Audit the sentiment of donors calling in and saving the losses
Problem :
30% of donors dropping out after calls (not donated in 365 days after the call).
Poor Ratings in Customer Satisfaction.
Long waiting times.
Data Science Challenges :
Huge text conversations, often in paragraphs.
Lots of misspellings, informal words etc.
450+ variables with text conversations. Had to include conversations as well as variables.
Approach/Solution :
Predict in advance before the call or during the call if the donor is going to drop. We will prioritize donors for whom we either expect a significant dip in donation or churn. So less waiting time for them and more senior call centre workers attending their call. We built this in the call centre CRM.
Gave a list of keywords given to employees which either lead to bumping in donation or drastic reduction. The workers with this list can prioritize calls and have a better idea of donation pattern and be more cautious/aggressive.
Match the personality of donors with call centre employees to improve engagement. Like-minded people feel more relatable with each other. We created segments by donor's personality type and matched them with the same personality of workers. Don’t worry about sounding professional. Sound like you. There are over 1.5 billion websites out there, but your story is what’s going to separate this one from the rest. If you read the words back and don’t hear your own voice in your head, that’s a good sign you still have more work to do.
Overview of the process
Results :
$90,000 surplus annual profit as 12% donors not churning after the call now
Customer Satisfaction ratings have gone up by 1.2 points (on a scale of 0-5).
Waiting time reduced by 2.2 minutes on average.
