How a Leading Telco Increased Call Deflection by 24% with AI-Powered Language Experiments

Telco providers facing a rapid spike in calls to their contact centers have an underused and powerful tool they can use to relieve pressure on agents and increase digital self-service—the most effective words. 

That’s what a European telecommunications provider found when it leveraged AI and machine learning to fine-tune voice digital assistant (IVR) language and encourage customers to use self-service payment channels. Including the date when the invoice was due in the IVR prompt gave the message emotional urgency that motivated action. When coupled with clear instructions, the right language helped the telco achieve a 24% increase in self-service adoption.

When coupled with clear instructions, the right language helped the telco achieve a 24% increase in self-service adoption.

Millions saved, while containing churn

The average customer service call costs more than $2 per minute when it involves human-to-human conversation, according to MetricNet. This adds up to significant costs for telecommunications providers. High call volume also results in long wait times and abandoned calls. In good times, these dynamics negatively affect Net Promoter Scores (NPS), customer satisfaction (CSAT), and retention. As COVID-19 spreads across the continent, the need to contain the gains made in migrating service volume to digital channels over the past decade, and continue to shift customers to self-service, is no longer just a promising way to reduce costs–it’s become an urgent, structural necessity for managing higher call volume with a remote call center workforce.

Only 30% of past-due customers complete payment through self-service

The European telco was looking for a way to reduce call center costs while still providing exemplary service. The company pursued dual goals of increasing payment of overdue invoices and increasing use of self-service options. Historically, nearly 70% of the telco’s customers who heard the IVR prompt about an overdue bill followed the prompts to pay it, but only about 30% of those customers completed the payment through self-service channels.

Learn more: The Changing Shape of the Telco Customer Journey During and After COVID-19

The right words increased self-service adoption by >24%

Persado conducted AI-powered language experiments to identify the best way to motivate customers to pay through self-service. This experimental design approach explored and tested different language combinations that tapped into different emotional motivations, including Trust-based emotional language such as Gratitude (“We appreciate your prompt response”) and Safety (“You haven’t made a payment yet, but I can help you”).

The winning version added the date when the invoice expired to the telco’s original IVR script to tap into the customer’s sense of Urgency.

The winning version added the date when the invoice expired to the telco’s original IVR script to tap into the customer’s sense of Urgency. The Persado version also added clearer instructions on what to do next—“Please say YES to continue”—so the customer could immediately act on that urgency. These adjustments led to 24% more customers making their payment without human help, saving costs while ensuring better customer experiences.

What telecommunication leaders should do now

It’s time to rapidly sharpen digital customer self-service experiences in three ways:

1. Identify high-value journeys where consumers are looking to interact with you
2. Start a Test & Learn program to unlock the power of words
3. Scale best practices while collecting actionable insights 

Using language to increase self-service payments and lower contact center costs, all while delivering a better customer experience, is a win-win for everybody.