ARPU fell eight per cent while people used more data, and one add-on carries all of it
ARPU on Postpaid Gold in New York has fallen to , where it held for the eleven months before. Data use per subscriber rose rather than fell. The base did not lose a subscriber. One stream carries the whole fall: roaming revenue per subscriber has dropped to , and the sessions behind it fell away from the first of April and have not come back.
So what? Revenue falling while usage rises is the one shape a discount cannot fix, and pricing against it would cost margin across the whole segment for nothing. One add-on carries the whole fall. Why those sessions stopped is not in these books, and answering it as a demand problem would be the expensive guess.
$131 vs $142
down eleven thirty a subscriber
$31.50 vs $40.90
$9.40 of the $11.30 fall
1.00 vs 1.00×
rules out a tier step-down
on its own level
the same tiers, the same shares
above its own level
rules out using less
since the first of April
the mechanism behind the money
ARPU by day · New York Postpaid Gold
Fourteen days, against the level the segment held all year.
ARPU on Postpaid Gold in New York has fallen to , from a level it held for eleven months. The base did not lose a subscriber over the same fourteen days, so this is revenue per subscriber falling rather than subscribers leaving.
All three are ruled out by a measurement rather than an argument.
| If this were the cause | The reading that would show it | What it actually did | So |
|---|---|---|---|
| People using less | Data-Attach per subscriber | rose | ruled out |
| Stepping down a tier | Downgrade Rate and Plan-Mix | both unmoved | ruled out |
| Subscribers leaving | the active base | unmoved | ruled out |
| An add-on falling away | Add-On Revenue, roaming | fell hard | this one |
People using less would show in data attach, and it rose rather than fell. Customers stepping down a tier would show in the downgrade rate, and it sits at with the plan mix has not shifted at all. Subscribers leaving would show in the base, and it did not lose a subscriber. All three are ruled out by measurement rather than argument.
The ARPU fall, by revenue stream
One stream carries almost all of it.
per subscriber, per month
Roaming revenue per subscriber has dropped to , which is most of the eleven thirty. The plan fee is exactly where it was. One add-on, on one segment, carries a fall that reads at first like a pricing problem across the whole book.
Add-On Revenue · roaming, by city and segment
Business subscribers, and only business subscribers.
| New York | Boston | Chicago | San Francisco | |
|---|---|---|---|---|
| Business | $31.50 | $32.10 | $31.80 | $30.90 |
| Consumer postpaid | on level | on level | on level | on level |
| Prepaid | on level | on level | on level | on level |
A tariff move reaches the plan fee, and the plan fee is exactly where it was. A retention offer reaches the downgrade rate, and it sits at . What moved is roaming, and the sessions behind it fell away from the first of April and have not come back — so the lever is the bundle, not the price.
$131 vs $142
every subscriber, every month
$31.50 vs $40.90
where the money went
1.00 vs 1.00×
nothing to win back with an offer
Revenue fell and usage did not
ARPU on the segment has fallen to while data use per subscriber rose rather than fell and the base did not lose a subscriber.
Why it matters: Falling revenue with rising usage is the one shape that cannot be a demand problem, which removes the most expensive explanation first.
It is not a tier downgrade either
The downgrade rate sits at and the plan mix has not shifted at all.
Why it matters: A downgrade is answered with a retention offer. This is not one, so that money would have been spent for nothing.
One add-on carries the whole fall
Roaming revenue per subscriber has dropped to , and the sessions behind it fell away from the first of April and have not come back.
Why it matters: It turns a book-wide pricing worry into one bundle on one segment, which is a product conversation rather than a tariff one.
Which destinations did the roaming sessions stop going to?
Says whether this is one corridor or travel in general.
How much of the business base travelled at all this month?
Separates fewer travellers from travellers spending less.
Did the roaming fall start on the same day in every city?
A single date points at a tariff or a policy rather than behaviour.
What do competitors charge on the same corridors now?
Tests whether the sessions moved rather than stopped.
Is any other segment's roaming revenue starting to slip?
Uses the same reading before the ARPU line shows it.