Sorting the signal from the noise is an increasingly tedious task for anyone spending time online. London-based startup Signal Media is using that as an opportunity to build a business out of automating real-time information gathering and filtering — crawling more than 75,000 online news sources, 3.5 million blogs, 100 social networks, plus research publications and other unstructured data sources so its customers don’t have to.
And doing so in a more flexible and accessible way that incumbent media monitoring services such as Precise and Cision, according to co-founder David Benigson. Signal’s SaaS business is being positioned between “unwieldy” and expensive legacy platforms on the one side, and freebie consumer apps like Google Alerts, Feedly and even Flipboard which are easy enough to use but don’t have powerful enough features for enterprise customers’ needs.
“When we founded Signal we saw a gap in the marketplace between those two groups. Could we apply some of the user interface design that’s more intuitive and used by the consumer tools but actually then use a really powerful technology so it was fit for purpose for an enterprise,” says Benigson.
The U.K. startup, which was founded almost three years ago, launched its commercial service around six months ago — after focusing on building outs its core text analysis platform, which uses NLP and machine learning to process and analyze millions of online news sources.
Unlike some legacy media monitoring techs, Signal is not just doing keyword searching either. Its focus is “a series of text analytics components” that aim to filter and “extract value” from the data as it flows through the system, says Benigson.
“Most of the incumbents it begins and ends with keywords and so you have issues like if you type ‘apple’ you get apple the fruit as well as Apple the company,” he says. “You can’t search for topics. There’s no de-duplication so you get vast amounts of duplication within the data. There’s no event detection or clustering. Their summarization technology doesn’t work so it’s typically just the first paragraph. So what we’ve done is we’ve looked at those individual problems and we’ve tried to build and leverage machine learning technology to try and solve each of those particular issues ones by one.”
Signal’s platform offers a machine learning-powered topic classification system that lets users define a topic they’re interesting in and then train the algorithm on what should and shouldn’t be in that topic. Another tech within its platform is called ‘entity recognition’ which analyzes the context of the language within an article to try to distinguish intelligently between apple the fruit and Apple the technology firm (for example). Signal’s platform also has a summarization feature to pull out salient facts from text so it can present digests to its users.
Signal has around five customers thus far, including Jamie Oliver, which is using Signal to power a curated food news website called Food Revolution Daily; and the Centre for Policy Studies, which is using it to monitor changes in the political landscape. Other extant customers include executive search firms and media companies, with discussions ongoing with PR and comms organizations.
It’s just announced a $1.8 million seed round, led by Frontline Ventures, and with various other investors joining the round — including Reed Ventures, Robin Klein and Jonathan Goodwin. The funding has been used to hire data scientists and software engineers to build out Signal’s core tech platform. It will now also be funneled into a sales push as it ramps up for the next push to grow the business. Signal is initially focused on the U.K. market but Benigson says he sees strong potential for international expansion.
“For the last two and a half years we’ve really been focused on building out the core [machine learning and NLP] technology that underpins the whole platform. We’re processing pretty much the entire world’s news media, blog content and a lot of social media as well, so it’s a lot of data to process and technologically very challenging,” he adds.
Benigson reckons Signal can have multiple applications — whether it’s in corporate comms for monitoring an organization, its rivals and a whole industry. It can also be used for market intelligence, or for looking at peripheral risk, or other monitoring opportunities within a particular domain. And for media applications — as a content marketing tool to curate content and republish it. “The opportunity to monitor, discover and draw insights from the world’s media has a vast number of applications across many different verticals,” he adds.
Keeping abreast of the news when there’s so much of it is also a growing challenge for senior executives in an organization — which he believes is another opportunity for Signal. In other words, for C-suite execs, the traditional daily email bulletin isn’t going to cut it for much longer.
“Increasingly executives, who are exceptionally time short, need information to do what they do effectively,” he argues. “Things move so quickly now. Reputational risk can emerge in a matter of minutes. And so can opportunities that they might have missed as well. So the executive, C-suite is another band across the organization that needs to be using tools like Signal to gather information and intelligence much more rapidly and effectively.”