Service transformation — US fixed-wireless ISP

Client details are anonymised under NDA. Figures are drawn from the engagement’s own analysis.

The brief

Understand the knowledge gaps causing service impairment, and define the service improvements, tech stack, processes and content needed to deliver omni-channel, single-source-of-truth knowledge services — the foundation for scalable, first-class operations, higher CSAT and better margins.

How I approached it

Interviews with senior leadership and a review of current digital services.
Manual sampling of 300 billing, account and tech-support calls, plus automated analysis of 5,560 calls to build a Voice-of-Customer intent breakdown.
Three days of workshops with 1st- and 2nd-line tech support, billing & accounts, truck-roll and tower management, and senior management.
A cost model tracing major call drivers through support channels to find where cost is truly incurred — and to model the impact of digital deflection and call-centre optimisation.
A market assessment of knowledge and chat platforms against single-source-of-truth, omni-channel needs.

What the numbers showed

~0%

digital deflection — no app, weak self-serve

24%

“no fault found” rate on $220 truck rolls

$475k

/ month wasted on avoidable truck rolls (NFF alone)

70%

of L2 time spent walking L1 through documented triage

32,000

calls received per month

5,560

calls analysed to map true customer intent

Underneath the numbers sat a consistent story: scarce measurement and a culture of anecdote-driven, knee-jerk change; engineer-led jargon that didn’t match how customers speak; no single source of truth for tower or customer data; little formalised knowledge or process consistency; and almost no self-serve — so agents were absorbing work the system should have deflected.

What I delivered

A Voice-of-Customer intent breakdown across 12 top-level intents and dozens of sub-intents, each with volume, share and average handling time.
A filter-bed of 58 ranked quick wins, T-shirt-sized by cost and benefit, so the client could act immediately while the bigger build was scoped.
A set of “big win” platform moves: automated triage, a centralised tower knowledge base, improved client comms, and a modern RAG-ML CMS / knowledge base for omni-channel single-source-of-truth delivery.
A cost model of the major call drivers, and a market analysis of candidate knowledge/chat platforms.
The MG12 Service Maturity Model — a framework mapping the operation from infancy to mature across knowledge base, channels, intent, engagement, control/measurement and strategy, giving leadership a shared picture of where they were and where to head.

Why it’s here

This is the proof behind ML consultancy: turning 5,560 calls into a quantified Voice-of-Customer picture and a prioritised, ML-ready plan — RAG-ML knowledge, intent analytics and digital deflection — that a board can fund and a team can execute.