Bright Pattern

AI call centers shift agency economics from seat-based resale toward outcome-based managed service, because autonomous resolution rates of 40-65% let one operator supervise work that previously required a full pod. The margin is not in the license; it is in the custom routing logic, vertical training data, and human escalation layer an agency builds on top. Agencies that resell the platform alone get commoditized within two renewal cycles.

Resolution Ceiling is the share of inbound contacts a platform closes without a human, and it sets the ceiling on what an agency can charge for a managed support retainer. Vendors publish wide bands: Nectar Desk's AI Voice Bot resolves 40-65% of calls autonomously, while Glassix routes conversations to AI agents by default and escalates only exceptions. The gap between those two numbers is not a product defect, it is the agency's pricing variable. Every point of resolution you add through custom intent trees, vertical vocabulary, and escalation rules converts agent hours into margin. The ceiling also moves: agentic control failures show autonomous systems acting outside intended boundaries, so a resolution rate without a containment policy is a liability, not a saving. Agencies that document resolution rate by intent, with a named human owner for each escalation path, defend the retainer. Agencies that quote a vendor's headline percentage without a measured baseline get repriced at renewal.

Oversight Premium is the margin you can charge for the human layer sitting on top of autonomous call resolution. Vendors publish autonomous resolution rates as headline numbers: Nectar Desk's AI Voice Bot resolves 40-65% of inbound calls without an agent, and Glassix routes conversations to AI agents by default across WhatsApp, email, chat, and voice. The remaining 35-60% is where agency revenue lives, because escalation handling, exception scripting, and vertical-specific training data are the parts clients cannot buy off the shelf. An agency that resells a seat license competes on price against every other reseller. An agency that owns the escalation playbook, the QA rubric, and the retraining loop on unresolved calls sells a retainer tied to outcomes the platform alone cannot deliver. The framework matters because resolution rates are rising, which shrinks the raw labor pool but expands the value of the oversight layer. Price the oversight, not the minutes.

Containment Economics treats the share of conversations an AI voice or chat agent closes without a human as the single number that sets an agency's retainer margin. Every point of containment removes a live-agent minute from the cost base, but the curve is not linear: the first 40 points are cheap, the next 20 require vertical training data and custom workflows, and the last 10 often cost more than the labor they replace. Agencies that quote managed support on a flat per-seat model get squeezed when containment stalls; agencies that price against a measured containment baseline keep the spread. Nectar Desk's AI Voice Bot resolves 40 to 65 percent of inbound calls autonomously, which is exactly the band where margin is decided, and Bright Pattern's white-label deployment lets a BPO agency resell that containment under its own brand rather than handing the client a vendor logo.

Price the managed service on the human oversight layer (transcript review, escalation design, intent retraining) before you quote the automation rate, because the automation rate is the vendor's number and the oversight layer is the agency's margin.

Treat containment rate as half a metric: demand the escalation rate, the warm-transfer path, and the post-transfer resolution rate before you put a number in a client proposal.

IF a client's call volume is stable enough to forecast monthly minutes and their vertical has repeatable intent patterns, THEN sell a managed retainer where the agency owns routing logic, escalation rules, and training data, priced on outcomes rather than seats. IF the client's volume swings more than roughly 40% month to month or they already run an internal contact center team, THEN resell platform seats and bill implementation hours instead, because the margin sits in configuration, not in ongoing oversight.

The choice here is a delivery-model decision before it is a feature decision: full white-label platforms let an agency put its own brand on the desk and bill a managed retainer, while CRM-native outbound tools fit agencies that sell speed-to-lead rather than a service desk. Margin erodes fastest when the platform's autonomous resolution rate is low and the agency still staffs for peak volume, so model the human coverage cost at the vendor's stated automation ceiling before quoting the retainer. Differentiation comes from custom workflows and vertical-specific training data, not from the platform logo on the invoice.

A productized offer that stands up an AI voice and messaging triage layer for a client's inbound queue, then runs it under agency oversight on a monthly retainer. The agency owns routing logic, escalation rules, and QA rather than reselling a platform license.