I recently came across a creator post on Instagram that got a ton of backlash in the comments for using generative AI. The post saw high engagement, no doubt. But most of the comments were negative. To make matters worse, the creator was using automated messages to respond to those negative comments!
So even the people who called them out got the same message: “Thank you for your interest! Check your DM.”
It was not a good look.
My point is: a high engagement rate doesn’t automatically translate to a positive outcome for your influencer campaigns. A post can have 10% engagement because everyone’s hating on it in the comments. That’s why performance tracking alone doesn’t tell you the full story. You need to combine it with sentiment analysis to reveal the truth behind those engagements.
Are people really engaging because they love the content? Or are they angry? Frustrated? Doubtful?
In this post, I explain why likes and comments alone could be misleading and how you can use sentiment analysis to get a clearer picture of your influencer marketing impact. Let’s take a look.
Everyone dreams of getting tons of engagement when planning their campaigns. And in most cases, those high engagement rates signal a positive outcome. It usually means people enjoy the content and want to engage with it or even share it with their friends.
But that’s not always the case.
One of the easiest ways to tell when your high engagement could signal a PR crisis? Look at the ratio. Are you getting “ratioed” in the comments? This basically means you’re getting more comments than likes. And it’s a simple “tell” that indicates that people aren’t engaging with the content because they enjoy it.
Instead, they’re engaging with it to criticize it or call you out for something. Even if it’s an issue completely unrelated to the content itself, the comments section is where people usually come to let out their frustrations and their anger.
And these folks won’t “like” the post, which means you’ll see more comments than likes.
Recently, a DC-based creator came under fire for casually filming a dead body and cracking jokes over it as she enjoyed a fancy anniversary dinner with her husband. The internet reacted exactly as you can expect. People were enraged over her insensitivity, calling her out of touch and leaving angry comments in her TikTok videos.
Although the video has since been deleted, people still kept “ratioing” her in her new videos. Most of the comments are hidden, but some videos got twice as many comments as likes.
And the backlash has extended over to other platforms like Instagram, where people continue to call out the influencer for her distasteful content. You can see people leaving hate comments and witty insults from all over the world.
Imagine if this backlash had been associated with a paid partnership campaign. Say the restaurant where the influencer celebrated her anniversary dinner had sponsored her post, expecting to reach local DC audiences and drive foot traffic.
With the kind of reaction her post got, the partnership would’ve turned into a PR nightmare.
Similarly, many brands might rule her out during the influencer vetting process because she’s been involved in such a major controversy. And partnering with her will pose a brand safety risk, as any potential brand partner could be indirectly aligning with her views and values.
What makes it even worse is that the creator refuses to apologize or even address the issue in the first place. She could’ve owned up to it and explained how, upon further reflection, the joke was in bad taste.
Instead, she chooses to be silent and simply keeps posting more content to bury the controversy. But her name will forever be associated with it.
Ratios are the easiest way to tell when your influencer campaign is turning into a PR crisis. But they’re not always accurate. For instance, a campaign involving giveaway contests where participants are encouraged to leave multiple comments could see more comments than likes.
That’s why it pays to dig deeper into the qualitative data to understand how people are actually responding to the campaign.
Sentiment analysis helps you do just that, analyzing the tone, language, and emojis of comments and conversations to understand the sentiment behind them. So you can quickly identify whether an influencer’s post is generating negative, positive, or neutral sentiment.
For example, it’ll detect negative words like “bad,” “sucks,” “disappointing,” or “pathetic” to understand if people are responding negatively to the post. Similarly, the presence of emojis related to anger or disgust could also be flagged as negative.
On the other hand, positive words like “lovely,” “excellent,” “brilliant,” or “winning” could be associated with a positive response.
This makes it easy to spot a negative response trend and be on alert for a possible PR crisis. Instead of manually tracking down every single comment on an influencer’s post, you can quickly detect rising negative sentiment and identify which comments are negative. So you can come up with a response plan much faster before the issue turns into a bigger problem.
Now you may be wondering whether these sentiment analysis tools can really be accurate, especially with the various nuances in human language. This is a valid concern because seemingly negative words don’t always have negative connotations and language is consistently evolving.
For instance, “sick” could mean something that’s really good. The same goes for words like “bad,” “insane,” or “fire.”
Similarly, someone saying, “Great! Another AI-generated slop is just what we need,” is using sarcasm to express their dislike for AI-generated content. But if a traditional sentiment analysis tool just focused on the word “great,” it could flag this as positive sentiment and fail to detect the sarcasm behind the comment.
Modern AI-powered sentiment tools have evolved from simple keyword matching. Now, you can find tools that use natural language processing (NLP) to understand context and detect sarcasm and slang. This allows them to analyze negative vs. positive sentiment much more accurately.
Though still evolving, AI has gotten much better at reading sarcasm and slang, with newer NLP models that can account for the subtleties and context of human language. So they can determine whether someone’s saying “Wow!” sarcastically or whether they really mean it.
Influencity’s technology uses NLP to run semantic and syntax analyses of comments in an influencer’s posts. It then ranks the sentiment behind those comments, allowing you to accurately see whether an influencer is getting praise or backlash.
It even shows you the timeline of sentiments to see how audience responses have changed over time. When did they get unusually high negative sentiment? You can then dig back to their posts to see which posts correlate with those timelines. So it’s easier to pinpoint the reason for their backlash and get more qualitative data on how people feel.
That way, you don’t just rule out the perfect influencer simply because of an online attack from groups you don’t align with.
Perhaps the comments are calling them out for very valid reasons. Or they might be getting hate comments for a stance that you support as a brand.
An influencer that often stirs up controversy may get plenty of engagement, but they’re not the best fit for your brand health. While you can’t always avoid controversy, avoiding polarizing influencers in the first place is an effective first step.
Influencity provides you with powerful sentiment analysis tools to help you spot them during the influencer vetting process. Here’s what you can do:
Influencity provides you with detailed brand safety insights to see if an influencer has received backlash in their comments. It lets you run a brand safety check, which will show you details like:
This overview gives you a fair idea of whether an influencer is “brand safe,” meaning they have minimal history of being involved in controversies. Ideally, you want to work with influencers whose top sentiment and audience sentiment are positive and have a low profile risk level.
Influencity also gives you a profile content analysis that further breaks down the tone and topics of an influencer’s content. It will give you an overview of the sentiment an influencer expresses in their content. You ideally want someone whose content is mostly positive.
It also analyzes the tone expressed in their content, showing you the distribution of tones like:
This is a very helpful graph to understand whether the influencer’s tone aligns with your brand’s identity. For instance, you might want to work with someone who’s usually casual, playful, and sarcastic to reflect your brand’s playful sense of humor.
Influencity also flags an influencer’s posts for safety risks and breaks down the risk rating for potentially sensitive topics like:
So a brand that’s looking for family-friendly influencers will want to steer clear of influencers with any type of flagged post across these topics. But a brand that’s targeting mature audiences could overlook a small percentage of offensive language. Flags for discrimination or hate speech, on the other hand, would be considered polarizing and, therefore, out of bounds.
A one-off spike in negative sentiment doesn’t always signal that an influencer is controversial. People sometimes make mistakes and recover from it. Some apologize and own up to their mistakes, often learning from it. So it’s not the best indicator of whether an influencer is truly polarizing.
Instead, you want to look at the sentiment graphs to get a better picture of the influencer’s patterns. What does their comment sentiment look like over time?
If they consistently receive negative sentiment in their posts, it’s a sign that they frequently get involved in some form of controversy. And you want to avoid working with them.
Traditional influencer vetting looked at profile metrics and rates to assess an influencer’s brand fit. Today, brands are more careful with their influencer partnerships, focusing on creators who are “brand safe” rather than just influential.
Sentiment analysis lets you go beyond engagement rate and provides you with qualitative data to understand if an influencer is getting a positive response from their audience. This helps you narrow down influencers that are great at generating not just engagement but the right kind of engagement.
Platforms like Influencity use NLP to analyze the language and tone in an influencer’s comments section, making it easy to spot polarizing creators when it matters the most.
Sign up for a free 7-day trial to see how it works.
Sentiment analysis in influencer marketing is the process of analyzing an influencer’s comments section and conversations around their name to understand the sentiment behind them. It helps you detect how people are responding to the influencer’s content and whether they’re involved in any controversy.
NLP or natural language processing in sentiment analysis, is a branch of AI that’s capable of detecting the emotions and context expressed in human language. This makes it more adept at detecting sarcasm, tone, irony, and modern-day slang for accurate analysis.
Sentiment analysis often struggles to detect sarcasm since many analysis tools take the literal meaning of words without understanding context or nuance.
A high engagement rate isn’t always good, as it could sometimes result from backlash and controversy.