Original conceptual editorial illustration.When the Spurs Put AI on the Phone
A missed payment can take several calls to resolve. The Spurs used a voice agent for routine outreach and kept their membership team available for questions that need a person.
The San Antonio Spurs used an AI voice agent named Brianna to follow up with season-ticket members about unpaid balances. Retell AI, the technology vendor, reports that the revenue recovered through payment agreements was 504.7 times the cost of the agent.
Editorial disclosure. Business Meets AI™ is independently owned and published by JGRobo Marketing, Inc. Its founder, Joshua M. Alexander, works professionally with Retell AI. The figures below come from Retell’s customer case study and have not been independently audited.
Reported results
Meet Brianna
The official Retell AI customer story plays above. Watch on YouTube.A member receives a call about an outstanding balance. Brianna uses the available account context and can send a secure payment link by text. For disputes, payment plans or questions that require judgment, a representative takes over.
The member receives an outbound call about an outstanding season-ticket balance.
The case study says Brianna has member and balance information before the call connects. It does not describe how identity is verified before account details are discussed.
A member ready to pay can receive a secure link by text. A dispute, payment-plan request or detailed question goes to a representative.
Why payment follow-up was hard to scale
The Spurs’ Membership Services team manages season-ticket accounts and member relationships. Retell’s case study describes how collecting one balance could take a week of intermittent dialing, voicemails and follow-ups. That time competed with renewals, member service and account questions that needed a representative’s judgment.
The Membership Services team sells season tickets, services existing accounts and maintains relationships across the season. The same representatives who follow up on balances also handle renewals, reactivations and account questions. Retell describes a pattern of dialing, waiting, leaving a message and trying again days later. Even one outstanding balance could take a week of intermittent effort.
The operational cost is the interruption. An unanswered call can create another follow-up task, while a connected call might uncover an issue that needs a person. When both kinds of work sit in the same queue, staff time goes to attempts as well as conversations. Retell says coverage was uneven and some balances aged while the team was trying to reach members.
Why the Spurs tested voice AI
Retell attributes the request to the Membership Services team. John Mansour, the Spurs’ senior innovation manager, describes a team looking for a better member experience and willing to test how voice AI might fit. Chris Monroe, vice president of membership, emphasizes the value of staff time and the importance of giving members accurate, personalized information. Their perspectives point to a specific test: could the team reach more members without asking representatives to spend more of their day dialing?
Dial → Wait → Leave voicemail → Schedule another attempt
Call and log the attempt → Send a link where appropriate or transfer to a representative
How Brianna handles a call
Retell says Brianna starts with account context, including the member’s balance, assigned representative and case history. It sends a secure payment link by text when appropriate and logs call attempts. Retell reads from and writes to Microsoft Dynamics 365 through KORE Software Connect, so call outcomes and follow-ups can stay in the tools the team already uses.
Members with disputed charges, payment-plan requests or more involved questions are transferred to a representative. Retell says the representative receives the transfer with the account and call history available. That handoff lets the team focus on the exceptions without asking members to start the conversation over.
How Brianna Works
Retell AI supplies the voice agent and handles the outbound conversation. Microsoft Dynamics 365 is the CRM named in the case study. KORE Software Connect is the integration through which Retell says the agent reads from and writes to Dynamics. The case study identifies the member’s name, balance, assigned representative and case history as available context.
That context supports two distinct actions. A member ready to settle can receive a secure payment link by text. A representative can take over a conversation that needs judgment with the account and call history available. Retell says call attempts are logged and outcomes and text follow-ups continue in the team’s existing system. The source does not specify the fields written, the payment processor or the identity-verification steps.
What Retell’s figures measure
Retell reports that 13.9% of calls ended in a payment agreement and that 18% connected live. A payment agreement is a call outcome, distinct from a completed payment. The reported 504.7× figure compares recovered revenue with agent cost. Business Meets AI has not audited the underlying transactions, measurement period or costs, and the figure does not establish net profit or a return on the full deployment investment.
An agreement is evidence that a conversation reached a stated next step. A completed collection requires a settled payment. For financial evaluation, connect the outreach record to the payment ledger and track how many agreements settle, how long they take and whether reversals change the amount ultimately retained. The published case study does not provide the underlying counts or enough detail to reconstruct that reconciliation.
The 504.7× comparison uses recovered revenue and the voice agent’s cost. A full evaluation would also need the measurement period, integration and operating costs, staff time for monitoring and escalations, and a baseline for payments that would likely have arrived without the agent. Those inputs would help distinguish incremental recovery from collections the team was already on track to receive. The case study does not provide enough data to calculate full deployment net ROI.
Track attempts and live connections by account as well as by call.
Separate payment agreements, completed payments, disputes and unresolved balances.
Review complaints, opt-outs, transfer completion and whether the representative received usable context.
The Business Playbook
- Choose one repetitive workflow with a clear outcome.
- Verify account records before outreach.
- Confirm calling and messaging permissions.
- Use secure payment and account handling.
- Route disputes and exceptions to a person.
- Test wrong numbers, stale balances, failed links and unavailable staff.
- Review completed payments, complaints and handoffs before expanding.
Use your own call volume, costs and customer outcomes to judge whether the approach works for your business.
What businesses can take from the Spurs’ approach
A business with recurring accounts could start with one measurable workflow, such as routine reminders for a defined group of customers. Track completed payments alongside agreements and live connections. Also measure whether representatives receive useful context on transfers and can resolve exceptions without repeating work.
Membership organizations and businesses with recurring account administration can evaluate the approach when reminders are predictable, records are reliable and enough calls recur to justify the integration and oversight. Start by checking whether the source system can identify the right account, distinguish a paid balance from an outstanding one and supply enough context for staff to take over.
A workflow is a poor fit when balances are frequently disputed, identity cannot be verified safely, outreach permissions are unclear or staff cannot reliably receive escalations. A small queue may also favor manual follow-up. Call volume, contactability, balance mix and existing staff workflows vary by business, so the Spurs’ reported performance cannot serve as a forecast.
Before dialing, confirm which calling, messaging, recording, privacy and consumer-protection requirements apply to the particular audience and jurisdiction. Design for opt-outs, wrong-party contacts, recent payments and disputes. Keep payment collection on an approved secure path and define what the agent can disclose before verification. These are pilot design considerations, not a conclusion about the compliance of any deployment.
Payment outreach depends on accurate account data, appropriate calling and texting permissions, identity checks, secure payment handling and clear handoff rules. Retell says the Spurs reviewed their process before calls began. Businesses evaluating a similar pilot should review the consent, privacy and consumer-protection requirements that apply to their own outreach.
Primary source. Retell AI’s San Antonio Spurs customer case study. Explore Retell AI in the Business Meets AI Tools Directory.