When it comes to customer sentiment, voice tonality (pitch and volume of a person's voice) is often used as an indicator of how a customer is feeling.
Your choice of words isn't the only thing that matters when you're communicating with someone — the tone of your voice can make a world of difference in how your listener perceives you. Whether you want to come across as warm and supportive or cold and authoritative, simply changing the inflection in your voice can make all the difference. The volume at which we speak also communicates certain types of emotions, like excitement or frustration.
This can be analyzed to understand how a customer is feeling. For example, if a customer has a higher-pitched voice, they might feel more excited or happy. If a customer has a lower-pitched voice, they might be feeling more frustrated or angry.
Pitch & Volume Are Unreliable Indicators
Imagine you are overhearing a conversation in a language you do not speak, let's say Italian. It gets very loud and passionate very quickly. You get the general gist of the emotional state of both participants, but you are not exactly sure if they are excited due to frustration that cannot be resolved or if they are passionately telling you about a problem that they encountered and need help with.
Voice Tonality Depends On Many Factors
One problem with using voice tonality to determine customer sentiment is that it's not always reliable. Everyone's voice is different. Some people are just naturally loud, and others are more soft-spoken.
The pitch and volume can be affected by many factors, such as the person's mood, the environment they're in (maybe they are in a shared office space or their baby is sleeping), or even the type of phone they're using. This can make it challenging to compare customer sentiment across different interactions accurately.
For us here at MiaRec, we believe that it is precisely that confusion that makes tonality an unreliable indicator for customer sentiment. It shows only half of the picture.
You Can't Teach A Computer Sarcasm (Not Yet Anyway)
Another problem with using pitch and volume to determine customer sentiment is that computers don't register sarcasm. And let's face it — sarcastic people exist. A lot of us use sarcasm as a way to cope with complex customer service experiences. For example, an angry customer might brightly say "Nothing is wrong whatsoever," but their biting sarcasm would go undetected simply because their tone was sunny and calm. So if you're only relying on pitch and volume to understand customer sentiment, you could be missing out on much important information.
Using Natural Language Instead Of Voice Tonality
Instead, MiaRec believes that natural language processing is a much more accurate way to determine customer sentiment.
Language Provides Context
Natural language processing is much more accurate, as it takes into account the words that are being said, as well as the context and progression of the conversation.
For example, if a customer says they are "frustrated", this can be taken differently depending on the context. If the customer has been waiting on hold for a long time, it's likely that they are angry. However, if the customer is talking about a challenging problem they're trying to solve, "frustrated" might simply mean that they're stuck and need some help.
Natural language processing can take all of this into account to give you a more accurate picture of how a customer is feeling. This is important because it can help you understand the customer's sentiment even if they're using sarcasm or other forms of verbal communication that might be difficult to interpret based on pitch and volume alone.
Overall, voice tonality is not the most accurate way to determine customer sentiment as it can be affected by many different factors. Additionally, sarcasm and other forms of verbal communication can be difficult to interpret based on pitch and volume alone.
Language, on the other hand, is a much more accurate way to determine customer sentiment, as it takes into account the words that are being said, as well as the context and progression of the conversation.