A sudden shift in tone—from collaborative to curt, from detailed feedback to dismissive comments—often signals a change in how someone views you or a shift in workplace dynamics. Documenting when and how the tone changed gives you concrete evidence that something shifted, not just your feeling that things got worse.
Sentiment analysis sounds technical, but it's actually simple: it's AI reading between the lines to understand the emotional tone of communication. When your manager sends you a message, there's what they literally wrote, and then there's what they meant underneath. Sentiment analysis picks up on that subtext.
Think of it this way: "We need to discuss your performance" versus "Let's chat about how things are going!" Both are about performance, but the emotional weight is completely different. Sentiment analysis catches these nuances, even when they're subtle or disguised as professionalism.
Hostile managers are often clever. They can't be overtly cruel (that would be obvious), so instead they use passive-aggressive language, backhanded compliments, or fake enthusiasm that masks contempt. A message might technically be "professional" while being emotionally cutting.
Examples: "That's an interesting approach" (subtext: that's wrong). "You're so ambitious" (subtext: you're overstepping). "We're all learning" (subtext: you messed up badly). When these pile up, you internalize them as personal failure—but sentiment analysis reveals the pattern of negativity underneath the professional veneer.
When you input a batch of manager communications into an AI tool, sentiment analysis scores them on a spectrum: very negative, negative, neutral, positive, very positive. The AI also flags emotionally loaded language—words that carry judgment or hostility. It picks up on:
Here's the genius part: sentiment analysis creates quantifiable evidence. Instead of saying, "My manager is mean to me," you can say, "In 18 direct communications over three months, the sentiment was negative or very negative in 14 instances (78%), with patterns of accusatory language and backhanded compliments." That's documented, measurable, and hard to dispute.
This is particularly valuable if you're building a case for HR or legal documentation. Emotional subjective feelings become objective linguistic patterns.
Try this: Take 10 recent emails from your manager. Paste them into Claude or ChatGPT and ask: "What's the emotional tone or sentiment of each message? Are there patterns in how negative, neutral, or positive they seem? What words or phrases carry emotional weight?" Compare your emotional reaction to what the AI identifies—you might realize you're absorbing more negativity than you consciously recognized.
Peri can explain this concept, give practical examples, help you decide whether it applies to your situation, or recommend a journey if appropriate.
Explore related journeys or tell Peri what you're working through.