Manuelita Maldonado was the 2020 winner of the Makovsky Best Master’s Thesis of the Year Award. This blog post is based on her thesis research, “The Rise of Intelligent Machines: How Artificial Intelligence is Transforming the PR Industry.”
Earlier predictions noted that by 2020, all our digital interactions would have generated 40 times more bytes of data than there are stars in the observable universe. As more clicks, shares, posts and comments hit the digital space, this staggering amount of data will continue growing exponentially, forcing businesses to look for novel ways to analyze the content of these new streams of information in order to obtain actionable insights. Artificial intelligence and natural language processing technologies are playing an essential role in helping organizations process and interpret large volumes of data in short periods of time, allowing them to better understand their competitors, consumers, media, and other key stakeholders.
These large volumes of data can be classified as either structured or unstructured. Structured data uses a predefined format where all fields are fixed, making it easy to export, store and organize information in databases like Excel or SQL. Unstructured data comes in all shapes and sizes and can be stored in the form of a video, audio, image or other formats.
The fields of marketing, advertising and public relations are usually more interested in analyzing unstructured data like social media posts, photos and videos, because it allows agencies to move beyond traditional metrics (such as AVE or number of impressions) and obtain valuable insights like market trends, customer preferences and feelings toward a brand. And it is only through artificial-intelligence-enabled technologies like natural language processing (NLP) that PR professionals can analyze large data sets to obtain results more efficiently and accurately.
1. NLP Is Helping PR Pros Analyze & Measure Audience Sentiment
Natural language processing can be described as a subset of artificial intelligence that gives computers the ability to process, understand and interpret human language. As such, NLP gives PR professionals the ability to analyze millions of tweets and other written social media posts to identify main topics of discussion or to measure audience sentiment (positive, negative or neutral) – helping them assess public opinion about their brands or clients.
According to the Chartered Institute for Public Relations study, “Humans Still Needed,” social media analysis is one of the most common PR activities impacted by artificial intelligence. This explains why almost half (44%) of the PR professionals who participated in the 2019 USC Annenberg Global Communication Report chose media monitoring as the most relevant tool for their current work.
2. Sentiment Analysis Is Not 100% Accurate – And Probably Never Will Be
One of the biggest challenges of sentiment analysis is how complex and subjective human emotions are. Sarcasm, tone and context can change the meaning of a text, and opinions are often subject to the interpretation of the reader. In fact, it has been estimated that when humans evaluate a text’s sentiment, they tend to agree 80% of the time, making it almost nearly impossible for human-made tech platforms to achieve a 100% accuracy level.
Some experts agree that sometimes measuring sentiment using a communications tech tool can be like flipping a coin, or only 50% accurate, since these platforms often struggle to measure more nuanced posts or are unable to filter and interpret the information through the lens of a company or brand.
Advanced communication tech vendors are creating more sophisticated sentiment analysis tools by using third-party services that hire real humans to classify large sets of data as either positive, negative or neutral, and also has them point out instances where the sentiment can be particularly harmful to a given brand. This classified data is then fed into AI algorithms, which learn to recognize patterns and interpret context through the eyes of a given brand. But this is often offered as a “premium” service because it is a long and personalized process that requires significant human intervention.
3. Other Tech Features May Help PR Pros Complement the NLP Gap
Even though NLP technologies are still at an early stage, there are other social media monitoring tools available – such as word clouds, topic clusters, and top hashtags and retweets – that can complement the insights provided by sentiment analysis tools.
It is undeniable that social media is playing a critical role in our economy, enabling companies to increase their reach and engage with potential customers. AI-powered tools are expanding PR professionals’ understanding of the social media landscape by allowing them to identify market trends and analyze people’s emotions – something that was simply not possible before. But these insights must be complemented with human reasoning, critical thinking and creativity.
Artificial intelligence has the power to process vast amounts of data to provide the “what,” and it is up to us, as PR professionals, to put those insights into context and understand “why” this is happening and develop our strategies accordingly.