The problem
The operator was dispatching field technicians for problems that did not require them, with a high no-fault-found rate on truck rolls that were the largest controllable cost after network. A prior attempt to build a tier 2 function internally had failed on recruitment, because the technical profile competed directly with the operator's own Santo Domingo network-engineering hiring.
What we did
- Placed tier 2 in Santiago, a second-city market that does not compete with the operator's Santo Domingo network hiring
- Gave agents read access to line diagnostics and provisioning, aligned to network-operations shifts
- Built a closure-data feedback loop with dispatch reason codes to attack no-fault-found dispatches
- Priced on the outcome the operator needed rather than on average handle time
Results
Technical contacts resolved without dispatch rose from 72% to 92%, avoiding thousands of monthly truck rolls, and the no-fault-found dispatch rate fell 72%. Tier 2 resolution became a new capability at 24% of contacts, and churn among subscribers with a technical contact fell materially. Released field capacity was redirected to a backlogged installation and upgrade queue.
Metrics
Before and after
- Contacts resolved without dispatch+20 pts
- Before72%After92%
- No-fault-found dispatch rate-72%
- Before46%After13%
- Tier 2 resolution shareNew capability
- Before0%After24%
- Repeat contact within 7 days-67%
- Before27%After9%
- Technical CSAT
- 3.2 / 5 → 4.3 / 5 (+1.1)
These metrics are shown as text only, because the client did not provide a comparable starting figure.
Bars compare the starting position with the measured result. Each row is scaled separately because the units are not the same.
Figures relate to the period stated and were provided by the client. They depend on that client's starting point, so they are not a forecast for another operation.
We had tried to build tier two ourselves and lost every candidate to our own network team. Putting it in Santiago solved a recruitment problem we had been treating as a training problem for two years.
What stayed in place
The diagnostic paths, dispatch reason codes and closure-data feedback loop are operator-owned process assets, and the Santiago second-city model has since been extended to two further support functions.