Julia Filsell-Chapman, head of digital at PR agency Speed: “Though AI hit the mainstream at the end of 2022 with the advent of ChatGPT, elements of AI, such as machine learning have been embedded in many of the measurement tools we use for much longer.

“Measurement sits at the heart of everything we do for clients, as circulation numbers just don’t cut it anymore. Central to this is using AI tools to help us understand large sets of anonymised data easily, uncovering key insights for clients - such as patterns in social analytics - that can help us deliver more sophisticated content and media strategies.”

Johnny Bentwood, global president of data and analytics at PR agency Golin: “AI enables faster and more accurate data modelling by removing human biases, automating tedious analysis tasks, and providing more nuanced insights using advanced algorithms. Specifically, AI can detect misinformation campaigns, make qualitative survey analysis cost-effective, improve sentiment measurement, and uncover strategic opportunities from vast datasets. My go-to AI tools for this are Talkwalker’s BlueSilk, Microsoft’s Copilot and Anthropic’s Claude.

“AI unlocks new growth opportunities by enabling intellectual processes not previously possible. Examples include detecting fake news and profiles, analysing open-ended survey responses to reveal audience perceptions, measuring emotion intensity in content, summarising brand and customer contexts, ensuring consistent reporting formats, and surfacing hidden insights from overwhelming data. However, AI remains a tool requiring human judgement, not a substitute for intelligence and intuition. Exceptional AI solutions from Cyabra and Canvs are my personal favourites to help in this area.”

Launa Ribeira, founder of PR agency Dauntless PR: “We use Critical Mention for tracking - it’s a combination of AI and automation. It’s a great piece of kit! It picks up where clients have been featured online, even in print, as well as working out the estimated audience and the value of the piece.

“My tip for anyone else considering using Critical Mention or any other AI is to remember that they are there to assist, and do sometimes miss things/get things wrong so it’s important to keep checking manually too. We use Critical Mention on a daily basis but will do a weekly spot check in case it has missed anything (which it sometimes does).”

James Cunningham, complete health optimisation coach at health and fitness community Total Shape: “While AI is powerful, it is not perfect. Human expertise provides nuance and context, which can be used to refine and validate AI-generated insights. Before fully relying on AI-generated metrics, it is critical to manually validate the results. This iterative process of validation, feedback, and refinement ensures that AI-driven insights are consistent with the complex and nuanced nature of public relations.

“Moving on to data quality, it's not just a box to check - it's the foundation of effective AI measurement. The old saying "garbage in, garbage out" applies here. No matter how sophisticated your AI tools are, if the data they're analysing is incorrect or irrelevant, the conclusions you draw will be flawed. Investing time in curating a robust dataset, ensuring its accuracy, and keeping it relevant is the foundation for gaining meaningful and actionable insights from AI.”

Allison Spray, global head of technology and intelligence innovation at PR firm Hill and Knowlton: “We’ve been using AI for some time, particularly within our data intelligence function to support measurement.

“We’re using AI in measurement across a spectrum of complexity: from leveraging the AI and ML in our existing data partners, to partnering with Pendulum to create Sonar, a disinformation and geopolitical risk tool; and finally building our own tools to solve recurring issues at scale.

“For example, we have a purpose-built pipeline in our data solution which allows us to automate the collection of media KPIs, enrich it with additional data sets and code for qualitative metrics like message pull-through using OpenAI safely within a secure Azure environment. This means we’re able to focus human time on analysis and recommendations rather than data processing and deliver insight at the speed required by modern strategic communications functions.”

Jack Neale, business analyst at PR Agency 72Point: “We've been aligning our AI stack to provide insight into the impact of campaigns we deliver. We're even implementing it in our own technology. In PR measurement, AI offers remarkable speed and efficiency in analysing data, yet its ability to accurately classify sentiment into positive, neutral, or negative is not flawless. Has it got to grips with sarcasm, irony, exaggeration, red herrings, or anything else that could be thrown at it? Not yet, in my opinion.

“While algorithms can recognise patterns, they lack the nuanced understanding inherent to humans, such as cultural context and subtle language cues. Human judgement remains indispensable for interpreting these nuances, refining AI systems, and correcting biases. The synergy between AI and human expertise is crucial for effective PR measurement. AI enhances efficiency, but human intervention ensures accuracy and relevance, guiding the interpretation of data and refining algorithms.”

Ailene Barr, acccount director at marketing agency BDB: "AI-powered tools have helped us elevate our PR reporting. Coverage monitoring software, like Signal AI, has enabled us to better track multiple competitors and key messages. This means we can see our clients' share of voice across key competitors whilst also tracking product names and any other key words. For example, we've been able to track share of voice across specific health benefit areas for our nutrition and pharmaceutical clients, providing valuable insights that previously relied on too many manual searches to be practical.

"With all the relevant brands and keywords being tracked in real time, our teams can easily see the volume of news and the stories behind the statistics which frees them up to dive deeper into the ‘why’ and give our clients actionable learnings and recommendations. AI for the data scraping plus humans for the interpretation has been a winning formula."

James Crawford, managing director of PR firm PR Agency One : “Embracing AI in the realm of PR measurement reflects a significant advancement, offering an unprecedented level of analytical depth and accessibility. We've experienced first-hand the transformative power of AI, particularly in its ability to handle data through .csv uploads and facilitate queries in natural language.

“This innovation effectively places a mini data analyst at the fingertips of every account executive, elevating the quality of outcome reporting to new heights.

“However, it's crucial to approach AI integration with caution. The fundamental principles of data analysis, insight generation, and measurement remain unchanged. The true art lies in measuring outcomes and weaving these metrics into a coherent narrative that traces the journey to these results. Whilst AI tools offer remarkable capabilities, they are not a panacea. Following the AMEC Framework remains, as ever, fundamental.”

Kate Geldart, online marketing specialist for international business Custom Neon: “We've included AI tools into our PR campaign evaluation process. Tools like Google Analytics for website traffic analysis have been particularly useful. They provided us with a report detailing the effectiveness of our PR initiatives on several media channels.

“Google Analytics is known for its comprehensive analysis and easy-to-use interfaces. A tip for leveraging AI in PR measurement is to clearly define your KPIs before launching campaigns. By tracking particular indicators, including engagement rates, reach, and conversion metrics, that correspond with your PR goals, you can use AI technologies more effectively.”

With a whole host of potential uses, AI has become an essential tool in PR measurement in recent years. The key takeaway, however, is not to forego the human touch. As the experts above agree, all AI approaches work best when used in conjunction with human interpretation.