The pandemic has highlighted the importance of credible and reliable information. However, the reliance on user-generated content online as a source of truth for news has increased the risk of exposure to bias, misinformation and potentially harmful content.
From trends to breaking news, celebrity culture to political and medical information, the variety of content available to consume is expansive. Many people use the internet on a daily basis for updates on what is happening around the world. Journalists have had to work even harder to navigate the challenges presented by online conversations and reporting.
Fast and fierce analysis
On average, there are 500 million tweets shared a day, and new content is increasing exponentially. The past two years is an example of this, with Covid-related content generated faster than humans can analyse or check for accuracy.
The reliance on brands and people, such as journalists or peers, to provide credible information poses a significant challenge for the internet. With Edelman finding nearly 1 in 2 people believe the government and media are divisive forces in society and over half of Gen Z and Millennials currently boycotting at least one brand – these factors are adding further expectation and pressure which can drive the spread of false and misleading narratives.
Marketers have had to learn to interpret and react to events quickly. Communications teams have had to frequently navigate through a sea of information and opinions to decipher emerging trends and ideas. The fight for faster analysis of changing opinion and intent is fierce. Making sense of content is essential for marketers to stay ahead of the curve.
Sentiment is no longer working
Previously, the manual analysis of mass mentions about a person, brand, or story has forgone the nuances of language, geo-local, political, and cultural context. Identifying new narrative trends and predicting future sentiment and intent was not possible.
Automated sentiment analysis was developed to overcome challenges in media monitoring. Although they’ve been updated and improved over time, sentiment analysis models are fundamentally flawed – they cannot understand the context. Inconclusive and harmful content that is potentially abusive or offensive is missed due to seemingly positive language.
Sentiment has been king for over a decade, but now there’s a new kid on the block. Businesses can leverage the modern capabilities of AI and other technologies to transform how they predict and stay ahead of narratives before they break.
Converting opinions into intention
The challenge for marketers analysing online content is to convert opinions into an intention. Progression in natural language processing has meant brands can understand the true intention of their audiences by looking at content in context.
Rather than simply classifying words objectively, stance analysis models consider them subjectively towards a given topic. By accounting for nuances in language, insights are created quicker, and predictions are more accurate.
Every opinion counts. When it comes to mass data analysis, no longer do teams have to endure hours of manual labour to account for the limitations of sentiment models.
Assessing risk with AI
New narratives can be identified as soon as they start to develop. If left unchecked, trends can grow and reach thousands of people with harmful consequences.
Artificial intelligence can assess the risk of online content about a brand or industry, driving efficiencies and increasing the quality of understanding conversations, perceptions, and intent. AI can also identify the influencers driving these narratives, enabling counter-measures to be deployed before they go viral and hit mainstream audiences.
People can say the same thing in different ways on an increasingly disparate set of platforms but these mentions are all driven by the same narrative. Narrative monitoring of this kind automates the process of identifying mentions with this shared meaning.
The future of media monitoring
Online platforms are defined by their interactivity, connectedness, and user-generated content. Today, social media is consumed as regularly as breakfast. The power of online platforms has meant the world is more knowledgeable but vulnerable to harmful and inaccurate content, which can easily proliferate if left unchecked.
Brands have had to improve their social listening to keep up with what is shaping public opinion and what is being shaped by it. Narrative monitoring has the power to assess the level of risk harmful narratives pose to brands and identify the influencers behind the content. Analysing stance over sentiment allows brands to get a deeper and more complex insight into audience opinion and intent to pre-empt stories before they potentially threaten brand and internet safety.
With user-generated content expanding and the metaverse set to radically increase information exposure – brands risk sleepwalking into a catastrophe of brand safety. Narrative monitoring can help with the fight.