provide real-time pictures spanning social and traditional media channels, and has its own patented algorithms for summarizing content. By far, the company’s largest segment is public sector, where, like Recorded Future, it helps government agencies monitor threats. But it also serves clients across energy, telco, manufacturing, financial services, travel, and technology sectors. On the roadmap, the company is seeking to add more predictive content, especially with its “Emerging Narratives” offering that helps clients anticipate narratives that are about to emerge.
