The market for circular mobile devices is projected to exceed $150 billion by 2027, while consumer appetites grow for second-hand, refurbished devices, especially as new premium device price tags continue to rise. This creates a lucrative opportunity for mobile operators, but accommodating the circularity requires intake to increase. Despite increasing openness to trade-in, intake is still lower with U.S. mobile trade-in attach rates between 50 to 60 percent and Europe’s average trade-in attachment rate at around 10 to 12 percent.

While there are higher volumes in some markets than others, the problem across all is that trade-ins are not offered dynamically and not centered around when it’s best for the customer, instead relegated to marketing campaigns (when it’s right for the mobile operator) or opportunistically when a customer visits a customer. Agentic AI can change that.

The Problem is Timing, Not Intent

Most trade-in programs wait for the contract renewal window or fixed promotional campaigns, then send a compelling upgrade prompt to an entire customer segment or advertise on media channels or they might wait for a customer to come into the store. For example, one customer’s device may experience rapid battery drain battery capacity and dropping calls, while another’s is performing perfectly. They are both trade-in targets and can both be converted, but it must be at the right times.

The consumer facing real degradation may not even consciously connect their frustration to an upgrade opportunity. The consumer with a well-functioning device has no compelling reason to act. In both cases, the program might not work. It’s not because of price, but because the intervention wasn’t built around what was actually happening on the device and to the customer.

What Agentic AI Changes

Agentic AI introduces a fundamentally different operating model for mobile operators who want to increase trade-in attachment rates. Rather than relying on pre-fixed timelines or opportunistic engagements for trade-ins, an AI agent continuously monitors the device and turns a trade-in program into a predictive and proactive one.

An agent engages with a customer right on the device, conversing in natural language to initiate a trade-in journey. Running a full objective assessment, the agent can then offer a highly tailored trade-in that matches the customer’s device, profile and interests. Grounded in objective device evidence and the customer’s voice, the offer lands as a solution to a problem the customer is already experiencing, not an unsolicited sales push.

TELUS-MCE’s AI program showed the value of deploying an agent. When consumers were engaged proactively by an agent around device health signals, they were more likely to convert on their offers, one of which included trading in for a certified pre-owned mobile device.

The Fuel: Device Telemetry

What drives the success of the agent is collecting both historical and real-time mobile device telemetry (diagnostic data) and intelligence. This allows the agent to know not just what the device is, but also when to trigger an interaction with the customer to create the trade-in opportunity.

Telemetry includes a variety of key signals that are often associated with the customer’s need to move to a new device – battery health and degradation curves, storage utilisation or software performance patterns, among others.

Without this critical data layer of the customer’s mobile device, an AI agent is relegated to chatbot status, only capable of engaging but not at the right time.

The Commercial Case

Higher trade-in attachment improves refurbishment pipeline supply, supports circular economy commitments and deepens retention. For mobile operators, not only does well-timed, automated trade-in retain the customer, but it also opens more opportunities:

  • Upsell, with 62 percent of trade-in customers in favor of buying more with the savings they get from a trade-in, with a mobile ARPU lift of up to $12 (MCE data).
  • Mobile operators who have single-play customers who only have TV/broadband can use an AI agent to poach wireless customers who are with competitors and add net new subscriptions.

The device signals needed to make this work are already being generated constantly on the customer’s device. The AI capability to act on them autonomously is mature and deployable today. So the real question for operators isn’t whether agentic AI can improve trade-in performance. It’s this: how much longer can you afford to leave that data sitting idle? Explore our AI solutions here.