React Score will use natural language processing (NLP) models to analyse conversations from more than 100 million sources to help spot risks and opportunities for brands.

The tool works by providing real-time ratings based on three key categories: harm, or damaging content that could affect brand image; spam, or irrelevant and promotional messages that clutter brand conversations; and emotion, looking at the emotional tone of conversations and gauging consumer sentiment.

Brandwatch said that React Score provides rankings across categories, helping brands prioritise responses, monitor reputation health and track performance. 

The release follows on from a recent collaboration between Brandwatch and Google Cloud, with plans to use Vertex AI and Gemini models.

Brandwatch will also launch its Iris AI query assistant in the first quarter of 2025, which will use generative AI to support query creation.

Matt Tippets, chief product and technology officer at Brandwatch, said: “It has never been more important for brands to have access to accurate and reliable insights.

“Brandwatch built this game-changing hybrid approach, combining generative and proprietary AI with human oversight to deliver accurate, reliable insights that help our customers confidently navigate the complexities of the digital landscape.”