Media monitoring — where news sources and other public information outlets are scanned regularly for mentions of specific organizations — is a well-established service used by companies for market intelligence and to measure sentiment around their businesses. Today, London-based Signal AI, which has built a substantial operation in the area, has raised $25 million to expand to newer frontiers: applying AI to that public data to also spot themes, risks and opportunities to make better decisions; and continuing to take that business to new markets.
The Series C is being led by Redline Capital, with previous VCs MMC Ventures, GMG Ventures (an investment firm linked to the Guardian Media Group) and Hearst Ventures also participating. The startup, which has now raised around $43 million, is not disclosing its valuation, but CEO and founder David Benigson said that it is “significantly higher” than before (it last raised $16 million a year ago), after growing revenues at well over 100% each year for the last two.
The presence of not one but two media-linked investors in the round points to the startup’s roots: Signal AI had previously been called Signal Media and worked mainly around the task of media monitoring in the more traditional sense: tracking how companies were being mentioned in the press.
Benigson said the reason for the rebrand was to “signal” to the world how the startup was widening its remit, both in terms of its sources of data and also in terms of its customers and how they now utilize Signal’s technology.
The challenge and opportunity that Signal AI is tackling is the fact that the world is awash in information, much of it unstructured and usually bombarding us from many angles, but tantalising all the same for hinting at the insights that it might hold if it could be looked at in a more comprehensive way.
“When we started six years ago, it was by aggregating news data and tapping the repository of global, traditional media,” Benigson said. “We have since broadened into social media, broadcast and radio, and regulatory information and started to apply more machine learning to structure that data.” In addition to selling services directly, the company now partners with third parties to build analytics around more targeted subjects, such as a changing regulatory climate in a specific area, which in turn is sold on by the third parties to other clients.
The company, for example, works with Deloitte’s tax division to monitor how tax codes are evolving and likely to move over time: the firm used to keep its own clients up to date verbally on these details, and now it sends alerts automatically with insights — a switch that Benigson said has saved the company $100 million a year in human and overhead costs.
Signal AI sits in a relatively new, not clearly defined area of business. It can be comparable with the likes of Meltwater, Cision (Gorkana) and even Dataminr when it comes to reading media in real time. But it also works a little like business intelligence or market analytics in its predictive analysis. The company refers to its specific area as “augmented intelligence”:
“There is a trend / emerging category that is far less crowded and defined than business intelligence or analytics,” Benigson said. “For me, it’s around taking those same values of BI and applying them to the world of data that sits outside the organization. There are very few companies that use augmented intelligence, although we are seeing management consultancy firms and others we potentially compete with convening around this space.”
It’s that open water that has attracted investors to the company.
“In this new digital era of news and content, having an adaptive platform to help the world’s leading organizations see around the corner is invaluable,” said Nicolas Giuli, partner at Redline Capital, in a statement. “Signal AI’s team of data scientists and engineers have been at the forefront of the AI revolution and we are excited to take this journey with them as they continue to scale across the world.”
In this day and age, data is indeed very much a hot commodity, but I’d argue that it’s also a hot potato. By that, I’m referring to the rise of security breaches, people’s growing awareness of how their personal information is being used (and too often misused) and regulation that now draws lines on how data can be used, after organizations failed to draw those lines themselves. All of these have made concepts like data analytics and data mining, even around supposedly anonymised information, feel more nefarious and unclear in their target purposes and ends. That potentially spells trouble ahead for companies that dabble in this space.
Benigson, for his part, was unequivocal on where Signal AI stands on any kind of anonymised or other potentially personal data:
“We purposefully avoid those data sets because we feel that the challenges are not being met,” he said. The exception, he noted, was in cases where a company uses its own internal data for its own purposes, but this does not feed into Signal’s AI engine, which focuses only on publicly available third-party content. “We have no plans to incorporate that kind of data ourselves. We have an opportunity to do this in an ethical manner.”