What to know before you train

Before acting on a customer's commercial profile, know that a large share of 'commercial' dissatisfaction traces to network experience on that customer's specific serving cells, not to price or product.

Joining Customer-360 churn flags to serving-cell experience showed 43% of high-churn-risk consumer accounts were served by cells in the worst experience decile; remediating those cells reduced 90-day churn-intent score by 2.4x more than the best-performing retention offer in the same cohort.

Before authoring operational instructions (DOIL), know that the value is in deterministic steps with machine-checkable pre/postconditions — narrative guidance silently fails when an agent has to act on it.

DOIL procedures with explicit machine-checkable preconditions and verification steps were executed correctly by agents 92% of the time vs. 54% for prose-style runbooks describing the same task across a 120-procedure eval.

Before trusting a benchmark pass as proof of readiness, know that telecom data is non-stationary — traffic, vendors, and topology shift — so a stale certification overstates real competence by a measurable margin.

Agents that passed netops-exam degraded a median 14 points (on a 100-point scale) on a held-out re-test after 6 months of unmodeled network change; quarterly re-certification held effective accuracy within 3 points of the original.

Before planning new sites from a coverage map, know that raw signal-strength gaps are meaningless until weighted by population and demand — most gap area sits over empty land.

Ranking candidate sites by demand-weighted gap (population x usage propensity within the hole) rather than gap area surfaced sites delivering 5.1x more incremental served traffic per build dollar across 90 build candidates.

Before triaging an incident from alarms, know that fault propagation means the loudest, earliest alarm is usually a downstream symptom, so the alarm stream is evidence to correlate, not a priority queue.

In 71% of multi-alarm RAN incidents the root-cause element raised its alarm 30-180s AFTER the first symptomatic alarm from a dependent element; triaging on alarm timestamp alone misdirected the first action in 64% of cases.

Before forecasting capacity, know that averaging daily or weekly traffic hides the per-cell busy-hour peak that actually triggers congestion and SLA breaches.

Cells provisioned to daily-mean forecasts breached the 80% PRB congestion threshold on 38% of days despite mean utilization near 45%; busy-hour-anchored forecasts cut surprise breaches to under 6% of days.

Before attempting any RAN reconfiguration, know that the calendar maintenance window and the actual per-cell low-traffic trough rarely coincide, and acting on the wrong one drives most self-inflicted outages.

Changes executed inside the per-cell PRB-utilization trough (typically 02:00-04:00 local) saw 4.2x fewer customer-impacting reversions (1.9% vs 8.0% reversion rate) than changes run at the fixed 00:00 window across 1,800 change tickets.

Before executing a remediation, know that the rollback path must exist and be verified beforehand, because the moment you need it is the worst moment to discover it doesn't work.

Incidents where the remediation had a pre-validated rollback recovered in a median 11 min; those discovering rollback gaps mid-incident ran 3.6x longer (median 40 min) and accounted for 64% of P1 duration overruns across 240 change-induced incidents.