True Resolution Rate - Telecom (TRR-T)

GMI-TEL-011TelecommunicationsPilot-ReadyPattern O-9: Skill Atrophy

Definition

Percentage of virtual-agent-resolved interactions where the subscriber's problem was actually resolved, separating true resolution from deflection that resurfaces through other channels.

The Measurement Gap

Measurement Gap and Decision

What TRR-T Measures: TRR-T calculates the percentage of virtual-agent-resolved customer interactions where the subscriber's problem was actually solved, after excluding abandoned sessions, re-contacts within 14 days, and channel switches to human agents.

Decision Supported:

Is the virtual agent genuinely resolving customer issues, or is the deflection rate overstating actual resolution? Without TRR-T, CX teams report First Contact Resolution (FCR) based on what the chatbot marked as "resolved," which includes sessions the subscriber abandoned in frustration, sessions where the subscriber called back two days later with the same problem, and sessions auto-closed by timeout.

Why This Is Unmeasured:

FCR is the universal metric for virtual agent performance, but FCR counts any interaction not escalated to a human as "resolved." The gap between FCR and actual resolution is the "deflection illusion". It makes the virtual agent look productive while hiding silent dissatisfaction. Measuring TRR-T requires joining chatbot session data with CRM re-contact records and cross-channel interaction logs, which crosses the boundary between digital channel management (which owns the chatbot) and traditional service operations (which owns the contact center). Most operators have both data sets but have never joined them to answer this question (Forrester, "The Deflection Illusion: Virtual Agent Resolution Quality in Telecom," 2024; COPC, "Customer Experience Standard: AI-Assisted Resolution Measurement," 2023).

Formula

Plain Language: Of all interactions the virtual agent marked "resolved," remove three categories: (a) re-contacts by the same subscriber on the same topic within 14 days, (b) channel switches where the subscriber contacts a human agent on the same topic within 14 days, and (c) auto-closed sessions with no subscriber confirmation of resolution. What remains is true resolution.

Notation:

TRR-T = (VA_resolved − Recontact − Channel_switch − Auto_closed_no_confirm) / VA_resolved × 100

Variable Definitions:

  • VA_resolved = interactions marked resolved by the virtual agent without human escalation (count, dimensionless)
  • Recontact = subset of VA_resolved where the same subscriber contacts on the same topic category within N days (count, dimensionless)
  • Channel_switch = subset of VA_resolved where the subscriber contacts a human agent (phone, store, email) on the same topic within N days (count, dimensionless)
  • Auto_closed_no_confirm = sessions that timed out or were auto-closed without explicit subscriber confirmation that the issue was resolved (count, dimensionless)
  • N = re-contact and channel-switch detection window (recommended starting point: 14 days)

Units:

Percentage. Compare TRR-T to FCR: the gap between FCR and TRR-T is the deflection illusion.

Sign Convention:

Higher is better. TRR-T of 100% means every "resolved" interaction was genuinely resolved.

Edge Cases:

  • Same subscriber, different topic: requires topic matching at the category level (billing, technical, account) to avoid false positives in the re-contact count.
  • Subscriber confirms resolution but the fix was a workaround (e.g., "restart your router"): TRR-T counts this as resolved from the subscriber's perspective. Track workaround rate separately if needed.
  • Minimum sample: 500 VA-resolved interactions per measurement period.

Proposed formula, pending validation pilots.

At a Glance

Unit
Percentage
Trend Direction
Higher
Reporting Frequency
Monthly
Category
AI-Driven Customer Value and Retention Integrity
← Sentiment Classification Equity Score (SCES)Defect Severity Grading Accuracy (DSGA) →

Assess Your Organization

The AI-GPS Assessment evaluates your AI governance maturity and recommends KPIs like this one for your implementation roadmap.

Take the Free AI-GPS Assessment