Circles shows how AI native telco experiences can drive commercial outcomes
A real world telco case highlights how using the OpenAI API and Codex for AI native personalization can lift revenue and reduce churn, with clear implications for business adoption, measurement, and rollout planning.
AI pilots often stall when they cannot prove impact on revenue, churn, or operational effort. One recent telco example, Circles, is explicit about the outcome measures it is targeting, and the AI components it used to build the experiences.
What changed in the approach
Circles built AI native telco personalization experiences using the OpenAI API and Codex. The intent is not just to add a chatbot, but to run personalization as an AI driven product capability that supports the customer journey and the underlying development workflow.
Business results to take seriously
Circles reports three commercial and operational outcomes. It says ARPU increased by 22 percent. It also says churn fell by 9 percent. On the build side, it reports improved development efficiency.
What teams should do next to adopt safely
If you are a business considering similar AI personalization, focus on measurement and workflow fit before scaling. Use the same structure Circles implies: define the business metrics you want to move, map those metrics to the customer workflows where personalization applies, and confirm that the engineering workflow improvements translate into repeatable delivery.
- Pick target metrics up front, then track them over comparable time periods
- Identify the exact customer moments where personalization changes outcomes
- Align the AI capability with engineering delivery, not just front end experience
- Document performance and efficiency results so you can scale what works