Marginal Moments was an AI R&D project, using AI personas to explore how election promises are engaging the public. We followed five AI generated voters throughout the election campaign, introducing them to a wide spectrum of national and local election materials, to understand how the speeches and pledges made by Parties impact their voting intention.
Which announcements ranked best across eight policy areas? Our Championship tested 56 of the key policies announced in party manifestos with the public. Using a message testing approach called Ipsos DUEL, each participant was shown a random selection of 16 policy pledges in their own mini-tournament, knocking out the worst performing to select an overall winner.
