Most helpdesks tell you what a customer asked. The empathy engine tells you how they feel about you, and how that has changed over time.
What the empathy engine is
The empathy engine is the part of our AI that works out how a customer is feeling. It runs on every conversation, and it writes what it finds into the Customer sentiment card at the top of the right-hand column.
It gives you a history of how this customer's experience has been over time. It is not a verdict on the message in front of you.
What the empathy engine measures
It works from two sources, and combining them is what makes it worth trusting.
- What people write.
- It reads the content of the messages in the conversation: the words, and the way they are used.
- What people score you.
- It also takes in the ratings customers give you in the survey after a conversation.
Message analysis on its own over-reacts to a blunt sentence. Survey scores on their own arrive too late to help. Together they give you something closer to how the relationship is actually going.
Why a history beats a score
One sharp message from somebody who has been happy with you for a year means something different from the same message from somebody you have already let down twice.
A single reading cannot tell those two apart. A history can. That is why the card shows a row stretching back rather than one number for today.
What it cannot do
- It does not judge your team. A negative reading is about the customer's experience, not about whoever replied last.
- It needs something to read. A new customer with one short message has almost no history, so a first reading carries less weight than a fifth.
- It does not act on its own from this card. Where sentiment does change what Gruvi does, that is covered in Where sentiment changes what Gruvi does.