As someone who has spent over 15 years working directly with public relations measurement and intelligence and more than a decade helping brands make sense of their media performance, I can say with confidence (and a touch of media analysis fatigue) that not all PR metrics are doing what we think they are doing. And when it comes to sentiment analysis, many of us have been led by tradition, not truth.
In my constant pursuit to help PR and comms professionals access metrics rooted in objectivity and research, I had to take a deeper look into how sentiment is currently being measured. After spending time digging into the methodology, analysing patterns, and comparing outcomes, it became clear: sentiment analysis by frequency has overstayed its welcome.
“Too often, we focus on counting sentiment rather than weighing it — frequency tells us how much, but deeper analysis tells us how much it matters.”
For too long, we have boxed sentiment into just three labels — positive, negative, and neutral — and then celebrated (or panicked) based on how large each segment appears. If a brand has 60% positive sentiment, someone somewhere is already serving small chops and cutting cake. But ask the hard question: what does that 60% actually mean? Does it carry weight? Is it impactful? Is it meaningful? I recall being in a strategy session where an agency CEO saw a 60% positive sentiment report and asked, “So… should I be excited or worried?” And truthfully, the data didn’t answer that. In another situation, a client saw 35% negative sentiment and wanted to escalate to crisis mode. Again, I had to ask, what kind of negative are we talking about?
“When it comes to sentiment analysis, it’s not enough to know the quantity of sentiment; you need to understand the intensity and quality of that sentiment. Without that, data can lead you astray.”
You see, frequency analysis doesn’t tell you intensity. It doesn’t ask, how positive is this positivity? Or how damaging is this negativity? In reality, a comment like “The brand dey try sha” (Nigerian slang for “they are doing okay”) and another saying “This brand saved my life!” are both tagged as positive but are clearly worlds apart in tone and impact. That is where the problem lies — we have focused too much on counting sentiment without weighing it.
Research provides a more meaningful approach. The empirical formula I recommend is:
Sentiment Score (StSc) = (Number of Positive Mentions – Number of Negative Mentions) / Total Number of Mentions
This gives us a normalized sentiment index between -1 and +1, where 0 is neutral, and the extremes show very strong positivity or negativity. So if a brand has 3 positive and 2 negative mentions out of 10 total, the score becomes (3 – 2)/10 = 0.1 — slightly positive. But if it is 8 positive and 1 negative, the score is 0.7 — that is significant. Now compare that to simply saying “80% positive,” and you see why frequency alone is not enough. The difference is in the depth of interpretation. This formula still isn’t widely used across the media intelligence space, but one company that’s already ahead of the curve is Truescope (North America) — where my friend and industry expert, Todd Murphy , serves as President of North America.
“Objective metrics that account for sentiment weight and distribution are what truly empower PR strategies. It’s not about having more positive mentions — it’s about understanding the level of positivity and negativity and its true impact on brand perception.”
To fix this gap in analysis, we have developed the Future-Proof Sentiment Score Framework – A P+ Measurement Services Proprietary Sentiment Score Framework. This includes a more advanced Sentiment Weight Score and Distribution Matrix, which doesn’t stop at “positive/negative/neutral,” but goes further to classify sentiment into strongly, moderately, and slightly — for both positives and negatives. This matrix brings clarity to brands and communications teams. It helps you know when to celebrate, when to adjust, and when to truly raise the red flag. Starting from Q2 2025, all clients of P+ Measurement Services will have access to this upgraded sentiment analysis dashboard, alongside a dedicated dashboard that tracks the media performance of competitive CEOs. And I can say with confidence — it changes the game.
“Let’s stop being impressed by pie charts that look shiny but don’t provide actionable insight. Understanding the meaning behind sentiment and the true impact on your brand is what matters.”
I will give you a practical example. A multinational brand we monitored recently saw 35% negative sentiment and was ready to call a crisis meeting. But our deeper analysis showed 80% of that negativity was slightly negative—things like delayed customer service or pricing feedback. Meanwhile, their strongly positive mentions were increasing daily, driven by user experience reviews. Instead of reacting emotionally, the brand realigned calmly. No panic, just action. That is the power of context.
So, let us stop being impressed by shiny pie charts. Let us stop reporting frequency without understanding what it means. A sentiment report that doesn’t answer so what? and what next? is simply not useful. This is why I always say: vanity metrics may look nice in a report, but they can’t guide strategy. Objective, research-backed metrics can.
“Vanity metrics can’t guide strategy. Only research-backed, objective metrics help you turn insights into action.”
At the end of the day, this isn’t just about a better dashboard. It is about moving our industry forward. For those interested in the technical side, I am happy to share more about lexicon-based sentiment scoring and resources like the Harvard General Inquirer—empirical research that goes beyond assumptions and digs into real language science. But even without the jargon, the message is simple: frequency tells you how much, but only deeper analysis tells you how much it matters.
