For CSPs, gen AI bots can boost sales and lower costs

Through work with several communication service providers, we’ve developed best practices for augmenting agents with gen AI-powered bots.

In the hypercompetitive communications market, live agents in sales and customer care play a key role—and face a time squeeze.

Agents field dozens of calls daily and juggle a formidable array of to-dos, from staying up to speed on new products and features to familiarizing themselves with the latest promotions and bundles. This is a challenge for the agents themselves, of course—and a pain point for the communication service providers (CSPs) that rely on them to gain and support customers.

Enter gen AI-enabled chatbots.    

In our work with CSPs, we’re building generative AI-enabled bots that help agents in sales and customer care to manage their time more efficiently and as a result handle up to 20% more calls.

The key is careful planning. While gen AI addresses many telecom needs, integrating it into live agents’ work requires zeroing in on strategic pilots, prioritizing bot accuracy, and doubling down on metrics and customer feedback.

Why conversions are king

Among CSPs’ efforts to make agents more productive, one goal stands out: Converting casual inquirers, or “shoppers,” into customers. Conversions are king, and they’re where inbound sales are won or lost, according to cross-industry research. Of the 2.5 million inbound sales calls that study analyzed, shoppers comprised roughly 40%. Although the shoppers were the largest of all calling groups, sales agents converted just 22% versus 36% of willing buyers.

That’s a gap CSPs want to close, and gen AI bots are the advances they need. 

As gen AI adoption grows, the technology is prodding companies to shift to a more customer-centric point of view, and CSPs are no exception. The bots fill two important functions for CSPs. One is as externally facing tools that engage directly with B2C and B2B customers, answering basic pre-sale questions about products and services and offering recommendations such as custom bundles or VoIP contract details.

These gen AI-driven customer bots provide big advantages over their more rudimentary predecessors. Backed by natural language processing and contextual awareness, the bots generate nuanced responses and even reply conversationally instead of using predefined scripts.

Gen AI bots: The new coach in town

But it’s as internal bots that gen AI really shines. Such bots can play a key role in end-to-end sales and service support—quickly verifying a customer’s identity, analyzing their profile, and then understanding their inquiry in real-time.

Equally important, the bot can act as a coach whispering in the agent’s ear, offering up product recommendations based on the customer’s needs and preferences and allowing the agent to provide more personalized service or resolve technical issues.

For example, the bot might inquire about a customer’s recent iPhone 16 purchase and suggest a pair of Apple AirPods. Or, based on data usage history, the bot could prompt a B2B customer about additional services such as high-capacity data plans: “We noticed your company has been increasing its data usage. Here’s a special offer on our 10 gigabit ethernet.”

Results tell the story 

In our work with CSP clients, we’re seeing gen AI bots produce impressive numbers among live agents in sales and customer care. At several client engagements, the deployment of external and internal bots has freed agents to handle up to 20% more calls. The bots have also enabled the companies to reduce average handle time (AHT) by up to 25%.

The gains add up to big benefits for CSPs: By shaving at least three to five minutes off call times that average 11 to 20 minutes, agents gain 60 minutes per day for additional calls. The added calls per agent mean more opportunities to interact with customers, understand their needs, and offer suitable products or services or resolve issues and convert more leads.

Among CSPs we’ve been working with on bot pilots, sales agents are handling 10% more calls. Combined with a 5% conversion rate, this leads to projected gains of up to 15% in added annual revenue.

Guidance for CSPs integrating gen AI

Like every industry, the CSP sector is still climbing the learning curve when it comes to executing gen AI. Hurdles include governance, data security, and regulatory and compliance. Questions remain about the accuracy of models and the problem of hallucinations.

To provide a solid footing for integrating gen AI into pre- and post-sales support, we recommend CSPs keep the following considerations in mind: 

1. Prioritize internal-facing bots for quick wins. Because they offer the least risk and the greatest returns, internal-facing bots provide the quick wins that CSPs need to confirm gen AI pilots’ ROI and scalability. CSPs can choose from dozens of internal use cases when it comes to gen AI, from field operations and deal pursuit to contract drafting. We recommend making internal-facing sales agent bots a priority for their potential to maximize returns by boosting revenue and reducing costs. External-facing bots can be developed and launched in parallel but typically take longer to ramp up due to the need for additional approval cycles from legal, risk and compliance.
   
   
2. Emphasize bot accuracy. When it comes to sales and customer support, accuracy matters more than ever. For example, an internal bot that mistakenly approves a customer for a discount risk can trigger a negative customer experience and potentially even a customer loss. We recommend setting accuracy targets of at least 95% and launching small-scale pilots that allow controlled testing to reach that target. Invest in rigorous testing programs that ensure bots’ automated responses and predictive sales assistance are laser precise.
   
   
3. Take time to track metrics and ROI, and to gather feedback. Given the continued buzz around gen AI, it’s tempting to rush into developing multiple gen AI bots and put off tracking metrics and customer feedback until later. Our advice: Don’t do it. Take the time to develop metrics to measure performance as well as ROI and business outcomes such as revenue increase, cost-per-call reduction, and net customer satisfaction increase.
   
   Standard key performance indicators work well here. AHT, new calls handled, and conversion rate all provide details CSPs can use to swiftly identify project shortcomings and to pivot their strategies if needed. Some CSPs have fine-tuned their measurements and explored alternative metrics such as upsell rate, deflection rate to AI bot, and agent cost-per-call.
   
   Metrics are only part of the story, however. Collecting customer feedback through post-call surveys, focus groups, and direct feedback channels provides a much-needed window into customer experiences and pain points.

Next steps

Building a solid foundation for gen AI bots not only eases the time-management squeeze for agents but also ensures CSPs are prepared to scale the effort—and net a healthier return on their investment.

When the deployment of pre- and post-sales support is backed by a careful strategic plan, CSPs are ready to expand the use of gen AI bots across functions such as technical support, proactive network monitoring, and field operations.