Revenue & Usage Analytics · Wisdom notebook
Ask the follow-up.
Agentic conversational BI · the whole-book scan neither brief runs
Across the whole base, where did ARPU fall more than 5% this month while the subscriber count held?
RKScanned every city × segment × plan combination, kept the ones where ARPU fell more than 5% and the active base moved less than 1%, then ranked by revenue exposure.
Six cohorts qualify, and they are not one story — the New York Gold · Business cell is the largest, but two others are moving for a different reason:
Revenue at risk per month — ARPU down more than 5%, base held
$3.6M/month across six cohorts. The two Business cells are 75% of it.
Split those by revenue component — is it the same cause in each?
RKDecomposed each cohort's ARPU change into plan fee, overage, roaming, VAS and device instalment, then compared the shape across cohorts.
No — there are two distinct causes hiding in one symptom. The Business cells are roaming; the Youth cells are the opposite problem:
Three cohorts, three causes: roaming collapse, overage erosion as bigger bundles land, and a discount cutting into Youth plan fee while their VAS grows.
Break the roaming drop by destination market — is it every corridor, or a few?
RKJoined roaming revenue and outbound session counts to destination market and partner network, indexed each corridor against its own pre-April baseline, and ranked by $ lost.
Four corridors carry 82% of it — every one an outbound business destination:
Roaming revenue change by destination market · Business subscribers
Four destination markets carry 82% of the fall, and they are the four the corporate travel policy named. Inbound roaming is flat and no partner's tariff changed — so this is our own subscribers travelling less, which no wholesale renegotiation or retail price move would recover.