Reading public sentiment in real time, as the news happened
Elections, court rulings, and crises move faster than traditional polling can follow. Over twelve months, Ripple Research worked with Digital Planet at Tufts to turn large-scale social listening into a sensing function for policy and crisis management.
The mandate
When a fast-moving event unfolds, a contested court ruling, an election, a platform upheaval, a public-health emergency, institutions need to understand how people are actually reacting while there is still time to respond. Conventional polling arrives too late and too thin. Digital Planet at The Fletcher School partnered with Ripple Research to adapt large-scale social listening and digital ethnography into a sensing function that policymakers and crisis managers could use in near real time.
What the partnership produced
Over twelve months the work applied the same underlying method, listening to online discourse at scale and reading it for sentiment and narrative, to a series of live and contested moments.
Inclusive crisis management during the pandemic
Drawing on more than 873 million online interactions, a demonstration of how real-time social analytics can serve as a sensing function for inclusive policymaking.
The US midterm elections
Reading the narratives and sentiment shaping political discourse through a contested election cycle.
The Roe v Wade decision
Tracking how public sentiment responded to one of the most polarising rulings of the period.
The acquisition of Twitter
Analysing how discourse shifted around a change of ownership at a major platform, and what it signalled for the information environment.
Trust in crisis
Examining the dynamics of public trust in institutions during periods of acute uncertainty.
What it showed
Across very different subjects, the same finding held: social analytics tools built for commercial marketing could be adapted into a rigorous instrument for public-interest understanding, giving decision-makers a read on sentiment and narrative at a speed and scale that surveys cannot match. The method treated online discourse not as noise to be counted but as a signal to be interpreted, which is what made it useful for policy rather than only for measurement.
The value was not the volume of data. It was the ability to read, while an event was still unfolding, how people were making sense of it.
What it left behind
The partnership produced a body of published reports, several available through Digital Planet, and a demonstrated method for real-time social sensing that has carried into Ripple Research's later work on health and political discourse. One strand of the underlying research was also published in a peer-reviewed venue, extending the approach into the wider evidence base.
Working on a similar question?
Most engagements start with a conversation about what you're trying to understand and what a useful answer would look like.