[Analysis of feeling]
Telecom quality of service – churn rate – collection
Telecoms players have a long history of using data. They face a triple challenge affecting several services, including customer service, marketing, sales and collection services.
Improve the quality of service in fixed and mobile telephony, increase the contract renewal rate and reduce the number of bad payers.
An operator in the telephony and telecom sector.
Machine Learning methods have been implemented to identify and detect technical variables that have a high customer impact (what motivates them to call the hotline? To become bad payers? etc.), to detect and correct bugs, pre-detect faulty devices and set up KPIs related to the customer’s perception.
And the results?
Thanks to the detection of bugs and incidents as well as the implementation of very effective Key Performance Indicators (KPIs) calibrated to customer feeling, the operator has registered a marked improvement in the quality of service: a large reduction in calls to the hotline (-15%), reduction in the number of developers working on bug detection (-25%), improvement of customer satisfaction rate (+10 points).
By setting up predictive maintenance processes for pre-detection of faulty equipment, the operator recorded a sharp reduction in the number of incidents, representing 5,000 devices per month (predictive swap).
In addition, this allowed for a reduction in the churn rate (loss of subscribers) and the number of outstanding payments thanks to the identification of the technical factors involved, representing 30% of the global factors.